Siol Mechanics

INDEX
S. No Topic Page No
Week 1
1 Lec 1 : Introduction to course content 1
2 Lec 2 : Stress acting at a point- Cauchy stress 7
3 Lec 3 : Stress acting at a point- Stress tensor 16
4 Lec 4 : Stress acting on a plane 26
5 Lec 5 : Stress acting on a plane example 30
6 Lec 6 : Transformation of stress tensor 33
7 Lec 7 : Stress invariants 44
8 Lec 8 : Relationship between stress invariants 61
9 Lec 9 : Principle stresses and Eigen vectors 68
Week 2
10 Lec 10 : Strain in soils 71
11 Lec 11 : Cause effect relationship 79
12 Lec 12 : Important constitutive relationship 88
13 Lec 13 : 3D to 2D idealization 108
14 Lec 14 : Mathematical formulation plane stress plane strain 117
Week 3
15 Lec 15 : Mathematical formulation axisymmetric 127
16 Lec 16 : Summary of Module 1 139
17 Lec 17 : Basics of shear strength 143
18 Lec 18 : Stress representation 156
19 Lec 19 : Shear strength granular soil - I 167
Week 4
20 Lec 20 : Shear strength granular soil - II 178
21 Lec 21 : Shear strength cohesive soil 188
22 Lec 22 : Shear strength cohesive soil - Stress strain 203
23 Lec 23 : Pore water pressure and Skemptons equation 210
Week 5
24 Lec 24 : Overall pore water pressure parameter 227
25 Lec 25 : Pore water pressure -plane strain-effect of sampling 234
26 Lec 26 : Pore water pressure estimation 248
27 Lec 27 : Triaxial test 253
Week 6
28 Lec 28 : Interpretation triaxial test-UU UCS 270
29 Lec 29 : Interpretation triaxial test-CU 287
30 Lec 30 : Interpretation triaxial test-CD 300
31 Lec 31 : Some additional aspects of shear strength 319
Week 7
33 Lec 33 : Stress path and representation 338
34 Lec 34 : Failure line in stress path 353
35 Lec 35 : Stress path-some common cases- I 368
36 Lec 36 : Stress path-some common cases- II 385
Week 8
37 Lec 37 : Stress path-triaxial test-drained 407
38 Lec 38 : Stress path-triaxial test-undrained 423
39 Lec 39 : Stress path-additional undrained case 439
40 Lec 40 : Stress path-field cases - I 455
Week 9
41 Lec 41 : Stress path-field cases - II 476
42 Lec 42 : Stress path problems 492
44 Lec 44 : Introduction-critical state soil mechanics 513
45 Lec 45 : Introduction-critical state soil mechanics 535
Week 10
46 Lec 46 : CSSM-2 D representation 557
47 Lec 47 : Peak state 579
48 Lec 48 : Soil yielding 598
49 Lec 49 : Cam clay 622
50 Lec 50 : Modified Cam clay 641
51 Lec 51 : Prediction of soil behavior from MCCM 665
Week 11
52 Lec 52 : Prediction of soil behavior from MCCM 682
53 Lec 53 : Strain from MCCM 704
54 Lec 54 : State boundary surface 722
55 Lec 55 : CSSM problems 741
57 Lec 57 : Closure of Advanced Soil Mechanics Course 769
Advanced Soil Mechanics
Prof. Sreedeep Sekharan
Department of Civil Engineering
Indian Institute of Technology, Guwahati
Lecture – 01
Introduction to Course Content
Welcome back. Now, in this part of the lecture, I will try to explain what you will learn from
this course.
(Refer Slide Time: 00:37)
Now, if you see like soil mechanics in undergraduate, once you have learned you know that
when you say soil mechanics, you need to discuss about the strength compressibility and flow
characteristics. So, to better appraise soil as a material, you need to understand these 3
characteristics. But in this particular course, we will be focusing mostly on the strength that is
the strength behaviour of the soil.
The reason is encompassing strength compressibility and flow characteristics, it will be a

humongous task and if we want to get into the details of each, then it will not fit into one
course. So, I have focused only on the strength part in this advanced soil mechanics course.
So, with strength as the focus the course is essentially divided into 4 modules.
1
So, the module 1 is about introduction to continuum mechanics. Now, why continuum
mechanics? Let me make it very clear that for any geotechnical postgraduate who is willing
to work in constitutive modelling or geo mechanics of the soil, he need to have or he or she
need to have the exposure to continuum mechanics. So, the basic knowledge of continuum
mechanics is mandatory for dealing with geo mechanical characteristics and modelling of
soil.
Let me tell you, this module is independent of the subsequent modules it is more like a stand-
alone. The basic aim of this module is to give some introductory briefing on continuum
mechanics. So, that the participants can develop upon it and can easily handle the higher
aspects of geomechanical modelling of soil. So, it is introductory and it is for orienting the
participants towards geomechanical concepts. And the basic contents of this module will be it
starts with the concept of stress at a point and introduction to the terminology stress tensor.
These terminologies are very important, when you deal with the modelling of soils essentially
constitutive modelling, one need to have certain exposure to these terminologies. So, that is
the primary focus of this module. Followed by stress acting on any plane then, we will
discuss about transformation of coordinates. Why because this is needed for understanding
stress invariants which are essential for any problem related to geomechanics. So, we will
discuss about stress invariants.
Essentially, we will focus on deviator stress and mean stress and its implication all of you
would have already known what is meant by deviator stress, as well as mean stress. But from
2
the perspective of stress invariants, we will discuss this more. This will be followed by some
discussion on 3d to 2d idealization in the form of plane stress, plane strain and axisymmetric
condition.
So, you can see that this will be a very introductory level lecture of continuum mechanics.
You would appreciate that continuum mechanics itself will be one semester full course, but
that will not be dealt in this particular module, it is introductory in nature.
This will be followed by module 2 shear strength. Now, the terminology shear strength is
nothing new to an undergraduate student but here we will focus more about stress
representation, what are the different ways of stress representation, shear strength of
cohesionless and cohesive soil a revisit and we will discuss a bit more in detail than what you
have learned during your undergraduate, we have seen that there are different misconceptions
depending upon how you conceive this topic of shear strength.
So, these types of misconceptions we will try to elevate in this particular course and you are
also welcome to give me feedbacks, if there are certain topic which is not in line with your
understanding, you can always discuss with me, this will be followed by pore pressure
parameters. As you will appreciate why soil mechanics is different from its allied solid
mechanics, most important aspect is the presence of void and the influence of pore water in it.
So, we will discuss about pore pressure parameters, which are equally important when you
consider soil modelling. Then it will be followed by interpretation of triaxial tests. All of you
3
would agree with me that triaxial testing is one of the most versatile testing for understanding
the shear strength of saturated soils there are different versions of triaxial testing, but in this
particular course, we will focus only on conventional triaxial testing and its interpretation
essentially.
3 types of tests were unconsolidated undrained, consolidated undrained and consolidated

drained. And the last part of it will be what is the influence of stress history on shear
strength?
The module 3 will be founded on module 2 different types of stress representations is not
enough to have better insight into the soil behaviour during loading, one need to study about
stress spots. So, this will be module 3 importance and representation of stress path. Stress
path for different cases of loading and in that we will be specifically focusing on traction
loading in the laboratory. And the final part will be stress path for different field conditions
only to show you how stress path is beneficial in dealing with certain problems in soil
mechanics.
4
And the last module, module 4 is the most important module of this course, which talks about
critical state soil mechanics and the contents would be critical state concept and its
importance parameters relevant to critical state, 2d analysis of critical state soil yielding
which is very important and prediction of soil behaviour with respect to stress history, you
will appreciate that this particular point is already discussed in the module 2, but here you
will understand how it connects in critical state framework.
Then what are the different elements of critical state model. Introduction of Cam Clay and
modified Cam Clay these are critical state-based models, but it will be dealt only at an
introductory level again. And finally, the closure would be 3-dimensional representation and
state boundary surface based on critical state model. So, these are the 4 modules which we
will be dealing in this course.
5
And these are lists of references you can see that the first 2 reference essentially will help you
to brush up the basics of module 1. Module 2, 3 and 4 follows this list of references. Some of
them are very, very important books. In fact, I have drawn a lot of inspiration from Munira,
Budhu book on soil mechanics and foundation Wiley’s third edition is available now. It is a
beautiful book.
Similarly, Holtz and Kovacs, all books are equally important, but I have laboriously followed
professor Budhu book and I have also taken some help from him in clearing my concepts
related to critical state. So, every book is important. I am just telling you certainly similarly
stress path you can follow Parry’s book it is also a very good book, if you want to learn more
about Mohr circles and stress path. So, these are the lists of reference which you can follow
and any other web materials relevant to each topic.
So, with that I would like to before winding up the course this lecture, I would like to
introduce my course teaching assistance, Mr. Bharat Ratan, who is a research scholar in
geotechnical engineering, Charakho N. Chah, Vineet Gajamer, Tharunkumar, Biswajyoti
Deka and Jishnu Choudhury. It is a vibrant team of students who is helping me in developing
this course. So, before winding up this session, I would also like to extend my thanks to
Centre for Education Technology IIT, Guwahati for facilitating me in doing this course.
Thank you.
6
Indian Institute of Technology, Guwahati, Assam
Lecture - 02
Stress Acting at a Point - Cauchy Stress
So, welcome back in this lecture, we will start the first module introduction to continuum
mechanics.
And the first lecture is stress acting at a point. So, if you see, stress is a very common
terminology, which all of us understand and we all appreciate that this particular terminology
is important in mechanics. Now, what is our conception or visualization of a stress and why
we need to discuss this particular aspect stress at a point. What is that concept all about, why
do we need it, so that is what we will see in this particular lecture and why this is important
that also will be discussed?
Now, action of any external force on a body what is its reaction the reaction is in the form of
some sort of stress that build up. It is same for all like, when we say we are stressed, it gives
an indication of some internal mechanism. It is something with respect to reaction not action.
So, when we are stressed it means, there are certain external factors which is happening
because of which our body is responding. So, there is something which is developed
internally the same with all materials.
7
Which means to say when a material or a body is acted upon by external stresses. So, for
example, in this particular case, it is action of external forces on a body it results in some
response within the body. So, that is what it is within the reaction or the response is
conceived as stress in very simple terms. So, defining stress is important to evaluate
equilibrium condition.
And stability of a body or a material when it is acted upon by external stresses, now why it is
stress at a point so when a body is acted upon by external forces, it is response will depend
upon the constitutive nature of that particular body and constitutive nature means, what is the
built of that body that will influence what will be its response and response here is stress. The
constitutive nature will dictate how the internal forces are mobilized.
So, now, it is not a time that you discuss about stress, when it is acted upon by some external
loads, what develops is the internal force. Now, internal forces which are mobilized within
the body. It will depend upon the constitutive nature of that particular body. So, the internal
force mobilization it will depend upon the material characteristics and the nature of external
forces.
So, what type of external forces are acting, how they are acting, what are the combinations of
external forces which are acting and also the material characteristics. So, these two factors
together determine what will be the response. And that response is the internal force which
gets developed, probably you will understand this, we have when we learned about analysis
of stresses, we have done this we were asked to find out what is the internal force.
8
Within the each of the stress member is similar to that. So, here we are making it more clear
and we are generalizing it. So, the internal force mobilization it will depend upon what sort of
material it is, what is its built as well as the nature of external forces. The manner in which
the internal forces are mobilized governs the deformation of the body. Now, what is the net
effect?
Now, how the internal forces are generated within the body will determine what type of
deformation will it undergo and that is very important for us. So, when a body deforms, so,
what causes its deformation. So, the way in which the internal forces gets developed that will
indicate what sort of deformation the body will undergo. Now, to study these necessitate
some basic requirements in continuum mechanics.
What are those basic requirements? For example, the first 1 is the equilibrium any body has
to be in equilibrium. So, the equilibrium conditions are very well necessary. Then comes the
compatibility condition. Then the constitutive behaviour constitutive behaviour has
something to do with the material properties itself. Then, the boundary condition boundary
condition in the sense what type of external loads are acting and how.
So, to understand or to have these basic requirements, one need to understand what is the
stress that gets developed within the body and to be very specific stress acting at a point. So,
stress acting at a point is mandatory in order to address the basic requirement which has been
discussed above.
9
Now, how do we do this, so before getting into the actual stress at a point, we need to discuss
about Cauchy's assumption, he put forth the hypothesize of internal forces, how do we know
that when a body is acted upon by some forces, there is an internal force which gets
developed. Now, this is based on Cauchy's assumption or rather hypothesis of internal forces,
the entire mechanics of stress at a point is built on this.
To be very specific, what is the assumption stating? There exist, this is very important, there
exist traction on any internal surface of the body based on which laws of motion or other
mechanics can be defined for any internal part of the body or the body as a whole. For
example, if you have a body it is not 3 dimensional, this 2 dimensional, what it means is that,
if there is an internal surface of the body, let us say this is the internal surface.
Now, what it means is that, there is a traction that gets developed on any internal surface of
the body, it can be any reference plane based on which the laws of motion can be defined for
any internal part of the body and any internal part of the body itself would confirm to the
whole body as a whole. So, the further discussion what we are going to see will entirely
depend upon this assumption.
Now, this is very important for defining the mechanics of deformable bodies. So, any body
which undergoes some load will try to deform. So, this particular aspect put forth by
Cauchy’s assumption is important for defining the mechanics of deformable body. Let us say
the equilibrium condition consider a deformable body under the action of external forces Fi
let us say why it is Fi because you have F1, F2, F3 and F4.
10
Now, this is a typical deformable body. Now, there is a plane which this deformable body is
cut into, that is plane A cuts the body into 2 halves 1 and 2. So, you have a deformable body
and a plane A cuts the body into 2 halves that is 1 and 2. Now, according to Cauchy’s
assumption, it states that there will be a surface traction, traction means a kind of stress or a
force there will be a surface traction on the surface of both the halves.
So, Cauchy's assumption state; that there will be a surface traction on the surface of both the
halves there. So, if you consider half 1 or half 2, there will be a surface traction which gets
developed and law of mechanics applicable to both the halves. So, if you consider half 1 or
half 2, it is one of the same the overall response of the body can be defined if you consider
half 1 or half 2. And what is more important and what need to be stressed here is there is a
surface traction, there is a traction that gets developed on the surface which is cut by the
plane A so, that is what we need to focus.
So, now, we will discuss more about surface traction which is represented by small t and
now, we are considering only half 1, so, this is half 1 F 3 F4 is there and the plane A is there.
So, this half of the body it is still in equilibrium under the action of internal force acting on
the plane A. So, you have external forces F3 and F4 which has indicated it also has some
internal force that gets developed which keeps the body in equilibrium.
The same is identical to half 2 as well. So, when it joins, it becomes the whole body it comes
consider the point x, y, z on the internal surface now, on the plane A we can consider any
11
point P which is denoted by P x, y, z and that is on the internal surface A. Now, consider dA
this is dA which is a small infinitesimal area in the neighbourhood of P so it is very close to
P.
And you consider a small area dA or infinitesimally small area dA in the neighbourhood of P
now, dF is the total force which is acting on dA let us say this dF is the total force which is
acting on dA and the most important thing and which all of you know is how to define a
plane now, we have plane A with us now, how will you define or how will you identify this
plane, we identify the plane using normal acting to that plane.
So, n is normal to the area dA and what is dA? It is the tangent plane at point P why tangent
because you are considering a point P so, normal should be with respect to that tangent. So,
that is why it is called tangent plane. So, n is the normal to the area dA, which we are
considering and there is dA is in the neighbourhood of P so, that is what we understand now.
Now, the traction vector t, is defined at point P on plane having normal n by
Now, this is similar to the definition of stress. Now, if you see here, traction vector t is
expressed as a function of x t and n what does that mean, traction vector t it is the function of
position, it is a function of time and it is the function of plane n is normal to the plane.
So, in a sense the traction vector t is a function of position. It is a function of time and plane.
So, it is always identified by some point in this space. So it is not as it is some position, it is
also defined by time, because you are considering a deformable body. So, what deformation
does this mechanics correspond too so, that is why time becomes important and it is also
dependent on which plane it is acting. So, this plane is identified by n. So, traction vector t is
a function of position x time t and normal.
12
Now, we can see that the traction vector t is always associated with a plane problem is there
are infinite number of planes passing through the point. So, what happens is, if you consider
point here you have different possibilities of planes. Now, if traction vector t is associated
with a plane, so which plane are we going to consider, so this is not a proper preposition. So,
is there a possibility of finding out a particular quantity which is a substitute or which
represents the traction vector?
At the same time, it defines the intensity of force at a point because traction does it, but
traction is associated with a given plane. Now, there are different planes now, can you find a
quantity which is not dependent on plane, but at the same time it would represent the internal
force. So, it should be a function of point x and time t and very specifically, it should be
independent of plane.
Is there any quantity as such this is very much possible this is defined by Cauchy’s stress σ
and represented by what is known as Cauchy’s formula and what is Cauchy’s formula?
It is in this manner,
Where, t is traction vector, is known as Cauchy’s stress, because the whole of the
formulation is based on Cauchy’s hypothesis and assumption. So, because of that, is known
as Cauchy’s stress and , this is very important and one should keep in mind for all
further discussion, so, this Cauchy’s formula and Cauchy’s stress these are the two terms
13
which we need to keep in mind further, so, what is the idea the idea is about finding a
quantity, which is not dependent on plane but it still represents the internal force.
Now, σ or Cauchy’s stress is such a term now, what is that so, σ is a function of x and t
whereas, traction is a function of all the three. So, we have eliminated the contribution of
plane here. So, traction t in this it linearly depends on A you can see in the expression and σ
transforms n vector to t vector. So, that is what it is according to Cauchy’s formula. So,
Cauchy’s stress σ defines the internal force intensity at a point for a given state of
deformation please keeps this mind for a given state of deformation.
That is why the time factor becomes important. So, this σ defines the internal force intensity
and that is what we want. And according to Cauchy’s formula, it ensures that the σ defines
the internal force intensity which traction also did at the same time it is independent of the
plane. The discussions related to Cauchy’s stress is taken from Singh, A.K mechanics of
solids book, this book I have listed in the reference.
It is a very good book for beginners in continuum mechanics. And it explains well so, most of
the discussions which I am following in module 1, especially in the initial half is from this
particular book you can refer to it.
So, now, we will come to the matrix representation of Cauchy’s stress which is given as
follows now in the matrix form the stress σ, which is known as Cauchy’s stress,
14
The matrix representation of Cauchy’s formula is in this manner.
So, you have traction vector then you have these stress Cauchy’s stress tensor or Cauchy’s
stress and this is direction cosine vector. So, direction cosines these are always associated
with the plane. So, you have . So, this is what the matrix representation of Cauchy’s
stress and Cauchy’s formula.
So, in summary, the knowledge of stress acting at a point is mandatory for modelling the
mechanical behaviour of solid or soils. In our case it is advanced soil mechanics. So, that is
why I have specified the term soils as well internal forces, it is entirely dependent on
Cauchy’s hypothesis. And according to Cauchy's assumption, it is stated that there will be a
surface traction acting on a surface of deformable body.
And the entire definition of stress acting at a point is dependent on this the traction vector t
represents the response of a deformable body to external forces, the traction vector t is a
function of position x time t and normal dependent on plane and Cauchy’s formula defines
stress at a point and popularly known as Cauchy's stress. Cauchy’s stress σ linearly
transforms n vector to t vector. So, this is the summary of whatever we have seen till now.
15
Indian Institute of Technology, Guwahati, Assam
Lecture - 03
Stress Acting at a Point - Stress Tensor
So, welcome back, this is the next lecture on stress tensor. So in the last lecture we have
categorically seen what is cauchy's stress, σ. And we have seen that definition of σ indicates
what is the internal force that gets developed within on a plane or within a body at a point due
to some action of external forces. Now what actually is σ. In fact σ is a stress tensor.
So, cauchy's stress it can be considered as a tensor. So, now we are going to define a new
term what is known as tensor. If you want to study or if you want to do modeling in
continuum mechanics, as the complexity of the problem increases, it is always convenient to
define what is known as tensor. And we have already stated stress is a tensor quantity. Now
what is a tensor? We know what are scalars, we know what are vectors. So tensor is also a
similar kind of quantity. So why tensor, because it is very convenient to express stress as a
tensor.
In short stress itself is a tensor. In simple terms, we can say that tensor can be defined as a
quantity with magnitude and multiple spatial directions. So, possibly you will think like what
is the difference between a tensor and a vector. Vector also has a magnitude and a direction,
but we will see that vector has magnitude, but it will have only one direction whereas in the
case of tensor multiple directions are there.
16
So that is the essential difference and tensor is a more general term. And the subsets of tensor
are scalar, vector and any other tensor of higher order. So different tensors which are
popularly used are that is what I told the first one is scalar the simplest tensor is scalar and it
is called zero order tensor. A quantity which have only magnitude and zero direction, scalar
we all of us know that it does not have any direction it has only magnitude.
So we call it as zero direction. And zero order tensor which is a scalar have 3 0, where 0
represents the number of direction. So 30 = 1 element and that is true, it is merely a number
which shows the magnitude. Scalar is a 0 order tensor. The second one is vector which is the
first order tensor. Vector has it is a quantity which have magnitude and one direction you can
see that vector has only one direction.
Accordingly the number of elements will be 31 which is equal to 3 elements. So if you have x
y z axis you have vector in 3 different directions. So that is possible. So that is what it means
it has 3 elements. So it has one direction every vector it is associated with only one direction.
And it has 31 which is 3 elements and specifically cauchy's stress is known as a second order
tensor.
Why cauchy's stress σ have magnitude and 2 directions. Now what are these 2 directions?
Now we will see it specifically how these 2 directions come into picture when you define a
stress component and it is very easy also if you remember cauchy's stress, we represent it as
σxx or σxy. So there are 2 symbols associated and that is why it is always associated with 2
directions it is associated with on which plane it acts, that means the normal to that particular
plane.
It is also dependent on which direction that particular traction acts. So we will discuss that a
bit later, only to specify here is cauchy's stress is a second order tensor. It has magnitude and
it has 2 directions. So this is second order tensor and it has 3 square equal to 9 elements
which we have already seen in the cauchy's stress tensor. There are 9 elements and the second
order tensor it linearly maps 2 vectors that also we have seen. We have seen that . So
it linearly transforms to vectors that is cauchy's formula.

17
Now, some aspects of tensor to be very specific, this may not be useful, but then this is
important to understand the tensor. Let us say there are 2 vectors u and v a tensor T is a
second order tensor if it linearly maps vector v to u as can be shown here v and the
second order tensor satisfies the properties of linear transformation. So this is what has been
written, t maps v to u or there is a linear mapping of v to u.
If you compare this with cauchy's formula, it is more or less the same thing that is how we
define cauchy's stress tensor is a second order tensor. Having said that, now the next job is to
interpret the components of Cauchy's stress tensor. We know that there are nine elements.
Now what are these nine elements. What does it represent?
18
So, for that we need to define the Cartesian coordinate. So you have a Cartesian coordinate x
y and z. And to make it simple a control volume is also shown, control volume is a very
common terminology which is used in continuum mechanics or any other form of mechanics.
In fact, this control volume is not required, but to make things simple and for one to
understand it has been shown.
So we have a Cartesian coordinate x y z. So you can consider positive x plane, now x is an

axis what is meant by x plane. It means the plane on which x direction is the normal to that
plane that is what is written here the plane whose normal is in positive x direction. So it is
called positive x plane.
So what will be negative x plane. The negative x plane will be here, because the outward
normal to this plane is in the negative x direction. So this is negative x plane. So you need to
understand this very carefully. Consider positive x planes, so we are talking about this
particular plane. That is positive x plane why because normal to this particular plane x this is
the yz plane, this is y, this is z. So this plane is yz plane.
Now for this yz plane the normal is in the direction of x. So that is what it means. So x plane
means, positive x plane means, yz plane which is shown here so, positive x, negative x plane
both are there. Now we will come back to cauchy's formula and cauchy's stress σ. So now,
normal vector to x plane this positive x direction. Please understand the normal vector to x
plane. So this is the x plane.
The normal is x. So you can easily write what is the normal vector. So this is the normal
vector. So for x direction it is [1 0 0]. So nT is given in this manner and similarly for y it will
be [0 1 0] and for z, [0 0 1]. So normal vector to x plane is defined that is n transpose is
given. Now what are the components of traction vector t x, ty and tz. You already have this to
be, that is tx, ty, tz is equal to σ and n.
So if you substitute the value of n that is for positive x plane [1 0 0] here, so it will be [1 0 0],
and do the matrix operation, you will see that t x = σxx, ty will be equal to σxy. So this is σxx,
σxy. So ty will be equal to σxy and tz is equal to σxz. So what does it mean, it means that the
components which are present in cauchy's stress tensor these are components of traction
vector in a given direction.
19
So if you see, you can see that σxx is the x component of traction vector on x plane. So there
are 2 references which are coming and that is why we said that there are 2 directions. It is the
x component that is the traction vector is in the x direction and it is acting on x plane. So
there are 2 things which are coming. Similarly, you have σ xy. σxy is the y component of
traction vector acting on x plane.
Similarly, you have z component of traction vector acting on x plane. Similarly, other
components of cauchy's stress tensor can be identified based on cauchy's formula. So that is
what is the meaning of each of the terms which are present in the cauchy's stress, it is nothing
but the components of traction vector acting in specific direction.
So, the component of cauchy's stress tensor in general is σ ij, it is the j component of traction
vector it can be x, it can be y, it can be z. So it is a j component of traction vector acting on i th
plane. So first index i, it shows which plane it is associated with. Which plane means, which
is the normal and j is the direction of that component, direction of the component of traction
vector.
So i is the plane on which traction is considered, j is the direction in which the traction
component is considered. So we can see the overall representation of cauchy's stress tensor.
So first is stresses acting on x plane. Now which is the x plane, this is the x plane. So there
are 2 x plane, this is negative x and this is positive x. So what are the stresses which are
acting, we have σxx in the direction of x. So, all of them are acting on x plane.
20
Then we have σ xy and σ xz. Similarly, on the other side other plane that is negative x plane we
have σ xy, σ xz and σ xx. It is identical but it is on the other side. Then we have stresses on y
plane. Now what is meant by y plane? A plane with y direction as the normal. So you are
talking about this and this. So you have positive y and this is negative y.
Similarly, so in this you have σ yy, which is the direction, in the direction of y and you have
σyx, σ yz, similarly σ yx and σ yz. Then we have stresses acting on z plane, what are the stresses
acting on z plane and what are the z plane, this is positive z plane and this is negative z plane
and this stress is acting as σ zz, σ zx, σ zy. Similarly, here also you have σ zz, σ zx and σ zy. So
these are the representation of the components of Cauchy's stress on a given control volume.
So, all these stress components are acting at a point. Now we need to keep in mind since I
have shown a control volume in the figure, and that is only for understanding how the
stresses are oriented. Otherwise it does not serve any purpose. We need to still understand
that whatever stress components are there in Cauchy's stress tensor, it is acting at a point and
the control volume, the cuboid is shown only to indicate the plane on which it is acting. So
that notion we should not forget. So it is a stress acting at a point. Now having said that, we
need to now define some sort of sign convention of Cauchy's stress tensor.
So the given sign conventions are the traction components on positive plane. So now we have
already marked what is a positive plane. So the traction component on positive plane acting
in the positive direction means the direction of x y z which is in the positive direction, so is
21
positive. So you have positive plane and the traction component is acting in the positive
direction, so it is positive. Similarly, if you have a positive plane and the traction component
is acting in a negative direction, so it is negative.
For negative plane, if the plane is negative and the traction component is acting in negative
direction, so it is positive. And the final case is negative plane traction component direction is
positive direction, it is negative. So this is one sign convention, you can see that there are
numerous sign conventions which are available and one may use at his convenience, but if
you follow one sign convention, you need to follow it for throughout.
So this is one convenient way of defining sign convention, there are assigned conventions
which are available based on movement also, sometimes it may be difficult to understand. So
this is very easy and very easy to define as well, one example is given here. So this is the
positive x plane, and the stresses acting are σ xx, σ xy and σ xz. If you consider σ xx, this is
acting on a positive x plane.
And σ xx is in acting in the positive x direction. So that is why it is positive, similarly σ xy and
σ xz. Now consider the case of negative x plane, if you consider σ xy, this is a negative plane,
negative x plane whereas, this is acting in positive y direction. So negative plane positive y
direction, so it is negative. Similarly, all the stress components sign can be assigned. So this
is the sign convention of Cauchy's stress tensor. So what are the summary that we have
understood till now.
22
There are 3 normal component or normal stresses, σ xx, σ yy and σ zz. You can see that in this
figure, you have σ xx, σ yy and σ zz, these are acting in the same direction as that of normals. So
there are 3 normal components or normal stresses σ xx, σ yy, σ zz or it is merely stated as σx, σ y,
σ z which is a common terminology, which we normally use in mechanics.
There are 6 shear components or shear stresses where (i j)
For (i=j),
(σxx, σyy, σzz)
So these are shear components of traction or shear stresses, it is written either in σ form or in
form. Cauchy stress tensor a second order tensor, quantifies the internal force distribution in
a body at a given position and time corresponding to a given deformation. Why time is
important, because we are considering the condition corresponding to a given deformation.
And internal forces, which that gets developed it follows the basic laws of mechanics.
Now one particular aspect why stress at a point that information is needed is to, define the
equilibrium equation. So it is a application of why you need to have the knowledge of stress
at a point. Now stress at a point is very important to define the equilibrium equation as we
have seen in the beginning, you have seen that there are certain requirements which need to
be satisfied like equilibrium condition, the compatibility condition and so on.
23
Now for defining the equilibrium condition, we need to specify equilibrium equation. I will
not go into the derivation of this equilibrium equation it is a very basic and which is mostly
seen by most of you. So by considering a given control volume, the equilibrium equation can
be represented as follows.
And you can see that the components of equilibrium equations are the stress tensor
components, the only new term is γ.
Where γ is a self-weight or the gravity stress which acts in the z direction or in the vertical
direction and to be very specific stresses are in terms of total stresses in this particular
equation, it is invariably necessary to have the knowledge of stress at a point for defining
equilibrium condition. Now based on equilibrium, we can say that τ yx = τ xy, τ yz = τ zy and τzx
= τ xz.
Therefore, the stress tensor is represented by 6 independent stress components, there are nine
components in the cauchy's stress tensor just because of this condition, we have 6
independent stress components, and they are 3 normal stresses, σ x, σ y, σ z and 3 shear
stresses τ xy, τ yz and τ zx, where τ xy = τ yx. So this is what it is. So that is how it boils down to
6 independent stress components.
24
So the final summary of what we learned in this particular lecture is cauchy's stress, σ is a
second order tensor. The element of stress tensor represents components of traction acting on
3 orthogonal planes according to a given Cartesian coordinate. σ ij means j component of
traction vector acting on ith plane. Stress tensor σ has 3 normal stress components and 6 shear
stress components. But based on equilibrium, there are 6 independent stresses were 3 normal
and 3 shear stresses.
All the stress components are acting at a point which is a very relevant point and which is
very important. The components of σ depends on coordinate axis, please note here as such σ
it is not dependent on coordinate axis, but the components of σ, I mean to say σ x, σ xy those
are the components or the traction vector components, they are dependent on the coordinate
axis. So there is a distinction which needs to be very clear, one should not get confused with
σ as a whole and the components of σ.
σ as a whole it is not dependent on any axis, but the components of σ keeps changing, but the
overall σ representation of internal force remains the same depending on the reference axis,
the components magnitude value keeps changing. Stress tensor σ at any point in the body
defines the internal force distribution of a body. So this is all about this particular lecture, we
will see in the next lecture.
25
Lecture – 04
Stress Acting on a Plane
In the last lecture, we were discussing about stress acting at a point and we have specified
Cauchy’s stress tensor Cauchy’s formula, wherein, the traction vector which is acting at a
point is related to Cauchy’s stress and the normal vector we have discussed this at length. At
the end, we discussed about one of its application of the stress tensor in mechanics is defining
the equilibrium equation. Following that, we will see today like what is the application of
Cauchy’s stress tensor with respect to stress acting on a plane? We will start.
So, the discussion for today is stress acting on a plane we have discussed about stress acting
at a point, but it is some effect of the whole body. Now, if you cut the body into 2 equal
halves or 2 halves by a plane by a cutting plane, we are interested to find out what are the
stresses acting on that plane? So, this is one typical application of Cauchy’s equation and
Cauchy’s stress tensor I am not repeating Cauchy stress tensor, it is a 3 by 3 matrix form,
what is the traction vector t acting on a plane at point P, point P is denoted by x, y and z.
This is the same plane which we discussed in the last lecture, where we evolved the Cauchy’s
stress tensor you can see here, this is the traction vector t. Now, stress tensor at a point P for a
given time t, why time is important, because the stress tensor corresponds to a given
26
deformation of the body. So, stress tensor at point P at time t is known, so, that is known to
us. So, this is represented by σij and the components are given this we have already seen.
The normal vector for the plane P passing through point P is known. Now, this is very
important, why because, the no any plane is always identified by its normal. So, normal
vector n is known. So, for a given plane which is tangent plane at point p and what is the
normal corresponding to that particular small elementary plane that gives the normal n? So, n
is known and the direction cosine or the normal vector n for the plane is known and that is
given as n vector.
Now, the question is what is the traction vector on the plane? So, I mean to say what is the
traction vector t on the plane? What are the normal and shear components traction vector
acting on the plane? So, these are the two questions which we have to discuss today. And for
that what are the minimal information needed as explained?
So, we will refer back to Cauchy’s equation where determination of traction vector t on a
plane is given by . So, this is the Cauchy’s equation. So, these are the elements of or
components of traction vector tx, ty and tz these are the elements which we need to determine
now, once we determine tx, ty, tz it is understood that the traction vector t is defined. now, that
can be written based on Cauchy’s equation as , normal component of traction vector
tn.
27
Now, as we have discussed in the previous slide t is any traction and then we have normal
component of traction vector now, tn is the normal component. So, normal component tn of
the traction vector. similarly, we have the shear component of the traction vector t s. So,
normal component of traction vector tn is obtained by the dot product of t vector and n vector
why because t this component is known.
So, this is known. So, when you take dot product, it is the same protection law that we
already know this is projected onto the normal. So, t vector or the traction vector and normal
vector the dot product gives you the normal component of traction vector. So, similarly, the
shear component which I have already shown here, the shear component of traction vector ts
is given by so, once we know tn.
So, tn is known now, we know that ts because these are orthogonal, we already know that it is
that means, the magnitude of ts that is the shear component of the traction vector ts is given by
Now, this determination of stress is acting on a plane when I say stress is acting on a plane it
means normal stress or the shear stress or if we tell more specifically it is the normal
component of traction vector and the shear component of traction vector can be determined.
So, tn and ts are the required components. So, here what is t? t is the magnitude of the traction
vector which is obtained as
28
So, to summarize knowing Cauchy’s stress tensor and normal vector, it is possible to
determine traction vector on a plane the traction vector can be used to determine normal and
shear component of traction. Similarly, traction component in any other direction can be
determined knowing the direction cosine vector. So, if the direction cosine is known, then
what is the component of traction vector in that direction can also be determined?
As of now, we have determined for normal direction, the normal and shear components of
traction are called normal stress and shear stress acting on a plane. So, whenever we say
stress acting on a plane, we define it in terms of normal and shear stress or the normal
component and shear components of the traction vector itself is considered as normal stress
and shear stress acting on a plane.
The stress acting on a plane is defined in terms of normal and shear stress. So, this is all about
stress acting on a plane and it is a typical example of applying Cauchy’s stress tensor and
Cauchy’s formula for finding out stress acting on a plane.
29
Advanced Soil Mechanic
Lecture – 05
Stress action on a plane example
In the last lecture, we have discussed about stress acting on a plane, we will see a simple
example, to demonstrate how to find out stress acting on a plane by knowing the stress tensor
acting at a point.
So, how do we do this? So, we will refer to 1 example problem given the stress tensor acting
at a point. we are asked to find out what is the traction vector because, without knowing
traction vector, it is difficult to find the stress acting on a plane. So, first thing is to find out
traction vector and then stress acting on a plane whose, normal makes equal inclination with
coordinate axis.
So, now, this particular sentence we need to understand carefully, we are asked to find out
stress acting on a plane now, what is stress acting on a plane stress acting on a plane is about
the normal and shear component of the traction vector. So, that is what we have to find out,
but for finding out we need to know the normal vector. Now, what is the normal vector. So, it
is told that it is a plane whose normal makes equal inclination with coordinate axes we have
learned this in the lecture.
30
So, what is meant by normal making equal inclination. It is similar to that of a space
diagonal, spaced diagonal makes equal inclination with the axes. So, here we have been given
with a stress tensor with its components. So, σ is given. So, normal making equal inclination
with coordinate axis it is same as the space diagonal. Now, what is that it is equal inclination.
So, we know that the normal vector n is equal to , 3, 3. This also we have
discussed in the lecture. So, how do you find the traction vector? So, traction vector t is equal
to σ transpose into n. So, that will give you traction vector.
So, traction vector t is equal to tx, ty, tz which is equal to the stress tensor into the normal
vector. So, do the matrix multiplication one will get t x, ty, tz is that is equal to - 3, - 3,
3. So, the traction vector is known. Now, we need to find out what is the normal stress.
So, how do we get that?
The stress acting on a plane means normal and shear components of traction vector. So, first
we will see how to get , that is nothing but the transpose of traction vector multiplied
by the normal vector.

31
So, that is what is done here, the traction vector t and the normal vector n. So, you do this
simple matrix multiplication, we will get -11/3. So, the normal stress or the normal
component of traction vector is now known. Now, what is left with we are left with the shear
component. So, how do we do that? We know now, what is the traction vector? We know
what is the magnitude of .
So, one can always find out the magnitude of shear component which is given by
where t2 is the magnitude of the traction vector and t2 this term. So, you will get τs equal to
that is 121 / 9, well that will give you 128/9 . So, that is about the normal stress and the
shear stress component, which has been asked. So, what we have done? We have done the
determination of traction vector, knowing traction vector one has obtained the stresses acting
on the plane.
32
Indian Institute of Technology – Guwahati
Lecture – 06
Transformation of Stress Tensor
In the last lecture we have discussed about stress acting on a plane as an example of what is the
application of cauchy stress tensor. So in today's lecture we will see what is mean by
transformation of stress tensor. Why this is important? Why this is important is we know stress
tensor rather cauchy stress tensor it is a 3 by 3 matrix and there are 9 elements we have already
seen that there are 6 independent stress components.
Now this corresponds to the axes σx, σy, σz. Now if I want to determine the stress tensor
corresponding to any other orthogonal set of axes can I do that? So this is explained in today is
lecture. And that is all about transformation of stress tensor or rather I can call it as
transformation of coordinate axes. So one axes transformed to the other. So what is going to
happen to the stress tensor the net effect of stress acting at a point is not going to change that
means the stress tensor as such is not going to change, what is going to change?
The components within the stress tensor will change. Now how it will change that goes by some
rule so that is what we will see in today is lecture.
33
So the focus is to determine stresses corresponding to a new set of orthogonal axes, the stress
tensor acting at a point does not depend on the coordinate axes I mean to say there are 2 things
here stress tensor and the components within the stress tensor, stress tensor is not going to
change depending upon the coordinate axes, I mean to say this some effect of some external
force acting on a body represented by a stress tensor acting at a point that is the internal traction.
And the stress tensor acting at a point that is not going to change because nothing else has
changed but the components would change, the components of stress tensor changes with the
coordinate axes that is what we need to see today. So it is important to compute the stress tensor
components corresponding to any coordinate axes you will understand the importance of this
subsequently.
So there are 2 sets of orthogonal axes one is x, y, z which we already know the other one is x 1,
y1, z1, this is what it is, so x, y and z this we already know now what will be the stress
components corresponding to a new set of orthogonal axes x1, y1 and z1 this is the task.
34
So to explain this, I will give you a clue that it has something to do with the direction cosines
because one needs to be mapped on to the other. So it is quite apparent that it is the direction
cosine which helps us doing this. So when we say direction cosine we should know what is the
angle now that is what has been discussed here, the angle between coordinate axes need to be
known so x, y, z.
So now what is the angle between x and x 1 that is given by xx1 now what is the angle between x
and y1 that is given as xy1 similarly xz1, yx1, yy1, yz1, zx1, zy1, zz1. So all these angles are known
this will be clockwise anti clockwise but the angles are known. So the direction cosine matrix
can be written in this form and this is represented by matrix A. Now what is matrix A, Matrix A
is the direction cosine matrix between x y z. And the new set of orthogonal axes x 1, y1, z1. Now
this is given by cos xx1. I have just translated from here to here the only thing is I have added
cos. So cos xy1, cos xz1 similarly up to cos zz 1. So this is the direction cosine matrix rather the
transformation matrix which will help us to transform the stress tensor components. So the stress
tensor components corresponding to new set of axes is given as now what is the new stress
tensor it is represented by σ1 what is the existing stress tensor that is σ this we know.
Now the new set of axes has been given the direction cosine matrix A is defined so the stress
components or the transformed components of stresses is given by
35
So this is how we determine what is the new stress element. So by doing this transformation we
will be left with new stress components which is given by σ1.
So in the process of transformation, there can be multiple transformations x1, y1, z1 can be any.
So there are different possibilities of transformation and accordingly the stress components in
this stress tensor also would change. Now in the process of transformation of stress tensor there
exists a particular case a unique case where only normal stress acts and the shear stress
components are zero, probably you will get a hint of what I am trying to mean here.
We are trying to say that only the normal stresses are acting and the shear components are all
zero in cauchy stress tensor we know the diagonal elements of course the stress tensor these are
basically the normal stresses acting on any point whereas all other elements other than diagonal
elements they are all shear components. Now shear components are zero that means it is a
diagonal matrix cauchy stress tensor becomes a diagonal matrix.
So one transformation is corresponding to a case where only diagonal elements are present or
only the normal stresses are present and shear stresses are zero the plane on which only normal
stress acts are called principal plane this all of us know this has been taught in undergraduate. So
the plane on which only the normal stress is acting is called principal plane. So the new set of
36
coordinate axes in the process of transformation which gives stress components all diagonal
elements or only normal stresses these are nothing but principal axes.
So the new set of coordinate axes for which shear competence are zero is called principal axes.
So one transformation gives us a coordinate axes corresponding to principal axes or the principal
stresses and any plane which represents that is called principal plane. So we have seen in one of
the transformations that means σ1 which we have just discussed that is the new set of transform
competence σ1 will be a diagonal matrix where shear competence as zero.
Such as such a transformation can be obtained by diagonalizing σ. So it is one of the same

whether you use the direction cosine matrix for transformation in this particular case it is quite
easy because we say we know that only diagonal elements are present. So a given matrix is
diagonalized and in the process only diagonal elements exists. So these are normal stress
components. So such a transformation can be obtained by diagonalizing σ matrix. And the
characteristic equation for diagonalizing is
=0
the equation takes the form as follows
λ3 - I1 λ2 + I2 λ - I3 = 0
Now this is the characteristic equation for diagonalizing the given matrix and here I1, I2, I3 these
are parameters of cubical equation λ is just representative of σ because ultimately you are solving
for σ only aspect is instead of repeating σ here I have used λ.
Now when you solve this cubical equation so there will be 3 roots, roots of this equation they are
known as Eigen values and these gives the principal stresses σ1, σ2 and σ3. So in the process of
diagonalizing we use characteristic equation it is a very common method of diagonalizing the
matrix = 0. Now solving this it results in a cubical equation and solving this equation
we get 3 roots of the equation these are called Eigen values.
And the Eigen values represents the principal stresses σ 1, σ2, σ3. So in essence what will be the
transformed stress tensor so it will be
37
so that will be the transformation that take place. Now coming to parameters I 1, I2, I3. What are
these expressions for parameters I am not deriving this rather the expressions are given?
So the first parameter I 1 =σx+ σ y +σ z is the summation of the diagonal elements of the
original stress matrix I 2 is expressed in this form. So the determinant you can see the summation
of the determinants which will give σ x σ y + σ y σ z + σ x σ z - τxy2 – τyz2 – τzx2. So this is I 2 and I 3 is
determinant of the same matrix which will give the expression
= σ x σ y σ z - σ x τzy2 - σ y τxz2 - σ z τxy2 – 2τxyτyz τzx
So that these are the parameters I 1 I 2 and I 3 in terms of the original stress tensor okay. So I 1 is
explained I 2 is explained and I 3 is explained and there are some meaning associated with these
which will be discussed later.
38
Now what we have done we have diagonalize the given stress tensor for what for finding out
principal stresses and that is nothing but a process of transformation. Now we have 2 sets of axes
one is the original x, y, and z again the next set is σ1, σ2, and σ3 you remember where we started
with now actually according to our earlier discussion σ 1x this angle all the angles should be
known and the A matrix is formed.
And then we go for transformation but here we have done it straight forward diagonalization of
the given stress tensor. Now but this is also equally important what is the angle of inclination
between x and the new set of axes which is σ 1, σ2, σ3 the principal axes what is the angle
between. So angle between x y z axes and principal axes how to find that? Now we know from
the diagonalization or based on the characteristic solution of the characteristic equation we know
we have obtained the Eigen value σ1, σ2, σ3.
Now if we substitute these Eigen values in the equation
{n} = 0
by solving this equation what we will be left out with because we substitute σ 1 in this equation
first solve this equation we will get 3 roots again. That means n1, n2, n3 which is represented by
l1, l2, l 3. So substitute σ1 solve this equation we will get an n vector that n vector corresponds to
l1, l2, l 3 in the same manner substitute σ2 in this equation, σ3 in this equation will get m1, m2, m3,
n1, n2, n3.
39
And these are Eigen vectors for Eigen values σ1, σ2, and σ3. So this n represents the Eigen
vectors. So the direction cosine matrix that can give principal stress metrics from stress tensor
can be written in this manner. So it is one of the same thing. So the transformation matrix A can
be written in this form l1, l2, l 3, m1, m2, m3, n1, n2, n3 which we get from the solution of this
equation and by knowing the Eigen vectors.
So Eigen vectors are direction cosines of principal axes with respect to xyz axes where I 1, I2, I3
are called invariants of stress tensor. Now I told that the parameters I 1, I2, I3 has got some
significance now what is the significance the significance is I 1, I2, I3 these are called invariants of
stress tensor σ. Now what is invariants? These are called invariants because its value is not
dependent on coordinate axes for that matter σ 1, σ2, σ3 is also now independent of any axes it
corresponds to only principal axes.
And here I1, I2, I3 which are the parameters of the characteristic equation these are called
invariants of stress tensor σ please make a note it is specified as invariance of stress tensor σ
because later we will have another set of invariants these are called invariants why because the
value of I1, I2, I3 these are not dependent on coordinate axes, components of σ depends on
coordinate axes.
We have already told several number of times why I am repeating this is that particular aspect
need to be very clear the components of cauchy stress tensor which is in terms of x y z these are
dependent on coordinate axes but principal stresses remain same whatever be the kind of stress
tensor in terms of x y z the principal stresses are going to be the same and they are independent
of axes not affected by coordinate axes. Hence σ 1, σ2, and σ3 can also be called stress invariance.
So we have invariants of stress tensor sigma I1, I2, I3 and we have stress invariants σ1, σ2, σ3.
40
So stress at a point can be fully defined by this cauchy stress tensor σ. The components of σ are
dependent on reference axes transformation to principal stresses is an Eigen value problem we
have already seen that considering principal stresses alleviates the dependency on reference axes
now probably you will understand why most of the problems in soil mechanics geo mechanics or
rock mechanics they are all in terms of principal stresses.
Because you need to handle only 3 stresses and other stresses are 0. But this these 3 stresses σ 1,
σ2, σ3 represents the same aspect of what a normal cauchy stress tensor presence that is the
convenience most of the soil and rock mechanics problems they are dealt in terms of principal
stresses or any other stress invariants because I1, I2, I3 they are also invariants. Similarly, there are
several other definitions for invariants which you can see in the literature.
So, any of these invariants are used in defining the failure criterion. Failure criterion is very
much essential for solving any sort of mechanical problems in geo mechanics or any other
mechanics. Most of the soil and rock mechanics problem they are dealt in terms of principal
stresses. So if all the principal stresses are unique then corresponding principle planes are also
unique please understand the sentence a but carefully.
If all the principal stresses are unique that means I have σ 1, σ2, and σ3 they are all different if that
is the case the principal planes also will be unique this is the normal case that we discuss. Now
41
the next case is if 2 principal stresses are equal. Now this is a very typical case in soil mechanics
in traxial testing where we have a traxial sample this is σ 1 and the confining stress is σ2 equal to
σ3.
So that is how we take. So if 2 principal stresses are equal as in traxial tests then one principal
plane is unique but other two are not. So this will be made clear now, consider this cylindrical
figure you have n1 which is the normal to this plane and n 2 and n3. So these are the 3 planes on
which are the these are the 3 axes corresponding to a given plane based on which we are
explaining this. So if 2 principal stresses are equal if σ 2 equal to σ3 which corresponds to the n2
n3.
That is the plane is always defined in terms of normal so that is why when I say plane the normal
has been introduced. So, n1 is a plane corresponding to this so σ 1. So in this particular case σ 1 is
unique whereas σ2 equal to σ3 what it means is that the plane n 1 plane n1 means the plane for
which the normal is n1. So the plane n1 is unique but n2 and n3 these can be any plane which is
orthogonal to each other only thing is this orthogonality has to be maintain it can be any plane in
this which is orthogonal to each other.
So when σ2 equal to σ3 or when any 2 principal stresses are equal then you will have only one
unique plane the other 2 can be any orthogonal plane if all principal stresses are equal that means
σ1 = σ2 = σ3 is a typical case of a mean stress that we will be explaining a bit later if all principal
stresses are equal then any 3 mutually perpendicular planes are principle planes there is no
uniqueness in that So all are equal. So any set of orthogonal planes mutually perpendicular
planes are principal planes.
42
So the summary of this particular part is if stress tensor is known let us say cauchy stress tensor
then the stress tensor components corresponding to any coordinate axes can be determined by the
process of transformation when we know the transformation matrix which is the direction cosine
matrix one can transform the components here please understand it is components stress tensor
components gets transformed.
Transforming σ into a diagonal matrix give principal stresses that means diagonalizing σ gives
principal stresses diagonalizing σ is an Eigen value problem Eigen values gives principal stresses
that means when you solve for it you get the Eigen values which are nothing but σ 1, σ2, σ3 these
are principal stresses and substituting Eigen values are the principal stresses in the equation we
get Eigen vectors.
And that gives the direction cosine for x y z and principal axes we have also defined invariants of
stress tensor and this stress invariants σ1, σ2, σ3 principal stresses are stress invariants. So that is
all for our transformation of stress stresses or stress components or transformation of coordinates
and what is the importance of transformation of coordinates?
43
Indian Institute of Technology - Guwahati
Lecture – 7
Stress Invariants
Welcome back all of you. So in this lecture, we will discuss about stress invariants. In the last
lecture we have seen transformation of a coordinate axis. We have seen the importance of
transformation. We have a given Cauchy stress tensor with stress components in it and we
have seen this stress components can be transformed to any set of orthogonal axis. The first
reference axis was x, y, z. It can be transformed to any other set of orthogonal axis.
We have also seen the importance of this transformation as one specific transformation gives
a stress tensor with only diagonal elements. What is the significance of these diagonal
elements? The significance is these diagonal elements represent principal stresses because all
the shear components in the stress tensor is 0. Now there are wide variety of use of these
principal stresses in geomechanics.
It is quite convenient in soil and rock mechanics to use principal stresses why because this
stress tensor alone is able to capture the entire effect of the external loading that is happening
on a body. So you need to know only 3 components and those are principal stresses. We have
also seen the process of diagonalization. We have also seen the characteristic equation based
on which you have the parameters I1, I2, I3.
And we term these as invariants or rather to be very specific invariants of stress tensor. Now
we will continue with our discussion on stress invariants because there are certain aspects
which we need to discuss more and the importance of these stress invariants, its geometrical
representation so that we will see in this lecture.
44
Let us start again with stress at a point. There is no harm in doing such repetition because for
those who are exposed to this for the first time it is very important for them to understand. So
let us start with stress at a point. Cauchy’s stress tensor can also represented in terms of
principal stresses. So, they represent the same stress effect. So, consider transformation of a
given stress tensor which is represented by , I is identity matrix.
So we are going to consider transformation of under transformation of coordinate axes.

Now transformation of coordinate axes is very well known to us from the last lecture. Now
we just want to see what will be the transformation of a given stress tensor which is
represented by , I is the identity tensor of second order.. So gives
We are going to transform this particular matrix. Now what do we have to know if we want
to transform this into some other axis? First of all, we should know some reference axis and
we should know the direction cosine matrix which we call it as transformation matrix that is
matrix A. So, we need to know matrix A which is the transformation matrix. So once the
transformation matrix is known for any rotation of axis, this can be to any axis depending
upon the components of A.
So we can write the new set of stress tensor that we get will be σ ‘. Let us assume is equal to
A transpose. is the stress tensor that gets transformed and A this comes from our
previous lecture? Now one can note that A is an orthogonal matrix because A contains the
normal vector. Its direction cosines, so it is A orthogonal matrix and it holds this particular
law
45
Now you come to this particular expression you have A transpose and A so that will give you
identity matrix. So what we are left with, we are left with That is when you do the
transformation you are getting the same stress tensor, we started off with and we ended
up in the transformation it is equal to .
Now we need to notice that in all the 3 diagonal elements are same or they are equal,
so this is the condition. So what does that mean? If you have a given stress tensor where the
diagonal elements only exist and all the diagonal elements are equal., then the transformation
of that gives the same matrix. Transformation means you are mapping it to some other axis.
That means rotation of axis is not going to have any effect on this particular matrix. So that is
what is summarized here. Stress tensor with equal normal components and 0 shear
components will be, this is the most important part, invariant for any transformation of axes.
So if you have a stress tensor with diagonal elements and all diagonal elements are equal,
then this happens to be an invariant as it is already proved.
Because whatever transformation you try to do, whatever be the value of A, it is not going to
change and hence you can conveniently understand that such a stress tensor will be an
invariant and it is not affected by the transformation of axes. This has again some relevance
which we will see in the next slide
46
Again, we will start with stress at a point the same stress tensor. This stress tensor now, now
we have known about Cauchy’s stress tensor, we have exclusively discussed the various
aspects of it. Now we are going to introduce another concept which is very important in
geomechanics which is the stress tensor can be expressed as a sum of something known as
hydrostatic stress and deviatoric stress.
Hydrostatic condition we know they are equal in all direction the something what we know
right from the beginning. When we started off fluid mechanics, we know hydrostatic pressure
it acts equally in all direction rather isotropic and hence we can represent the whole of this
stress tensor as a sum of hydrostatic stress and deviatoric stress which is given by this
particular expression. So where hydrostatic stress is and
So combining these two you will get the stress acting at a point. Now in this, that is the
hydrostatic stress is associated with the volume change of a given material or a body. The
represents the volume change that is happening and σd represents the change in shape or
deformation. So these two aspects have been decoupled. The stress tensor as you know it is a
sum effect of all these.
But we are decoupling these two and writing stress tensor as a sum of hydrostatic and
deviatoric stress where hydrostatic stress part represents the volume change that happens in
the body, only volume change, and the other part that is the deviatoric stress that represents
the volume change that is happening. Now what does this mean? This means that if you are
discussing about shear component or the deformation. What causes deformation?
Essentially the shear causes deformation. That deformation is captured by deviatoric stress.
So we should understand that the component of deviatoric stress is very important in soils
and rocks. We already know because in triaxial testing this component or this stress is very
familiar to us, the deviatoric stress. So what causes shearing? It is the deviatoric stress or the
unequal stress condition that causes shearing.
So deviatoric stress is very important in soils and rocks. Now the hydrostatic means stress is
associated with the stress tensor, how? Now as I told this stress can be decoupled, so let us
see how. The hydrostatic mean stress is given by this expression that is the summation of the
47
diagonal elements divided by 3 okay, so it is the mean stress or average stress or hydrostatic
mean stress associated with the stress tensor is given by (σ xx + σ yy + σ zz) / 3.
And it is same as (σ 1 +σ 2 + σ 3)/ 3 why because both of them are same. The entire stress
tensor can be represented by principal stresses. So the hydrostatic stress tensor can be very
conveniently written as
So one part of the stress tensor we have obtained, one part of the stress tensor is the
hydrostatic stress tensor , where σ m is the hydrostatic stress.
So we can very easily write based on our previous slide whatever we have discussed we can
easily write now this diagonal components are equal. So we can very well write that
hydrostatic stress tensor is invariant and it is not affected by transformation of axes. So one
invariant again we are stressing upon is hydrostatic stress tensor. So hydrostatic stress tensor
is also called volumetric stress tensor.
So these are different terminologies which one should keep in mind. Mean stress, hydrostatic
stress, means stress tensor, hydrostatic stress tensor, volumetric stress tensor all represents the
same thing that is given by and is an invariant, so this we have to keep in mind.
48
So the next part is deviatoric stress tensor. Now what is deviatoric stress tensor? It is given by
σ d which is equal to the total stress tensor that is stress tensor minus the hydrostatic stress
tensor, it is very easy. So one is the summation, so σ d happens to be the difference of stress
tensor and the hydrostatic stress tensor this is what you get, again I do not need to explain it.
So what you ultimately get is
So this takes the form of the original stress tensor itself, only thing is the diagonal element
changes or in the principal stress notation one can write we just replace it that is also pretty
straightforward. Now you can write σ m is already known to us, we have already discussed.
So deviatoric stress tensor and hydrostatic stress tensor is used for developing failure criteria
of different materials. Now it is very convenient to represent the failure criteria in terms of
deviatoric as well as the hydrostatic stress tensor, but to be very specific for materials like
soils and rocks deviatoric stress tensor is very important for defining the failure criteria. We
will not be dealing with the failure criteria, yield criteria in this particular course.
The whole idea is to let you know that these components or this knowledge is important for
understanding or for defining the failure criteria okay. So what we are concerned about is
deviatoric stress, the reason is soil is a granular material and it basically fails in shearing and
shear is represented by the deviatoric stress component because that influences the
deformation. So that is why deviatoric stress tensor becomes important in the development of
failure criteria for soils and rocks.
49
Now we have already seen in the last lecture what are the invariants of stress tensor? We
have seen the characteristic equation. Now this deviatoric stress is very much similar to the
stress tensor itself. So for the same hydrostatic stress tensor one can easily find out what are
the invariants. The same logic of characteristic equation one can use and obtain the invariants
J1, J2, J3. When I discussed I1, I2, I3 I have specified this point, what is that?
I1, I2, I3 are invariants of stress tensor whereas here J1, J2, J3 are invariants of deviatoric stress
tensor, so we need to specify that and the process of obtaining J1, J2, J3 is same as that of what
we have adopted for I1, I2, I3. So, I will not go into characteristic equations here, rather we
will go ahead and write J1, J2, J3. So here σ d is σ 1 – σ m, so this we have already known and
this is written in another form which is equal to S1, S2, S3.
So S1 = σ 1 –σ m, similarly S3 = σ 3 –σ m. So the expression for J1 is S1 + S2 + S3, I1 is σ 1 + σ 2

+ σ 3. So S1 + S2 + S3 can be written in this form, we can see that ultimately J1 becomes equal
to 0. so J1 = 0, it comes from this expression. J2 is expressed as – S1 S2 + S2 S3 + S3 S1. There
is a lot of importance of this stress invariant J2. I mean to say J2 is the invariant of deviatoric
stress tensor.
This J2 has got a lot of importance in failure criteria or in geomechanics, so please pay a lot of
attention on J2. It is also expressed as half S12 + S22 + S32. This expression is very important
which is
50
You will see that this particular invariant J2 keeps repeating in various other forums.
Please note that J2 is very important and we will try to see these expressions again as part of
assignments of this particular course. We will not discuss that right in this lecture, but will try
to include this in the assignment the expressions. Similarly, J 3 = S1 S2 S3. So these are the
three invariants of deviatoric stress tensor.
Now we will go to the geometrical significance of invariants. What do you mean by these
invariants? Does it have any meaning? Does it represent something? So that is what we will
see here. There are different ways in which the invariants are defined and used in failure
criterion. So whatever we are discussing in this lecture it is not conclusive, there are different
ways by which the invariants are defined in the literature.
So, we will see only very specific and the basic ones. Since only three elements are there, it is
always convenient to represent state of stress at a point by principal stresses, we have already
discussed this quite a number of times. Now the σ 1, σ, 2 σ 3 space is called Heigh-Westergard
space. It is just a terminology which we would like to include why because there are certain
other terminologies and the geometrical significance which we have to discuss.
So first of all we will understand σ 1, σ 2, σ 3 axis and the space is known as Heigh-
Westergard space. Now we are going to see what is a space diagonal in Heigh-Westergard
space. So space diagonal you can see that now this is the orthogonal axis where σ 1, σ 2, and σ
51
3 is discussed. There is a line OD, it is any such line OD in space, so it is the line which
makes equal angle of inclination with all the principal axis.
So OD is a line which makes equal inclination with σ 1, σ 2, σ 3 axis. We can always find one
such line which is equally inclined with σ 1, σ 2, σ 3. This line is known as space diagonal OD.
So what is space diagonal? It is same as hydrostatic line. So a space diagonal or hydrostatic
line is a line which makes equal inclination with σ 1, σ 2, σ 3 axis.
So OD makes equal inclination means the inclination is with σ 1, σ 2, σ 3 axis. You

will be wondering how comes because the normal vector in σ 1, σ 2, σ 3, if it is n1, n2, n3
we know the relationship So we know these are same and equal, n1 = n2 =

n3. So you have , so we can write n = , so that is how it comes, so it is .
So it is the locus of all points for which σ 1 = σ 2 = σ 3. This also is a very important
information. So on space diagonal or hydrostatic line you have also an information that σ 1 =
σ 2 = σ 3. So what is the geometrical significance? So we will come to that again okay.
So what is the geometrical significance of invariants? Now we know σ m ‘ that is the mean
stress or hydrostatic stress which is also denoted by p’, I have specifically introduced this
letter p’ why because in most of the textbooks you will see that the mean stress is written in
terms of p’, so σ m’ is not that common usage but p’ is, you have p q plot so p’ that is what p ‘
represents.
52
So it is (σ’1 + σ’2 + σ’3) / 3. Only thing is that effective stress component has been
specifically mentioned here, it can be in total stress as well. So (σ’1 + σ’2 + σ’3 )/ 3. Now for
OD we know that it is σ’1 = σ’2 = σ’3 that is equal to σ’m or p’. So OD is a line which
represents what σ’m.
Now let us introduce the terminology what is known as deviatoric plane Dev in short. This is
a plane which is perpendicular to the space diagonal. So any plane that is perpendicular to the
space diagonal which is given as Dev is the deviatoric plane. Now let us say M is the point of
intersection of deviatoric plane with the space diagonal OD. Now P is a point on the
deviatoric plane which represents the current stress state. What is the current stress state?
The current stress state is . So this is the current stress state. This current stress
state is represented by the point P and that is in terms of σ 1, σ 2 and σ 3. So now we have
distance OM, we are talking about the geometrical significance. So we are just trying to see
whether there is any sort of significance of these parameters σ’m or J2 for that matter and what
is that.
So we will first see what is the distance OM. So OM = p’ or σm. So it is the component
of stress state along the space diagonal, what is meant by that? If I draw it here OM and this
is P, so what is the component of OP on OM direction, so that this particular component is
what has been talked about. So it is the component of stress state along the space diagonal
and distance OM = p’.
This distance is given by p’ and distance MP = . So both p’ or σ m or the hydrostatic

stress and the second invariant of deviatoric stress has got some geometrical meaning. So first
one is a measure of distance of OM and MP or the J 2 is the measure of distance of MP that is
the distance of the current stress state from the intersection of space diagonal and deviatoric
plane.
So mean effective stress or hydrostatic stress is the measure of distance of the deviatoric
plane from the origin along the space diagonal in principle effective stress space. Please
53
understand that it is specifically mentioned it is measure of distance, it is not equal to the
distance why it is in terms of p’, we will again see how p’ in the assignment problems
when we will discuss that. The second invariant of deviatoric stress tensor it is the measure of
stress state from the space diagonal.
It is the distance of the current stress state from the intersection of space diagonal and
deviatoric plane. So both J2 and p’ has got some meaning. So deviatoric plane is useful while
defining yield criteria or yield function. Again the yield criteria and yield function will not be
discussed in this particular lecture or in this particular course. The idea is the knowledge of
deviatoric plane is important for one to understand what is yield criteria and how it is defined.
Now the next aspect of invariants is octahedral stress. Now what is octahedral stress? Now to
understand octahedral stress one need to understand what is space diagonal that is just what
we discussed. Octahedral plane is a plane which is equally inclined to principal axis. Can you
relate this statement with our previous discussion? Now octahedral plane, now plane cannot
have any inclination, what can have?
The normal to that particular plane can have an inclination. So octahedral plane is a plane
which is equally inclined to principal axis. For a given coordinate axis there are eight such
planes which is given by in this figure. You can see that this makes 8 planes and all the 8
planes together it is called as octahedral plane and this is σ 1, σ 2, σ 3.
54
So each of the plane there is you can draw a line which is normal to this plane. That particular
normal when you draw this plane we can always draw normal to this plane, now this makes
equal inclination with principal axis. it is same definition as that of space diagonal. Normal to
octahedral plane makes equal inclination with principal axis. Now this normal is nothing but
the space diagonal that we discussed in the previous slide.
So the direction cosines of the normal making equal acute angle with principal axis is , se
have again seen this in the previous slide. Now what is of importance is we have seen before
any plane, any stress acting on a plane is defined in terms of normal stress and shear stress. In
the same manner, the stress acting on an octahedral plane is defined in terms of normal and
shear component acting on the octahedral plane of the traction vector.
Now whenever we want to find normal and shear component, we need to know what is the
traction which is acting on that particular surface. So normal and shear component of traction
vector that is I am marking it as T octahedral. T octahedral is the traction which is acting on
the octahedral plane. So normal and shear components of traction vector is the normal stress
and shear stress acting on octahedral plane and these are known as octahedral stresses.
So what is meant by octahedral stresses? The normal and the shear component of traction
vector which is acting on an octahedral plane is called octahedral stresses and whenever we
say stress acting on a plane it is normal and shear stress, so this is what it means. So this is
the typical octahedral plane with traction T octahedral acting on it. Now it includes the
normal stress, σ octahedral.
This is the normal to this plane and the stress acting is the normal stress, normal octahedral
stress, σ octahedral and octahedral shear stress which is tau octahedral. Now octahedral
stresses are again used in some of the failure criterion and they are highly relevant in
geomechanics problems.
55
Now let us talk in detail about octahedral normal and shear stress. Again the same plane is
noted here. We have normal vector to octahedral plane, this particular aspect normal vector n
is given by n1, n2, n3 and this can be arranged in this manner
this is coming from the previous slides. So one can always find out what is the traction
component of this particular representation.
So you have T octahedral, where does this comes from? This comes from the Cauchy’s
formula. So the components of the traction tx, ty, tz it can be written in terms of the stress
tensor, earlier we discussed this in terms of Cauchy stress tensor, here we are replacing it by
principal stress tensor into the direction cosine So you get what is the traction component
So one can find out the octahedral normal stress as again T octahedral transpose into normal
vector. You are mapping T octahedral onto the normal, so that is given by σ octahedral. So
octahedral normal stress is obtained as T octahedral transpose into n and that is given by
56
Finally you will get (σ 1+ σ 2+ σ 3) / 3 which is nothing but the mean stress or the hydrostatic
stress. So σ octahedral is same as the mean stress.
Now we have discussed one part of it, we will see what is octahedral shear stress. We know
once we obtain the σoct and we know the traction we can always find out the magnitude of
octahedral shear stress and that is what is written here. Toct2 – σoct2, both are known so one can
obtain Toct and this is given by
This again both the expressions it comes from the previous slide. Again if you expand this
you will get in this particular expression, I strongly advise all of you to work it out and see
for yourself, it is a straightforward exercise so I am not spending time. So ultimately, we get
the expression
Now this closely resembles the expression that we got for J2. So here distance MP we have
written it as , we have already seen in the previous slide. Now this closely resembles J2,
hence it can also be written equal to . So both = . So please try to derive

this, we will again try to discuss this as part of assignment.
57
So here this is the situation deviatoric plane M, the current stress rate P, so octahedral stresses
are also invariants. So we have discussed the relevance of tau octahedral with respect to this
particular distance. Also the σoct is a measure of the distance OM. So hence we can also
conclude that octahedral stresses are also invariants.
Now octahedral shear stress forms the basis of von Mises yield criteria. I have thought of
introducing this why because there are certain definition for deviatoric stress. The definition
of deviatoric stress q which you can see in certain textbook equal to yield stress according to
von Mises yield criteria because this form of expression for q which I am going to discuss is
rather popular. So . This is defined in this manner according to von Mises
yield criteria.
So if you follow this, J2 is known, then one will get the expression for q.
This form of expression of q is important and you will see this in some textbooks. So that is
why the relationship of q and octahedral shear stress just wanted to highlight because this
particular expression comes from this definition and that is according to the von Mises yield
criterion.
So this is also known as equivalent stress or von Mises stress, what q is known as equivalent
stress or von Mises stress. So one can easily find out the relationship between tau octahedral
58
and q because you have the expression for q in terms of J 2, you also have the expression for
tau octahedral in terms of J2, substituting both one can always obtain (Refer Slide
Time: 39:13)
So we will summarize what we have seen till now related to stress invariants. The stress
invariants are independent of coordinate axes and useful in defining failure criterion or yield
criterion. For any materials that you encounter in geomechanics for soils and rocks stress
invariants they are very important and it is a very convenient way of defining a given
problem. Hydrostatic stress or mean stress or volumetric stress is a stress invariant.
I1, I2, I3 are invariants of Cauchy’s stress tensor and J1, J2, J3 are invariants of deviatoric stress
tensor. Mean effective stress is the measure of distance, please underline measure of distance
it is not equal to distance, of the deviatoric plane from the origin along the space diagonal in
principle effective stress space, along the space diagonal means hydrostatic line. Second
invariant of deviatoric stress tensor again it is the measure of current stress state what is the
given current stress state from the space diagonal.
So that is the second invariant means J2. Octahedral stress act on octahedral plane which is
the plane with normal equally inclined to principal axes. Octahedral shear stress is correlated
to deviatoric stress. Now there are different ways of defining invariants for failure criterion
which you will see in geomechanics now all of which are not discussed here. So I want to
make this point very clear the discussion on stress invariants is not conclusive.
59
There are a lot of other aspects which one can discuss, but in this particular course it is a bit
difficult to go into the details of stress invariants. So whatever bare minimal understanding
one should have to start continuum mechanics or geomechanics problems has been discussed
here. So that is all for now. We will see in the next lecture.
60
Lecture – 8
Relationship Between Stress Invariants
In the last lecture, we have discussed about invariance. We have seen the stress invariants.
That means invariants of stress tensor I1, I2, I3 and invariants of deviatoric stress tensor J1, J2,
J3. In this session, we will try to see some of the important aspects of these invariants and its
relationships okay.
So we are now given a principal stress matrix where σ is equal to σ 1, σ 2, σ 3. We have seen
that the principal stress matrix gives the same meaning and representation as like any other
stress tensor in a more simplified form because we have only diagonal elements. We can
easily write what is the mean stress which we know already it is (σ 1 + σ 2 + σ 3)/ 3.
So we can formulate the mean stress matrix or isotropic stress matrix which is given as
. We can always write deviatoric stress matrix because you know the given
stress matrix is decomposed into σ m and σ d which can be written as σ = σ m + σ d. So, one
can always write the deviatoric stress matrix which is given as
61
which is this particular matrix.
So, the characteristic equation just like we have done in the previous case what was the
characteristic equation? It where I is the unit vector. So here characteristic equation
is . So one can easily get J1, J2, J3. So that is what we have done. So we have I1,
I2, I3 and then J1, J2, J3.
So now first we are asked to derive J2 which is the second invariant of deviatoric stress
J2 = (S12+ S22+ S32)/2. So let us see how to derive this. We know the first invariant J1 = S1 +
S2 + S3. What is S1? S1 = σ 1 – σ m, S2 = σ 2 – σ m, S3 = σ 3 – σ m. Now if you take a
summation of this you will get σ 1 + σ 2 + σ 3– 3 σ m.
Now σ m = (σ 1 + σ 2 + σ 3)/ 3. So if you substitute it here this will be equal to 0. So that is

why we know the first invariant J1 = S1 + S2 + S3= 0. If you take the square of J1 you can
easily write S1 + S2 + S3 the whole square which is equal to 0. So expand this. So once you
expand this S12+ S22+ S32+ 2 S1 S2 + 2 S2 S3 + 2 S1 S3.
So if you take 2 outside you can write –2, so if you bring S1, S2 , S3 on the other side you can
always write –2 (S1 S2 + S2 S3 + S1 S3) = S12+ S22+ S32. We also know from the solution of
the characteristic equation just like we had I2 = σ 1 σ 2 + σ 2 σ 3 + σ 3 σ 1, we have
J2 = – (S1 S2 + S2 S3 + S1 S3). By substituting it in the equation we finally get
62
So it is proved. Now these relationships become important towards the end of this lecture
where you may have to use these equations for further proving certain other relationships. So
please make a note of this.
The next relationship is to derive
You would have noticed by this time that this expression is very important because this gives
us the notion of deviatoric stress. So we will see how to derive this. Let us first see what is
(S1 – S2), which is equal to (σ 1 – σ m) – (σ 2 –σ m). So this will give us (S1 – S2)= (σ1 – σ2);
(S2 – S3)= (σ2 – σ3); (S3 – S1)= (σ3 – σ1).
So if you write (S1 – S2)2 + (S2 – S3)2 + (S3 – S1)2 that will give us 2 (S12 + S22 + S32 – 2(S1 S2
+ S2 S3 + S3 S1)). What is this S12 + S22 + S32? We have seen in the previous slide it is nothing
but 2×J2 and (S1 S2 + S2 S3 + S3 S1) is this is –J2. Now on the RHS we get, 6×J2.
So you can easily write J2 = ((S1 – S2)2 + (S2 – S3)2 + (S3 – S1)2)/ 6, (S1 – S2) is nothing but
(σ1 – σ 2). So J2 final expression is
So this comes from because (S1 – S2) = (σ1 – σ 2). Now this equation has been used by various
failure criterion or yield criterion for defining deviatoric stress.
63
And one we have already discussed that in von Mises yield criteria the deviatoric ,
so that is what we have already evolved in the previous lecture.
The next relationship is I2 = S1 S2 + S2 S3 + S3 S1+ 3 σ 2

m . Now here it is the relationship
between the second invariant of stress tensor with the deviatoric stress elements and σ m2. So
let us see how to do this. We already know S1 S2 + S2 S3 + S3 S1. If you substitute that for S1
is σ1 – σ m, if you substitute that and expand you will get
S1 S2 + S2 S3 + S3 S1 = σ 1 σ 2 – σ 1 σ m – σ 2 σ m + σm2 + σ 2 σ 3 – σ 2 σ m – σ 3 σ m + σ m2 + σ 3
σ 1 – σ 3 σ m –σ 1 σ m + σ m2.
I request you to please work it out and see for yourself, it is just a very simple equation and
rearrangement. If you simplify this you can get
Now we know from the characteristic equation of stress tensor Cauchy stress tensor I2 = σ 1 σ
2 +σ2σ3+σ3σ1
I2 can be substituted. So this is replaced by I2 and then what is σ 1 +σ 2 + σ 3, that can be
written in terms of 3 σ m. So σ 1 + σ 2 + σ 3 = 3 σ m. So if you substitute it here it will become
–6 σm2+ 3 σm2. So you can write now this is–3 σm2.
So I2 = S1 S2 + S2 S3 + S3 S1 + 3 σ m2.
So this is also equally important, we will consider this equation further.
64
The next relationship is between the third invariant of stress tensor associated with the
deviatoric stress elements. So I3 = S1 S2 S3 + σ m (S1 S2 + S2 S3 + S3 S1) + σ m3. So we will
first start with S1 S2 S3 whose expression is
So by expanding we get
Now by replacing it by suitable expression. We know

I2 = σ 1 σ 2 + σ 2 σ 3 + σ 3 σ 1 = S1 S2 + S2 S3 + S3 S1 + σ m3 which we have already derived in
the previous slide. So this also gives σ m 2.
Now the next σ m 2 = σ 1 + σ 2 + σ 3 – σ m 3 can be written as 3 σ m 3 because σ 1 + σ 2 + σ3
can be replace it by 3 σ m. So 3 σ m × σ 2 m gives 3 σ m 3 – σ m 3 that is 2 σ m 3. So that will
be the relationship. So then you substitute back all these information, you will get
S1 S2 S3 = I3 – σ m (S1 S2 + S2 S3 + S3 S1 + 3 σ m 2) + 2 σ m 3
S1 S2 S3 = I3 – σ m (S1 S2 + S2 S3 + S3 S1) - 3 σ m 3 + 2 σ m 3
By rearranging the terms,
I3 = S1 S2 S3 + σ m (S1 S2 + S2 S3 + S3 S1) + σ m3.
65
Now we have a relationship between the second invariant of deviatoric stress tensor and the
first and the second invariant of Cauchy stress tensor. So let us see how to do this, it is
already proved in the previous slide I2 = S1 S2 + S2 S3 + S3S1 + 3 σm2. So it is already known
to us. We also know I1 = 3 σ m and σ m2 = I12 /9. So it is a simple information.
What is S1 S2 + S2 S3 + S3 S1 that is nothing but –J2 and if we substitute for σ m2 = I12 /9. So
we will get – J2 +3×(I12/9). Simplifying further we get
So this is one relationship between the both invariants.

66
The next expression and the last one is
Let us see where to start from. This we have already proved I3 = S1 S2 S3 + σm (S1 S2 + S2 S3
+ S3S1 )+ σm3. We know that J3 = S1 S2 S3, -J2 = S1 S2 + S2 S3 + S3S1 and σm = I1/3. So
substituting we get
Now substituting J2 expression we get
Simplifying further we get
So that is the required expression. So what we have done is we have discussed about the
relationship between different invariants both for stress invariant and for deviatoric stress
invariant okay. That is all.
67
Lecture – 9
Principle Stresses and Eigen Vectors
In the last few lectures rather we have discussed about stresses acting at a point,
transformation of coordinate axes, determination of principal stresses and the eigen vectors
which represents the direction cosine metrics which will be useful for transformation. In this
lecture, we will try to see a small example demonstrating this.
So we are discussing about determination of principal stresses and normal vector. So given a
stress tensor acting at a point which is σ and the components are given. We are asked to
determine the principal stresses and the normal vector that is eigen vectors for the specific
principal axis. Now we know the characteristic equation is |σ - λ I | = 0 which would
diagonalize the given stress matrix.
If you write that |σ - λ I | will give you lambda I, I is an identity matrix. So
Now if you expand this, you find the determinant, you will get
–λ3 + 18 λ2 – 45 λ = 0.
The last component becomes 0. I strongly suggest you to go through and you do it by
yourself.
68
So solving this equation you will get the eigen values λ = 0, 3 and 15. Now what are these
eigen values? These eigen values are nothing but the principal stresses and you can arrange
this in a diagonal matrix form. So this is σ1 = 15, σ2 = 3 and σ3 = 0 is the principal stresses or
the eigen values. It is an eigen value problem which we have obtained.
So now we are left with determination of normal vectors or eigen vectors. Now if you use this
equation, this also we have explained during the lecture. |σ - λ I |{n} = 0 is the equation. Now
what we have to do? We have |σ - λ I | which is this, the normal vector that is nx ny nz that is
equal to 0. Now for the first we have the eigen values lambda which is already there which is
15, 3, and 0.
So first let us put λ = 0 in this equation and you will get same matrix nx into 0. You will get a
system of equations where 8nx – 6ny + 2nz = 0. Similarly, there are other 3 equations. Now
we already know that the determinant |σ - λ I | = 0. This means that the row or the columns
they are not independent; hence these equations may not be independent equations.
So we may have to use or not we may, we have to use this additional equation
nx2 + ny2 + nz2 = 1
So this we already know. So using these equations, we can solve for nx, ny, nz by putting λ =
0. This will give nx = 1, ny = 2, nz = 2.
69
Similarly, we have to put λ = 3 and 15, the other eigen vectors can be determined. So it will
give you nx = 2, ny = 1, nz = -2 and nx = 2, ny = -2, nz = 1. So this is how we find the principal
stresses and the corresponding transformation matrix okay. So, with this we finish this part of
the problem. We will see in the next lecture. Thank you.
70
Lecture – 10
Strain in Soils
Welcome back all of you. Till last lecture, we have discussed about stresses and in today's
lecture we will see a bit about strain, strain in soils, and when I say we have discussed about
stresses, I would like to make a disclaimer here like whatever we have discussed is just a tip
of the iceberg, like this is only for orienting the listener or the participant of this course
towards the higher limbs of continuum mechanics.
So, whatever we have learned, it is a very elementary discussion related to stresses because in
this course advanced soil mechanics, we cannot go further beyond because we have many
other steps closely related to soils which need to be discussed.
Now in that whatever we have discussed about stresses, it holds true for these strain in soils
as well. So, as you can see much of the concepts discussed for stress is applicable for strain in
soils as well. Now when we discuss about stress and strain, one would wonder, which is
happening first and which is the manifestation of the other. What we know is that we have a
body on which a set of forces or a force act.
71
Now, this application of the force can bring about some changes in the body. Now this
changes one can visualize as the strain that is happening within the body. Now, when such
changes happen, what is the internal response of the body that we defined in terms of stress.
So, both are manifestation of the load acting on the body where you can always visualize
strain that is happening in the body and it is more like it is a reality.
Whereas stress is the internal manifestation, you may not be visualizing stress, but it is there
in the form of response of the body. So, that is how we need to conceive this whole aspect.
Now, deformable body subjected to forces. Now forces it can be body force and surface
traction, we have not categorized these forces in our earlier discussion and it was not required
also, only when we discussed about equilibrium equations we mentioned about body forces,
which becomes very important in soils.
So, geostatic stress or the self-weight stress becomes very important in soil and rocks. So,
deformable body subjected to forces, it can be either body force or traction force, it
undergoes rigid body motion and deformation. So, the two responses are rigid body motion
and deformation. Now, when we say rigid body motion it includes translation and rotation.
We will not get into the details of rigid body motion right away.
We will discuss at length about the deformation part of the deformable body. Now
deformation in general it refers to change in shape and size. Now, possibly you will
understand the implication when I say shape and size because we have already discussed this
in the previous lectures. Size basically it relates to volumetric response and shape it basically
represents the deformation or the shearing aspect of the material.
So, when I say change in size is a volumetric response and it is decoupled. So, we study an
object which undergoes volume change where change in shape is not discussed as a
volumetric part of it and the other one is deviatoric part which we discussed earlier. So,
deformation in general it refers to both in general, but for soil mechanics or for rocks, the
deformation which is associated with change in shape or shearing becomes very important.
Now, strain displacement relationship which ensures the compatibility requirements in the
body. Initially we discussed about stress and we have seen the equilibrium stress equation
that is one part of the story, the other part is strain displacement relationship, which will help
72
to ensure the compatibility. All these elements are very much important for the study of
continuum mechanics.
Now, let us say that this the given material, a compatible material when it is subjected to
different forces is in this manner. So, this it is not failing as such but it undergoes
deformation, hence for the given situation we can call it it is compatible. In this you can see
that certain fractures are being made under the given condition of external forces. Now, this
can be termed as noncompatible.
So, strain displacement relationship helps to assess the compatibility requirements of a

material when it is subjected to different loading.
Now, compatibility is defined using strain displacement relationship. Now, we can say that u,
v, w these are displacement of a point in x, y, x direction okay. So u, v, w for example if I
have a body like this and I am discussing about this point. This point may undergo, it is a
microscopic phenomenon, I am just analyzing it and showing you, so this change this
displacement if this is an x direction, then this corresponds to u, so that is what it means.
So u, v, w are the displacement of a point in x, y, z directions and infinitesimal displacement

of a point. This is the actual displacement and infinitesimal displacement of a point is defined
by some continuous functions and infinitesimal displacement in x, y, z direction is given by
u, v and w. Now, let us consider this given axis x, y and z and this is a given part of the
body. First let us define what is linear strain components.
73
Now, as I told in the beginning whatever we have discussed four stresses are applicable here.
So, I am introducing the term strain tensor and you need to understand it is one of the same or
it is very similar to stress tensor. Now, for that in the stress tensor we had normal as well as
shear components, same way in the strain tensor also we have normal components that we
discuss in terms of linear strain.
So, linear strain components, linear strain components to define that first let us see this is the
point A on the body and displaces by around u, u is the infinitesimal displacement in x
direction. So, u is represented A is displaced to A’, so that is what it means. So, epsilon x
which is the linear stream in x direction is given by u / x. Now, this is the same as that of
the definition of strain.
So similarly, we will have εy = v / y and εz = w / z. So, these are called normal strains or
linear string okay. So, this is u / x, u is the infinitesimal strain, this displacement
happening in the x direction and dimension is given by u. So εx, εy, εz. So, these are called
normal strains.
Now, next we will move on to the shear components of the strain tensor. Now, again I refer
to x and z axis. Initially there was a body and due to some sort of deformation, due to some
response what happens is there is a kind of deformation that is happening for the point C’
74
which is given u. Similarly, there is a deformation happening in terms of w here and A and
C points gets moderately shifted to A’ and C’.
So, there is a kind of shear happening on this body. So, here u is the kind of displacement
that is happening in the x direction and w is the displacement that is happening in the z
direction, all those things remain same. Now, here it is all about deformation and hence the
rotational aspect of this becomes relevant here. Now, this is in z, this is x. So, we will first
discuss about pure shear strain or it is called tensorial shear strain.
So, what are those? Now, the pure shear strain εxz that is in xz plane, u / z. So, what is εxz?
εxz is u / z. So, this is nothing but this particular angle. So, u / z is given as εxz. So, what
will be then εzx? Which is w / x. So, these are known as pure shear strain or tensorial shear
strain. Now, there is another definition of shear strain, which is known as engineering shear
strain.
Where engineering shear strain γxz = εxz + εzx. So, this is the representation of sum total of the
deformation that happens to the body. For example, if I rotate this about this point, one can
get something like this, so this becomes γxz. So, it is a sum total rotation that is happening is
represented by engineering shear strain and that is the εxz + εzx.
Now, we can take εxz = εzx. In that case, γxz = 2εxz. So, one can write the tensorial shear strain
is half of engineering shear strain. So, in the tensorial form, it is represented either in terms of
ε or in terms of γ whatever be. Hence, we can write now εxz = εzx = 1/2 γxz. What is γxz? γxz =
εxz + εzx
Now, what is εxz? That is equal to u / z which is written here and the other one is w / x
which is written here. So, that is equal to half, half of γxz, it is very clear from this expression.
Similarly, we have
75
So, what we have done? We have discussed about the nominal strain as well as shear strain
components. So, there are 6 strains, which are functions of 3 displacements. Which are the 6
six strains? Three normal and 3 shear strain components and they are function of 3
displacements and hence they are not independent.
For compatible displacement to exist, now what is the whole idea? When some force act on a
body whether it is under stress equilibrium, whether it is compatible, these are some of the
questions we need to answer. So for compatible displacement to exist, all components of
strain and its derivative should exist which means it should be bounded. It should be
continuous to at least second order.
Now, these are certain mathematical requirements to understand whether a given system is
compatible or not. The displacement field must satisfy any specified displacement or
restraints imposed at the boundary that is satisfying the boundary condition. So, all
components of strain and its derivatives should exist, again the mathematical requirement for
continuity and it should be continuous to at least second order and displacement field it
should satisfy some specified boundary condition.
For elastic continuity following conditions should satisfy, that is displacement continuity, the
way in which we define u. Slope gradient continuity, the slope gradient continuity which is
2
given as u / x and the curvature continuity which is the higher order derivatives u / x2
should exist. Similarly, the other partial derivatives also should exist.
76
Now having done that, it is quite easy now for you to visualize. Since you have done that
strain tensor it is very easy to visualize strain stress as well, we do not have to discuss more.
So, this is the representation of strain tensor εx, εy, εz, these are linear strains or nominal strain
components and other components εxy, εxz these are shear strain components of strain tensor.
And to expand this, you can see that it is expressed in terms of pure shear strain or tensorial
shear strain.
And hence, we have half of engineering shear strain, which is expressed here. I will not go
into the details, it is self-explanatory, just like we had stress invariants it is true for strain also
it has strain invariants, but again I will not go into the details of it and a very prominent strain
invariant is the volumetric strain εv, which is the sum of the linear strains or εx + εy + εz and in
principal strain is ε1 + ε2+ ε3.
Again, deviatoric strain as we have seen previously there are different ways by which
deviatoric strain or stresses are discussed or it is defined. The same way deviatoric strain is
also defined and one such definition is
This is similar to whatever deviatoric stress that we have discussed before. Now, deviatoric
means we understand that it is going to create some sort of sharing component and
deformation the associated deformation. So, again octahedral stress, we have defined here
octahedral strain
77
So to summarize the strain acting in the soil, similar to stresses there are 3 normal strains and
6 shear strains that form a strain tensor. It can be expressed as linear strain or it has a tensile
strain engineering shear strain. So, one is normal component and the other one is shear
component of strain which is tensorial shear strain or it is known as engineering shear strain.
The rule of transformation of coordinate axes we have discussed for stresses.
Strain invariants, principal strains, volumetric strain, deviatoric strain; all are valid for strain
tensor as well. So, this is all about strain acting in the soil with relevant to this particular
course and there are a lot more which you will study when you do a course in continuum
mechanics or in geomechanics okay. So, this is all for now in this particular course related to
strain in soils.
Now, why we have discussed about stress? We have discussed about stress; we have
discussed about strain. Now, where are we leading to? We are leading to what is known as
cause and effect relationship, which we will see in the next lecture. Thank you.
78
Lecture – 11
Cause Effect Relationship
Welcome back. So this lecture is about cause-effect relationship. In the last lecture we
discussed about strain and the lecture previously we discussed about stress. Now in this
lecture we will be combining these two so that is known as cause-effect relationship okay.
So solution to geomechanical problems it should satisfy the first one is stress equilibrium
which is considered as a cause. Now as I discussed before stress is not actually the cause, the
cause is the external loading, but the internal mechanism is understood as stress. So do not
get confused with that, stress is a response, so how it can become a cause, but this is due to
the external load which is acting. The manifestation of external load is conceived as stress.
So stress is the cause that is how it is taken and the effect would be strain. So we need to
satisfy stress equilibrium condition which is conceived as the cause. Then we also need to
understand the strain displacement relationship which is defined as the effect and to ensure
compatibility. Then it is cause-effect relationship and in most of the cases for simplicity it is
assumed linear for elastic behavior.
79
Now why I have specified it here is to make you understand cause-effect relationship really
talks about the material behavior. Now cause-effect relationship, why do you need that such a
relationship that also will become clear subsequently? We need to study about stress, we need
to study about the strain, why you need to connect these two? So that it will be clear in some
time. So it is assumed for linear for elastic behavior.
Now it is all about our understanding about a given material. We know that soil and rocks
they behave nonlinearly, but there are certain situations and conditions where soil can also
exhibit linear behavior. So if you assume the cause-effect relationship essentially depends
upon what kind of response the material give for a given set of loading and depending upon
the constitutive behavior of that particular material.
Now step c that is what it means, the step c involves material constitutive behavior, whether a
material should respond linearly is mostly dictated by the material itself. A very good
example would be of that of over consolidated soil. It is the same soil but it has undergone a
lot of stress in the past. So the present stress what it is subjected to, is in fact very small for
that material and in those conditions it can exhibit a sort of linear behavior.
So what type of response a material should exhibit will depend upon the constitutive nature
of that particular material and then we need to also satisfy the boundary condition both in
terms of force and displacement. So these are some of the minimal requirements to solve
problem in geomechanics. So in that we have discussed about stress, we have discussed about
strain, displacement relationship. So in this lecture we will specifically understand the cause-
effect relationship.
80
So to summarize this is about the stress equilibrium equation which we have already
discussed before, just for completeness I am again repeating it. Again the strain-displacement
relationship which is the effect linear strains εx, εy, and εz and shear strain which is the
engineering shear strain is given in this manner. I will not spend time here, we have already
discussed this.
Now considering stress equilibrium and strain-displacement relationship, there are 6 stresses,
independent stresses, 6 strains and 3 displacements, altogether there are 15 unknown
quantities. Now; 3 equations of stress equilibrium, 6 strain displacement equations amounts
to 9 equations, again number of equations are way less than the unknown quantities. So for
solving the problem, we need 6 more equations and this comes from the constitutive behavior
or constitutive relationship, why?
81
We need to connect stress and strain or cause and effect because that process will give you
additional 6 equations with which one can solve the given geomechanical problem. The
constitutive relationship is extremely important for any material and we need to spend a lot of
time learning if you want to master geomechanics, this constitutive relationship is extremely
important. So constitutive relationship represents how a material respond to external stresses.
For a given problem, rather a given linear problem this is the cause, this is the effect and this
is the constitutive behavior or the constitutive relationship is ensured through this particular
matrix. For here it is given as 6 by 6, we will come to this a bit later and for materials like
soils which are highly nonlinear it is always ideal to talk in terms of incremental stress and
incremental strain.
So this is what is denoted here Δσ and Δε, the other thing remains the same, the constitutive
matrix that remains the same.
So cause-effect relationship: Now it is always easy to start with most simple part of the
constitutive relationship or cause-effect relationship that is why we want to discuss for
perfectly linear elastic material with the assumption of homogeneous and isotropic. I hope all
these terms are known to you. Homogeneous means spatially it is all same and isotropic it is
same in all directions and it is a linear elastic material.
82
This is a very fundamental relationship between stress and strain where εx which is the linear
strain in x direction you can write in terms of stresses as
where σx is the normal stress in x direction, E is the modulus of elasticity, μ is Poisson's ratio.
Similarly you have that is what is written here Young's modulus, Poisson's ratio. There is one
more term called G which is the shear modulus.
So similarly you have εy and εz. In addition, one can always write the shear strain γxy = τxy /G.
Accordingly γyz = τyz /G and γzx = τzx /G. So these are the set of constitutive relationships if
you consider a body to be linear elastic homogeneous and isotropic where G is the shear
modulus and it takes the relationship E / 2(1+ μ).
One important thing we have to note here is in all these expression E and μ they are the only
two parameters which need to be known to define the constitutive behavior. G is a dependent
parameter and it depends on the value of E and μ. So we have two independent parameters
we should know. Now when you take into account the cause-effect relationship, we are left
with now 15 unknowns and 15 equations and the problem can be solved.
So cause-effect relationship it is written in matrix form where ε is the strain, E is the

constitutive matrix and σ is the stress matrix, E represents the material characteristics and
here again we will be dealing with homogeneous isotropic linear elastic behavior. This
reduces E matrix to comprise of only two unknown parameters which related to material
behavior which is E and μ and which can be determined using suitable laboratory tests.
83
So that is how it looks like εx, εy, εz, γxy, γyz, γzx. The corresponding stresses and 6 by 6 matrix
in of material behavior and for linear elastic material one can always obtain and this
formulation of matrix comes from the previous slide. We have the relationship, so that is
rearranged in matrix form that is all. So this is a simple relationship wherein we talk about the
linear elastic behavior and that is why this particular model is so popular.
So if one can define the behavior of the material using linear elastic, then it considerably
simplify the problem and it becomes less computationally intensive.
So volumetric strain under stresses σx, σy, σz. This we have already seen and if you expand
you can write as
So expression for bulk modulus now there are two more parameters, sometimes it is
expressed in terms of K and G, in some cases it is represented in terms of E and μ. Whatever
be all these terminologies parameters should be familiar.
So expression for bulk modulus corresponding to hydrostatic stress condition is just to

understand the expression for K which we all already know, but just to have a discussion in
terms of let us say whatever we have learnt till now. Now what is meant by hydrostatic stress
condition? Stresses are equal in all direction. That means σx = σy = σz. So bulk modulus how
do you define?
84
It is volumetric stress upon volumetric strain and this is the condition for hydrostatic
condition, this is the expression σx = σy = σz = σv. If you substitute this back into this equation
which is the expression for volumetric strain you can get the expression
So K is nothing but E / 3(1- 2μ) and you can see that both K and G are dependent on E and μ,
so it is not independent.
So now we have E = 3K(1 – 2μ), E and K are positive because they are material constants
which means to say (1 – 2μ) > 0 ,because E has to be positive. For fulfilling that condition,
we can write that Poisson's ratio is always less than half. Now in general Poisson's ratio
varies between 0 and 0.5. Now there are certain materials which assume the value of
Poisson's ratio close to 0.
For example, cork is a classical example and in general for most of the problems in
continuum mechanics related to soils and rocks the variation we will consider around 0.1 to
0.5. We do not talk about μ close to 0 and you will be surprised to see there are certain
materials where μ is even negative. These are called auxetic materials, you can refer to
literature and basically these are engineered materials mostly and which exhibit negative
Poisson's ratio, but that's not relevant for this particular course.
85
Now for μ = 0.5, in the previous relationship if you substitute μ = 0.5 that is Poisson's ratio
equal to 0.5, then the volumetric constant or bulk modulus it tends to infinity which implies
that the volumetric strain is equal to 0. There is no volumetric strain, the material is
incompressible, body undergoes no volume change it is basically incompressible. This
definition has got some meaning in soil mechanics as well.
So undrained condition, we will be discussing about drained, undrained condition in detail,

but since this is an undergraduate portion all of you will be knowing what is meant by an
undrained condition, loading under undrained condition. So what happens during undrained
condition? You are not allowing the water to move out of the soil during loading. So in that
case is a typical case of no volume change.
When there is no expulsion of water, then you cannot have volume change, what will
happen? Pour water pressure will increase. So there is no volume change, the whole of the
soil mass behaves incompressible, so what does it mean? It means that for undrained
condition one can take the value of Poisson's ratio = 0.5. So the whole discussion is basically
for that.
So in geomechanics we actually refer to the μ value from 0.1, 0.2 to 0.5. For purely undrained
case for soils undrained loading, one can always presume μ = 0.5.
To summarize whatever, we have just told about cause-effect relationship, we have solution
to geomechanical problem involves stress equilibrium equations, strain displacement and
86
cause-effect relationship, so 15 unknowns, 15 equations. Considering above there are 15
unknowns and 15 equations for solution. Cause-effect relationship for linear-elastic,
homogeneous, isotropic material has been discussed in this particular lecture.
There are only two material constants for the above case and that is the simplicity of this
model where we discuss in terms of E and μ and Poisson's ratio μ is = 0.5 for undrained
incompressible condition. So, that is all for this lecture on cause-effect relationship. We will
briefly discuss about some classical constitutive relationship which you normally see in
geomechanics again, it will be a very elementary discussion which we will see in the next
lecture. Thank you for now.
87
Lecture – 12
Important Constitutive Relationship
Welcome back in the last lecture, we were discussing about cause effect relationship. Today's
discussion is a bit of extension of that particular lecture, where we will discuss about some
important constitutive relationship. Again it is not an exhaustive session here, we will
introduce certain important relationship between stress and strain and which is relevant in geo
mechanics. So, cause effect relationship, it describes how stress and strain is related and we
have seen that the most important aspect is constitutive metrics.
Now, depending upon what sort of model you adopt for modelling a given situation, the
constitutive matrix keeps changing. So, we will see some of the important constitutive
relationship in today's lecture.
Again, we will stick to the very same issue that is solution to geomechanical problems, it
should satisfy stress equilibrium, strain displacement relationship which ensures
compatibility the cause-effect relationship which is assumed to be linear for elastic
behaviour, this one and the step c involves the constitutive behaviour. So, today's discussion
on constitutive relationship will focus around this particular point.
88
That is the constitutive behaviour and we will skip this we have already discussed set of
discussion, why it is essential for one to understand the problems or the solutions in
geomechanics involves these 3 steps. The first 3 steps essentially and plus this in some cases,
you will see that maybe stress equilibrium is satisfied but the others may not. There are
several problems in geomechanics or maybe specific to soil mechanics as well, where we do
not achieve the other 2 but stress equilibrium is obtained.
So, how a particular method got evolved, it depends on that. Now constitutive relationship in
geomechanics, it is a cause effect relationship. That is what we in the beginning I told is an
extension of last lecture session. It is a relationship between stress and strain or you can see it
is a relationship between force acting on a material and the resulting displacement, the real
stress strain relationship of soils or rocks, it may be complex in nature.
In most of the time, we have seen and we everybody who is an engineer or a non-engineer,
they will be able to understand this very fact that soil is pretty complex and very hard to
understand and whatever we say, still, there will be some gap left in our understanding, there
is so much of uncertainties are there associated with soil behaviour. So, a real stress strain
behaviour is beyond question, what we can do is to understand or to develop models which
are close to or which closely simulate the real soil behaviour and this is simplified.
So, we understand the fact that it is complex. So, through some idealization, these particular
relationships are simplified by idealization and are required for the development of
mathematical models.
89
So, first let us begin with a very simple model which is linear elastic behaviour. One may
always tell why we should discuss linear elastic model at all in soil mechanics, because most
of the cases we understand soil behaves nonlinearly but there are certain aspects just because
of its simplicity and simple mathematical formulations and minimal requirements. This
model is very popular.
So, linear elastic model, we always start with that and there are very specific problems which
consider it to be elastic in nature. Now, how far it is true, that is a question mark but because
of its simplicity and to the fact that the results are not really off, under certain circumstances,
we can even apply linear elastic model for simulating the behaviour of soils but not always it
is far from reality.
As you can see, the y axis is the stress and the x axis is string and this is a simple linear
elastic behaviour, where it is represented by a straight line and we know that it is non
hysteretic means it traces back the same path. So, linear elastic model, the load results in
stresses which are well within the yield limit. Now, this is the point. Now, if the loading
acting on the soil is fairly less, as compared to its strength, we can still see that it is within the
elastic limit.
And this is the whole basis of working stress method and the logic of applying factor of
safety, we apply a high factor of safety and ensure that the stresses which are acting on that
particular body or soil is well within or maybe it is at par or slightly below the elastic limit or
to be very specific in geomechanics the yield limit. Now yield limit is the point beyond which
90
the plastic strain sets in or the plastic behaviour starts happening, let us say that we have
marked up to here at up to this particular point.
Now, if the same material is further loaded, it may exhibit a nonlinear behaviour. So, this
point can be treated as yield limit. So, we will come to that a bit later. So, linear elastic
consideration ensures or we make a presumption that the stresses are well within the elastic
limit or the yield limit. Yes, I have already told this it is popular due to its simplicity and the
requirement of less number of parameters.
Obviously, this we have seen in the last lecture, we have seen that for linear elastic cause
effect relationship, you need to know only 2 parameters and what are they? Modulus of
elasticity E and the poisons ratio μ. So, with this one can readily model the behaviour of soils
or maybe geo materials. Now, there is an inherent assumption involved in most of these
models which is directions of principle incremental stress and incremental strain coincide.
Means, the direction in which the incremental stress is it is the same direction for the
principle or incremental string. So that is what it means both coincide. Now, when it comes to
plastic formulations, it may or may not coincide. So, there are various other models but in
linear elastic for that matter, various other elastic models will have this inherent assumption
that the incremental stress and incremental strain are in the same direction incremental why
incremental. We are talking about a very small increment of stresses that we will come to
know a bit later why incremental is important and this model essentially corresponds to
homogeneous isotropic low stress that is what I mean, if we want to ensure that it is working
stress, then it has to be under a low stress condition and material parameters independent of
stress level. Now, if you consider this particular plot, you can see that E the value of E
remains constant up to this particular limit yield whatever be the strain level.
We can see that this are different levels of stresses and this various levels of string or
whatever be the stress level and the strain level, if your body is considered to be linear elastic
in nature, the parameters or the material parameters which is E and μ is considered to be
constant or rather it is constant and it is independent of the stress level. So, linear elastic on
only this particular requirement is there and it essentially confirms to low stress.
91
So, if the material behaviour is linear, then the modulus of elasticity and Poisson’s ratio are
constants which I have already discussed. So, it does not simulate the important aspects of
soil behaviour. So, in the absence of other complex models, one can always use linear elastic
model but with a proper understanding that it is not going to simulate the real soil behaviour
but, it is not fully true the statement.
There are certain situations where soil exhibit some sort of elasticity and one good example is
an over consolidated state where the soil is fairly stiff and when it is stiff, it can exhibit
elasticity. Again that elasticity is also subject to certain conditions beyond certain point it
again subject to yield. So, it is condition specific, it is not like other materials for soils, what
type of behaviour it would exhibit, it is also condition specific under certain conditions it can
exhibit elastic behaviour.
So, this model would be more relevant for structural elements like retaining walls because
retaining walls are very widely used and it is an important aspect of geotechnical engineering.
So, for modelling those linear logistic models may be important. So, now, we will go to some
other extension of the linear elastic model.
Now, we will come to bi-linear model in fact, bi-linear models are or it represents non
linearity but, you are doing some sort of adjustment or you are showing that non linearity
using 2 different lines. So that is why it is called bi-linear model material behaves linear
elastic till it reaches yield stress, we will be discussing about yielding yield stress in detail in
92
module 4 but here we just need to understand it is the point beyond which the plastic
behaviour starts.
So, material behaves linear elastic till it reaches yield stress and then remains constant with
strain. So, what it means is that the stress strain behaviour, it is of certain manner till it
reaches yield limit or yield stress and beyond which it remains constant. Now, this constant
means, depending upon the formulations, there will be some sort of changes as constant, we
will come to that in the discussion. Material parameters remains constant till it reaches failure
condition.
So, here we will discuss 2 bi-linear model. The first one is elastic, perfectly plastic. So, you
can see here that we are not talking only about elasticity is a pure elastic model. It is elastic,
perfectly plastic, it includes plasticity behaviour as well. So, here it is not a pure elastic
model, it is rather elastic and plastic response, both are taken care. But since it is an
approximation using linear lines, it has been discussed now, you can see that there are well
defined 1 and 2, 2 well defined behaviour one is in this manner. One is in this manner, there
are 2 straight lines. So that is why it is called bi-linear model and elastic perfectly plastic.
You can see that from this point to this point. You can see it is a linear trend and the material
constants E and μ remains constant. So, this part takes care of the elastic behaviour. Now,
after reaching the yield point then nothing changes. It has gone to its plastic state or just
directly from the elastic limit, it goes to its plastic state or the failure state.
So, here you can see that stress remains constant, whereas, the strain increases to a large
deformation but at constant stress condition. So, here it remains constant, whereas,
immediately after the yield point, it remains constant, the stress remains constant. So, one
more requirement in this particular model is one should be knowing what is the yield point.
So, apart from E and μ one should be knowing what is the yield point the Mohr Coulomb
failure envelope using c and φ is a good example of this for setting the yield point.
So, this we know that this is the point at which the soil will fail. So, it is assumed up to that
particular point it behaves elastically. This is also far from reality. In real condition, soil is
not going to reach to its plastic state just at its yield point or at its failure state at its yield
point, there are certain other type of behaviour as well in the real soil behaviour. Now, we
93
have discussed about elastic perfectly plastic a similar model or maybe a subset of this, is
rigid perfectly plastic.
Where the material is considered to be having no deformation till it reaches the yield point
you can see here. There is no elastic portion of the curve here in the earlier case you have
elastic portion here there are, there is no elastic portion directly it goes to its failure state at its
yield point. So, in this case, only one parameter need to be known and that is the yield stress
or the yield point or the failure state whatever, but the response both are same. So, here it
jumps to its yield condition directly.
Now, these are different models which are used. So, once it reaches failure, elastic parameters
are set to a value close to 0. Now, I would be very cautious to use this term it is close to 0,
because, you will see that in certain finite element modelling, if it is set equal to 0 then there
will be a problem. So, it is given a very small value close to 0. So that is why I am used this
particular term and we have written here not necessarily equal to 0, even though we know
according to the model, it is 0 but it is set close to 0.
If this stress rate is unloaded, exactly at the yield point, it comes back to its pre-failure value
which means I am not telling any point here and telling at this particular point, if the
unloading is done then it gets back to its original. So that is the issue related with this
particular model. Once at yield it undergoes some sort of plastic deformation but here even it
has reached at its yield point and if you unload, it is bound to come back to its pre failure
state. Elastic parameters defines the pre-failure behaviour.
So, pre-failure behaviour is defined by elastic parameters. Further, it needs parameters to

define the failure surface that is what I told you need to know the yield point as well. This
behaviour is elastic and perfectly plastic and these are used in limit analysis problems that
you use in soil mechanics, where in the case of limit analysis, it also satisfies the stress strain
relationship but it needs to know what is the limit value according to certain theorem.
So, we will not go into the details of limit analysis but this becomes the foundation block for
limit analysis used in geotechnical engineering, if the elastic deformation is negligible, then,
it is assumed to be rigid, perfectly plastic, only under the condition that we neglect the elastic
94
deformation it is, so less that it is neglected and straight away it is going to its failure. So,
limit analysis of slope stability is a classical example of this.
Now, we will quickly discuss about certain other model which are in line with this bi-linear
modelling which is linear, elastic, linear hardening, now I am, I am introducing another term
which is known as hardening, hardening behaviour, it is a typical behaviour exhibited by
plastic response of material after it reaches its yield limit, where the stress increases with
strain as you can see here.
So, this particular point is the yield limit, where it shows the transition from elastic state to
the plastic state, obviously, the stiffness portion has reduced now, this portion is called the
hardening state and this is the linear elastic portion but it also tried to capture the linear
hardening. So, this in fact, it is not linear but here it is approximated as linear hardening. So,
this is linear elastic linear hardening.
Now, another aspect of it is linear elastic, nonlinear hardening. Now, this particular portion is
represented by nonlinear behaviour. So, both captures the hardening portion of the soil
behaviour which means, I will again add here hardening it refers to a sort of volume change
that is happening during sharing, is the kind of response in linear elastic or maybe the bi-
linear model that we discussed, this portion of volume change during sharing is not
incorporated.
95
So, beyond yield point, there will be a volume change now, this volume change results in
such behaviour which is hardening or it the another aspect is softening. So, we will discuss
those in detail later but here we just need to understand this particular model takes into
account the hardening behaviour by certain line and in this case, it is a nonlinear behaviour.
We have discussed about linear elastic. Now, the next set which is a bit of refinement over
the linear elastic model is nonlinear elastic behaviour.
Where the material response is considered to be nonlinear but elastic in nature, how do you
ensure this, this is a nonlinear response is a curve response but, if you unload at any point of
time, this particular curve trace back. So that is the one very simple definition of elastic
behaviour, it does not show hysteretic behaviour, the manner in which you load it traces back
So that is what is shown here, you. This is the loading and this is the unloading.
So, non-linearity is in the curvature, it is there but it exhibits the elastic behaviour, it does not
result in any sort of permanent deformation and it traces back this is more realistic than the
linear elastic behaviour for soils. Material parameters are dependent on stress and or strain
level. Now, this we need to understand carefully and that is why the incremental strain
becomes important like at every point, if you consider along this curve and we draw a tangent
at these points.
For example, if I draw a tangent here, if I draw a tangent here, if I draw a tangent here, we
can see that this stress by strain or rather the elastic constant that keeps changing at every
point on the curve. So that is why we call it as tangential increment. The increment is
96
represented by tangent at a particular point on the curve and that gives you what is the elastic
constant or material parameters.
So that is what it is written here. So, material parameters are dependent on stress and or strain
levels. So, if you consider this is a particular stress level or this is a particular strain level
similarly, you have this as the stress level and this as the strain level. So, at every stress strain
level, the material constants or the material parameters keeps changing. So, isotropic
assumption is valid to ensure that it has only 2 parameters.
If it is an isotropic then again the number of parameters keeps changing because in different
direction parameters will be different, any of the two material properties are chosen. Now, we
know that this is an elastic analysis and hence the very straightforward choice is E and μ. But,
we can also represent the model in terms of bulk modulus and shear modulus, why
specifically bulk modulus and shear modulus.
Because one represents the volume change behaviour and the other represents the shearing
behaviour because it is associated with shear in soil mechanics K and G are preferred as the
soil behaviour is represented in terms of mean stress and deviatoric stress. So, when it comes
to model that uses invariance it is found that mostly K and G are preferred, why because, one
the same reason what I just discussed, it represents the volume change aspect and the
shearing aspect.
We have seen in the earlier lectures that we have already told these are decoupled which
means to say we study the from the given stress tensor, we decouple the mean stress and the
deviatoric stress and we study the volume change behaviour keeping the shape constant and
we study the shearing behaviour keeping the volume constant. So, these are taken one at a
time. So, volume change and shearing in a decoupled manner and this is purely from
simplicity point of view.
The actual soil behaviour, it does not follow a decoupled behaviour. So that is all about
nonlinear elastic behaviour, another refinement of this is nonlinear elastic, nonlinear
hardening. So, this portion gets added up here, it is not it is up to yielding. So, beyond a
yielding, the hardening response is also treated as nonlinear. Another model is K-G model,
where K stands for bulk modulus and G stands for shear modulus.
97
So, the tangential bulk modulus that is incremental at a particular point and shear modulus are
defined in terms of stress invariance, how it is defined in terms of P’and J2. What is J2? J2 is
the second invariant of deviatoric stress tensor. Now, probably you will understand why we
have discussed these invariants before. So, it is just not discussed, why because these
invariants are very much used in your mechanics modelling or to express certain models in
these invariants are used. So, these invariants are very important.
So, here in the K-G model, the both the bulk modulus and the shear modulus they are
expressed in terms of stress invariance that is the mean stress P’ and root J2 where cannot
alpha K0, G0, αG and βG these are all parameters of the model. So, we need to determine these
parameters in order to apply K-G model, the stress strain curve for this model is given that is
P’ versus εv.
Let us mean stress upon volumetric strain, it is always associated with volumetric strain and
the deviatoric stress that is is associated with deviatoric strain. So, E d here represents the
deviatoric strain. So, what is deviatoric strain? The expression is given here again it depends
upon the various definitions of deviatoric stress or deviatoric strain there is no specific
guideline as such different models assumed different forms of deviatoric stress strain
expression.
For this in this particular model, the deviatoric strain is given by
98
So, stress strain curve one important aspect which has we need to note here is in P’ versus v,
the slope keeps increasing in this direction which means to say the bulk modulus keeps
increasing.
Whereas, in the case of was deviatoric strain you can see that the bulk modulus keeps
reducing as the soil gets more and more stressed or sheared. So, you can see that the shear
modulus reduces whereas, the bulk modulus increases. So, this is typical stress strain curve
representation which we use in K-G model.
So, next is hyperbolic model; hyperbolic model is a relationship between accumulated stress
to accumulated strain. Now, till now, we have been talking mostly about the incremental
stress and incremental strain. Here it is the cumulative effect of stress and string it was
originally used it was a hyperbolic model was developed to present undrained triaxial test
based on 2 parameters in which Poisson’s ratio is equal to 0.5.
I think this point you remember in the previous lectures, we have already discussed this
particular point Poisson’s ratio is set to 0.5 for undrained condition why the bulk modulus
becomes infinity and then there is no volume change. So, original model is represented by
this hyperbolic equation:
σ1 σ3 major and minor principle stresses, ε is the axial strain and a, b are the material
constants. So, it looks like this and the initial tangent model is shown here Ei and the
asymptote is shown here (σ1 - σ3)f. Asymptote means, the final value at which the failure
99
happens is represented by this curve becomes asymptotic and that particular value is shown
by (σ1 - σ3)f and the initial tangent modulus Ei is shown here.
Now, it is shown that the initial tangent modulus Ei = 1 / a and the failure value that is (σ1 -
σ3)f that is the asymptote value is equal to 1 / b. So, both these material constants a and b has
some meaning in the hyperbolic model that is 1 / a is Ei and 1 / b is (σ1 - σ3)f. So, based on
our stress strain data, if we can determine a and b then the constants Ei and the final value is
known.
So, for that, if we rearrange the hyperbolic model in this particular form that is:
Now, this is a linear response . So, plotting laboratory data to this re-arranged form a
and b can be determined how if you plot this portion as a function of ε this then we
are likely to get a straight line and the intercept is a and the slope is b. So, once you get a and
b, Ei and (σ1 - σ3) at failure or the asymptotic value can be determined.
100
So, another model is small strain stiffness model which is based on the understanding this is a
relatively newer model as compared to the earlier discussed models which were very much
there in the literature. Now, this particular models small strain stiffness model is based on the
understanding that the soil undergoes significant changes in stiffness when initially subjected
to a small change in string.
That means, if it is subjected to for the first time or initially a small changes in strain, it can
bring about a significant change in stiffness. Now, none of the above models what we have
discussed can capture this particular behaviour. So, this particular model has been developed
the existing models they are not capable for capturing the stiffness change at small strain. So,
small strain stiffness model it was developed to simulate the soil behaviour in small strain
range not applicable for large strain where failure occurs.
So, frankly we are not discussing about the failure in this model. We want to capture this
small strain behaviour. The variation of bulk can share moduli are expressed as secant
modulus. Now, what is secant modulus? It is expressed as Gsec and Ksec. So, what is the
secant modulus? So, if this is σ versus ε and it represents a nonlinear curve, so this is the
initial tangent modulus and this is tangent modulus.
So, if I want to find out the modulus here, the secant modulus is this, so this slope is the
secant modulus. So, here the variation of bulk and shear moduli are expressed in terms of
secant modulus and the expression is given as
101
So, there are lots of parameters along with this small strain stiffness model. So, practically it
becomes difficult in certain cases where the numbers of parameters are too large. So, you
need to determine all of those. So that becomes sometimes a deterrent for choosing certain
models and why models like linear elastic models are so popular. So, all these are material
constants whatever has been discussed here, they are material constants and hence, those
issues make these model less popular but for taking care of specific behaviour one need to go
for these models.
So, where Ed in this expression is deviatoric strain we have already explained εv is the
volumetric strain and these are strain invariance.
Before ending this lecture, we need to actually discuss what is the real behaviour and how it
can be captured for problems in geomechanics and that is elasto-plastic behaviour, a more
realistic model, different version of elasto-plastic behaviour we have discussed in bi-linear
section where we discussed about elastic perfectly plastic and rigid perfectly plastic, this is
yet another and more refined and considerable improvement of those where we discussed
about the elastic as you can see here.
This is the elastic portion here the yielding happens, the plastic behaviour sets in so elastic
yielding and plastic. Now, if we try to unload it from this particular point, it is not going to
102
trace back. Now that is the difference between nonlinear elastic and elasto-plastic behaviour.
In this case it is not tracing back rather it takes another route and this is more or less parallel
to this linear portion.
So, it comes down and if it is again loaded, it traces more or less the same parallel as that of
elastic behaviour, why possibly, it is because it has been loaded up to this. Now, whatever
loading you are making here, it is always less than the load which it has been subjected. So
that is why here this is the elastic response, this is the plastic response, here it is unloading
and reloading and then beyond a certain point again the plastic response sets in. So, here
another important aspect to be looked at as this is the string.
Now, from here we have unloaded. So, whatever string it has regained that is elastic portion
of the strain and whatever it has undergone the permanent setting or permanent deformation
that portion is given by the plastic strain. So, the total strain in elasto-plastic response is the
summation of elastic strain and plastic strain. So that is what is shown here. So, knowledge of
stress strain behaviour beyond yielding is important to understand failure this we have
already seen, unless we know where it is going to fail, it is difficult.
So, now, here it there is a sort of behaviour which the soil exhibit beyond yielding. So that is
captured in elasto-plastic models stress strain relationship, it is not unique beyond yielding
which means to save once it has yielded, then the stress strain response is not unique the way
we have it for elastic response and strain at a point not only depends on the stress level but it
also depends on the loading history.
That is the point P, what it means to say is, let us say it has taken this is the elastic portion,
then there is plastic deformation that is happening at this point it is unloaded. Now, let us say
that it reaches this particular point P. So, we have reached this particular point P. So, what it
means strain at a point that is strain at point P, this one it not only depends on the stress level.
Now, here at this particular point, you can say that, this is the stress level.
Now, from here it is loaded, it can take this particular route to P or it can go it can yield and it
can come back here and here there is a point P. Now, the strain at a point here it has crossed
the strain and then it is unloaded. So, strain at a point, it not only depends on the stress level
103
but also the loading history at the point P. So, it is not only the stress level, if I know a
particular stress level, this is the strain it is not like that.
That strain would be a result of the stress history as well or the loading history as well and in
soils this is extremely important, what is the loading history and that is how the concept of
normally consolidated and over consolidated or dense or loose state comes into picture. So,
what is the stress history or loading history also play a major role in defining what is the
strain which the body has undergone in this case, it is purely elastic strain, whereas, in this
case, it has undergone some plastic deformation and certain regaining has happened.
So, totally these are different path which it has followed. A realistic model for soil behaviour,
where both elastic and plastic behaviour is incorporated the models can account for yielding
hardening and softening we will come to hardening and softening later. So, here it is yielding
and it is hardening now, this is a typical behaviour of ductile material where with stress the
strain is increasing.
So that is a typical behaviour of hardening and in the case of softening after yielding the
stress reduces with strain. Now, why this behaviour we will see it later in the subsequent
modules. So, this is a typical behaviour of brittle materials. So, yielding with hardening
yielding with softening. So, these 2 aspects can be captured by different elasto-plastic
models.
104
So, some similarities between hardening and oedometer results. So, if you re-plot the stress
strain plot like this, so what I have changed, I have interchange the x and y axis now, the
strain is on y axis and stress is on x axis, you can see the same plot as before. So, A to B and
then yielding and then the plastic behaviour happens from here it is unloaded. So, this is
unloading reloading path CD which is more or less parallel to AB.
Now, if you compare this with a given oedometer results, you can find that there is a sort of
similarity here it is E versus log σ’v. This is a very popular representation of odometer results
where you talk about the virgin consolidation line and swelling-reloading line you can see
that this particular response of this is more or less similar to this behaviour this particular
behaviour. Whereas swelling-reloading line it is more or less mimicking this particular
behaviour.
So, swelling-reloading line is similar to the unload reload plot of this particular stress strain
behaviour CDC. So, you can consider this as the yielding point which is somewhat similar to
here or maybe similar to here. Swelling-reloading line is similar to the unload-reload plot that
is CDC of strain stress plot because it is plotted on y axis here, this particular line and this
particular line it closely resembles and the way in which it responds, it is also similar.
CDC exhibit elastic behaviour, so does the swelling-reloading line and it has something to do
with the stress history. Virgin consolidation line induces permanent plastic strain, so beyond
BC, so this is the yield point. So, beyond yield point it induces plastic deformation and same
is the case with the virgin consolidation line where there is permanent setting which is
happening and this is similar to the yielded strain hardening behaviour of BCE. So, this is
yield point and this is the hardening behaviour BCE. So, this and the virgin consolidation line
behaviour these are same.
105
So, to summarize today's lecture, the simplistic linear isotropic elastic models require only 2
material parameters and that is why it is simple. In nonlinear elastic models, material
parameters vary with stress strain level, whereas, in linear elastic model it is constant. The
nonlinear elastic models it cannot capture the volume change behaviour of soils during
sharing, we are talking about the hardening and softening behaviour. All the elastic models
inherently assume the same direction of incremental stress and strain.
Most of the linear elastic and nonlinear elastic models cannot capture the real soil behaviour,
we have already explained that. Elasto-plastic models they have been developed to capture
the nonlinear stress strain behaviour of soils fairly well, it can account for the strain
hardening and strain softening response during shearing and beyond yielding the behaviour of
soil is not unique.
But, having said that, we also need to keep in mind that as we proceed from linear elastic to
elasto-plastic models, the complexity of mathematical solution also increases and the
computational effort also increases. So that we have to keep in mind. So, the various aspects
related to anisotropy is not considered in this course. So, mostly we will confine all our
discussion to isotropic soil behaviour.
So that is all for this particular lecture, we have just discussed some important aspects of the
cause-effect relationship in terms of constitutive behaviour or constitutive model behaviour.
So, we have discussed about different models which relates the stress and the strain. But, let
106
me again make a point here that this is not an exhaustive discussion, we there are several
aspects to what we have discussed now.
This is only for you to get oriented towards such a constitutive behaviour. So, with this we
will end this particular lecture and till now, we have been focusing mostly on 3 dimensional
aspects. Now, for actual problem solving, we need to transform this 3D to 2D behaviour for
simplicity. So, for certain problems, one can always idealize a given 3D problem to 2D. So
that will be seen in the next lecture.
107
Lecture – 13
3D to 2D Idealization
Welcome back. So, now we have already completed stress acting at a point, strain, stress strain
relationship rather cause effect relationship and we have discussed some important constitutive
relationship which is handy for geomechanics. Now, in most of the discussions, we have focused
on 3 dimensional state or we have discussed x, y, z 3 dimensional coordinate system. Now, the
scale of the problem is in 3 dimension.
There are certain cases or there are certain problems where you can convert 3D to an equivalent
2D in modeling. This will be quite beneficial from computational point of view as well as
simplicity point of view. Hence, in this today’s lecture, we will see 3D to 2D idealization.
So, 3D to 2D idealization of geo mechanical problems: As we have seen, the geo mechanical
problems are essentially 3 dimensional. It is not 2D. This makes the solution to the problem
complex and computationally intensive, reality is 3D but when you conceive 3D, it is going to be
a bit more complex than 2D and it is computationally intensive. So, any effort to solve the
problem with reduced dimension through some idealization, it is a welcome move.
108
The idealization is possible depending on geometry of the problem, loading conditions and
symmetry. So, based on these 3 factors, one can idealize a 3D to a 2D problem, how it is done? It
may be possible to neglect some components of stress tensor. It is not stress tensor we are
neglecting, some components of stress tensor, strain and or displacement. So, essentially, the 2D
idealizations which are considered in geo mechanical problems are plane strain, plane stress and
axisymmetric condition.
For completeness, we are discussing all the 3. What is this 3D to 2D idealization based on these
3 formulations? But, in geo mechanical problems or in soil mechanics, what is more important is
plane strain and axisymmetric condition. So, these are the 2 important ones when you consider
3D to 2D idealization.
So, first let us discuss what is plane strain condition and how you can idealize a 3D to 2D
problem? Based on conceiving a geo mechanical problem as a plane strain. So, what is meant by
plane strain? You can consider geo mechanical problems in which 1 dimension of it is very long
as compared to the other 2. For example, retaining wall, dams and tunnels you can see here those
x axis, y axis and z axis a typical gravity dance structure is shown.
109
You can see that this is the cross section, compared to cross section, this direction is quite long.
So that is what it means one dimension is very long. So, z direction is pretty long as compared to
the its cross section. So, it is possible to consider strain acting in the longest direction to be
negligible as compared to other two directions since this length is quite long. So, you can always
consider the strain acting in z direction to be negligible.
In the process, a 3D problem can be idealize; to 2D by following some assumptions. So, what are
the following assumptions and what are the conditions that are inevitable for considering plane
strain condition. one of the dimension of this structure should be very long say z as compared to
x and y in this particular example. Loading is perpendicular to z axis that means, there is no
effect of loading in the z direction and when I say loading, it includes both body forces as well as
surface traction.
Now, these forces are perpendicular to z axis, because of which it does not have any effect in the
z direction. Also, loading is not a function of z dimension, body forces and surface traction are
only a function of x and y and it is independent of z. So, it is a function of only x and y, it does
not vary with z. Now, the fourth condition is ends are constrained or it is infinitely long in z
dimension, why? This condition is needed for achieving the most important condition for plane
strain.
That is strain in z direction or any 1 direction is 0 or close to 0. So, how it is made, either it is
end constraints are there, or the length is infinitely long, so that when you divided by its original
length, the strain becomes negligible. So that is how ɛz = ∂z / z which is approximately equal to 0
and the other strain components in z direction ɛxz, ɛyz is equal to 0. So, whatever is there,
whatever strain components associated with z is neglected.
110
So that is about plane strain condition, we will have some basic formulations done for plane
strain and plane stress. But in todays lecture, we will understand how and what is plane strain,
how does it exist which problems it is relevant. So, similarly, we will also know plane stress
condition, it is not that important in geomechanics but for completeness, we will see what is
plane stress condition.
For problems where one dimension is very thin. So, in the plane strain, it is very long, here it is
very thin as compared to other two. For example, thin plates, you can see here this is a thin plate
and dimension along z direction that is thickness is very less as compared to x and y. So that is
what it means, here less means it is relative. So, as compared to xy plane, the thickness is quite
less, so, it is becoming like a thin plate.
So that is what it is you can see here the thickness is very small as compared to the area of cross
section in the xy plane. So, because of this one can always approximate the stress acting in the z
direction σz as well as shear stress components τxz and τyz this can be approximated to 0. Stress
tensor components in z direction are 0, this idealization is called plane stress. So, stress in one
direction is 0 that is plane stress, strain in one direction 0 is plane strain.
So, 2D stress can be represented by in this manner. So, we are left with σx, σy and the shear stress
component acting in the xy that is τyx and τxy. So, what are the example is plate with a hole under
111
uniform load, as you can see here, masonry wall under vertical load as you can see here. So,
these are some of the examples where you can approximate the given problem to 2D plane stress.
Strain tensor for plane strain can be written in this manner. Now, we have strain in one direction
to 0. So, all these components have become 0. So, you have ɛx, ɛxy, ɛyx and ɛy. Similarly, you
have stress tensor for plane stress now, here the other components are all 0; you are left with σx,
σy σxy, σyx.
So, the third idealization is axisymmetric condition to analyze 3D problem which is symmetric
about its axis. Now axisymmetric condition is not new to geotechnical engineering is most of the
112
problems this particular aspect comes into picture knowingly or unknowingly we have already
dealt in detail, the axisymmetric conditions or axisymmetric problems, this is also nothing but
you are idealizing a given 3D problem to a 2D problem, how? It is based on axis of symmetry.
Let us see how there are several problems in geomechanics which have rotational symmetry. For
example, a very good example is triaxial samples, everyone has worked with triaxial samples,
you can see that there is a deviatoric stress acting actually and all round stress acting now,
because of this all round stress, there is a sort of axis of symmetry, with respect to geometry as
well as with respect to loading.
So, triaxial sample is a very good example of axisymmetric condition, where you change the 3D
to a 2D state. Then we have single piles actually loaded single piles, caissons, circular footing
with load acting at the center, tunnel cross section and borehole that is what it is. So, along this
direction, if you consider this is acting at the center. So, along this axis, it is axis of symmetry,
same is the case with actual loaded pile or triaxial sample which we have already told stress
around a circular tunnel.
So, this is an overly simplified representation of tunnel where this is the rock medium and this is
the tunnel, you can see that with respect to the center, if you take there is an axis of symmetry
along this longitudinal direction, if you take the longitudinal axis, you have an axis of symmetry.
So, axisymmetric problems these are mostly defined in terms of polar coordinates in terms of
radial distance r, vertical distance z and the angular plane which is given by θ, so, it is basically a
polar coordinate approach.
So, axisymmetric problems, it also satisfies actually symmetric geometry and axisymmetric
loading. So, the symmetry can be both, because you need to have symmetry with respect to
geometry that is the first condition and it should also be symmetric with respect to load, because
if I place another load somewhere here, then there will be sort of eccentricity acting, so then you
cannot consider it exactly to be an axisymmetric problem. So, both geometry and loading are
important to define the symmetry.
113
So, here is the representation, let us say this is the reference axis and this is the radial distance r
and this is the vertical distance z. So, if you consider a cylinder like this and this is the reference,
so you have θ beginning from here and then all around to 360 degrees. So, one need to have the
properties to be seen, if; you go from here to 360 degrees. So, if this is considered as a plane, if
you rotate this plane from 0 to 360 degree, nothing should change; neither the loading nor the
geometry should change.
So, then we say that the axisymmetric condition is satisfied. So, displacement, strain and stress in
z and r are independent of θ. That is what I just explained. What it means, if you consider stress
displacement or strain in r direction and z direction, then if you rotate θ, this is not going to
change rather, whatever is there in r and z direction, it is independent of θ, the value of θ = 0, θ =
90 things are seen. So that is how we say that the axisymmetric condition is met.
So, strain tensor input polar coordinates in general, you can represent it by in this particular
form,
So, it takes the form. So, basically, you have epsilon r, epsilon z, epsilon θ and the shear stress
components. Strain tensor for axisymmetric conditions specifically, what is there what will go
and what will not be there?
114
So, you can note that the shear stress component, εrθ, εzθ becomes 0 and since it is symmetric, so,
this is also 0. But another important aspect is epsilon theta is not equal to 0 in plane stress plane
strain, in fact that was also 0 in axisymmetric condition, εθ is not 0. Why? Displacement acting in
θ direction, it is 0. So, if I represent it, so, this is the reference, this is θ. Now, θ direction means
this is the θ direction and this is the tangential location.
So, if you draw a tangent here, we define theta based on the tangent. Now, what it means is that
displacement in this direction is 0, if displacement is 0, then why? Epsilon theta is non-zero,
because it has got component from some other axis for example, radial displacement induce
circumferential strain εθ, epsilon theta is the circumferential strain. Vertical component it will not
induce any sort of circumferential strain.
But radial displacement that is displacement in this direction, this can induce εθ or
circumferential strain. So that is why epsilon theta is present. So, stress tensor for axisymmetric
condition is σ is equal to σr, σrz, σzr, σz and σθ more like the strain tensor.
So, now, we have just introduced how and what are the possibilities of 3D to 2D idealization. So,
we will summarize a 3D geo mechanical problem can be transformed to an equivalent 2D
depending on geometry loading and symmetry. Plane strain, plane stress and axisymmetric are
115
the 3 ways for transforming 3D to 2D. Plane strain and axisymmetric idolization is common in
geo mechanical problems. Strain in one direction is negligible for plain strain.
There is an axis of symmetry about which attributes remain same for axisymmetric problem. So,
this is all about 3D to 2D idealization. In the next lecture, we will see some basic mathematical
formulations relevant to plane strain, plane stress and axisymmetric cases. Thank you.
116
Lecture – 14
Mathematical Formulation Plane Stress Plane Strain
Welcome back, in this lecture, we will see the mathematical formulation for plane strain and
plane stress. In the last lecture, we discussed how a given 3D problem can be idealized to a 2D
cases. And we understood that it is plane strain, plane stress and axisymmetric condition. So, for
a simple linear elastic isotropic case, we will see how the mathematical formulation of plane
strain and plane stress would look like.
So, we will start with the general linear elastic equations where εx is related to σx, σy and σz as
given. So, εx = σx / E - μ / E (σy + σz) this is where we start with a typical linear elastic problem.
Similarly, we have εy, εz the engineering shear strain γxy, γyz and γzx, one can also represent γxy in
terms of the pure shear strain or tensorial shear strain.
Which is given by εxy the relationship is it is 2εxy = γxy. So, γxy = 2εxy. So, one can also replace
this by this in that case, you will have this particular side it will change to γxy = 2εxy. So, then it
will be εxy = τxy / 2G. So that is the difference. So, we will start with this now, for plane strain or
plane stress now, in the case of planes train case, the given condition is the strain is equal to 0.
117
So, we need to have the inverse of this relationship as well that is stress in terms of strain. So,
first let us now this is strain in terms of stresses. So, we will try to find the inverse of this that is
stress as a function of strain. So that is what we will do first, now, we need to do some
mathematical rearrangement and then find out what is the expression for σ. So, we will start this
we need to obtain the equations of the form σ = Eε.
So, for that, I strongly suggest all of you to follow these steps and work it out for yourself,
because in this it may look a bit abstract. So, when you when we discuss the whole formulation,
then it will be easy for you but I strongly suggest all of you work it out on your own. So, first
what I will like to do is I will try to eliminate one of these stresses. So, one of these stresses gets
eliminated for that what are the different steps to be followed. The first step is multiply equation
a by μ. So, you will have μεx = μσx / E – μ2 / E (σy + σz).
Then do g + b. So, this equation is g, why should you do g + b? In the process of doing g + b. So,
this is g + b we will eliminate σx. So that is the whole idea in whatever manner you can eliminate
you can do that but this is one particular step g + b will give μεx + εy = μσx / E – μ2 / E (σy + σz) +
σy / E - μ / E (σx + σz). Now, if you simplify this, the left hand side remaining same you will get
σy / E - μσz / E – μ2 / E (σy + σz).
118
So, in the process of this, we have got rid of σx. So, μεx + εy = (1 - μ2)2 σy / E, if you combine
these 2, this one and this one, you will get this - μσz / E (1 + μ). So, again if you rearrange this
and this, you will get this expression. So, now, you are left with 2 stresses, we need to eliminate
again 1. So call this as h, then similarly, you do g + c.
So, g + c once you do you will get μεx + εz = σz / E – μσy / E – μ2 / E (σy + σz) and rearranging,
you will get (1 - μ2) σz / E – μσy / E (1 + μ), you can see a lot of similarity between h and i. Now,
it is very easy for us to again simplify this multiply h * (1 – μ) / μ, you multiply this particular
equation by (1 – μ) / μ. So, we will be left with (1 – μ), μ and μ get cancelled off, So, it will be
(1 + μ) εx + (1 – μ) / μ εy = (1 - μ2)* (1 – μ) / μ *σz / E + σz / E (1 - μ2). So, μ again gets cancelled
off from this particular expression. So, call this as j.
Now, add i + j because you will have this is j, this is i, one of the stress we will get again
eliminated that is why we have multiplied it by (1 – μ) / μ. So, when you do i + j we will get εx +
(1 – μ) / μ εy + εz = (1 + μ) / μ * σy / E [(1 – μ)2 – μ2] simplifying, we will get μ (εx + εz ). So, you
can combine these 2 because this multiplied by μ. μ (εx + εz) + (1 – μ) εy. The same expression by
μ, here again you can do the simplification (1 + μ) * σy / E [(1 – 2μ)]. So, this μ and μ goes away
finally, we will be left with the expression for σy. So, in the process σz also got eliminated. So,
you can see that σy = E/ (1 + μ) (1 – 2μ) * [(1 – μ) / εy + μ(εx +εz)]. Now, once you got the
expression for σy it is more or less the same for σx and σz.
119
So, this is equation A similarly, you can write σx and σz the expression which is A, B and C. So,
this is the required form for applying the condition for plain strain, you can also do by after
applying the condition for plain strain you can simplify. So, we have done it before, so that it
becomes very easy to just apply the condition for plain strain. So, you have linear elastic
equations in terms of epsilon that is strain in terms of stresses and now, we have inverted this and
we have got now stress in terms of strain.
So, these are the equations, you can note that these remains same E/ (1 + μ) (1 – 2μ) and here it
is [(1 – μ) / εy + μ(εx +εz)].
So, now, it is easy, we know for plane strain conditioned strain in one direction is 0. So, εz = 0,
γxz, γyz = 0. If we put this in the expression, we will get that is we are substituting in equations A,
B, C we will get σx = E/ (1 + μ) (1 – 2μ) * [(1 – μ) εx + μεx]. So, σz the expression is this is
slightly different from the earlier one because here the contribution of σz goes away.
So, you are left with μ E/ (1 + μ) (1 – 2μ) * [εx + εy] and τxy = G γxy, earlier we have written γxy =
τxy / G or it is 2G εxy where you are replacing γxy / 2, εxy. Now, one important point is like even
though σz = 0, σz is not equal to 0 for plane strain condition. So, these equations are known as the
constitutive relationship for plane strain for linear isotropic elastic material.
120
So, constitutive relationship, we know that is the relationship between stress and strain. And it
represents the material characteristics as well. Here, this is specific to plane strain condition, we
have derived or we have formulated equation the constitutive relationship for plane strain
corresponding to a simple linear elastic isotropic material.
So, we are just rearranging that the whatever expression that we got for plane strain in matrix
form. So, you have σx, σy, τxy these are the 3 stresses which are present whatever we have got in
the previous slide, we are just rearranging it and you have E (1 – μ)/ (1 + μ) (1 – 2μ), E μ/ (1 + μ)
(1 – 2μ). So, these are symmetric the other components are 0 and there is 2G corresponding to
τxy. In addition, we also know σz = μ (σx+ σy).
So, if you substitute this in the expression you will get this particular value as well. So, σz is not
an independent stress. So, if you just insist on this particular matrix that is sufficient for solving
the plane strain problem, because σz is again a function of σx and σy it is not an independent
stress. Hence, the above matrix formulation is sufficient for plane strain condition.
121
Now, constitutive relationship for plane strain in terms of principle stresses. So, some more will
go away. So, you have σ1= E/ (1 + μ) (1 – 2μ) *[(1 + μ) ε1 + μ ε3] what is the difference in earlier
formulation and planes principle stresses, the shear stress component goes away that is the only
difference replace σx, σy, σz by σ1, σ2, σ3.
So, here σ3= E/ (1 + μ) (1 – 2μ) *[(1 + μ) ε3 + μ ε1] and σ2= μ E/ (1 + μ) (1 – 2μ) *[ε1 + ε3]. And
σ1 + σ3 if you take σ1 + σ3 you will be left with E/ (1 + μ) (1 – 2μ) *[ε1 + ε3]. Now, you please
compare this equation with σ2 it is same. So that is how you can very well replace σ2 equal to
only this μ is additional here remaining E/ (1 + μ) (1 – 2μ) *[ε1 + ε3] is σ1 + σ3.
So, one can always write σ2= μ (σ1 + σ3). Earlier we have returned the same kind of expression.
So, it is similar to σz= μ (σx + σy). So, here in terms of principle stresses one can write σ2= μ (σ1
+ σ3).
122
So, constitutive relationship for plane strain in matrix form one can write this is exactly in terms
of 2D, see the effect of σ2 is not considered here, you can write
here we have considered only 2 dimensional stresses. So, what we have done is the mathematical
formulation for plane strain.
Now, let us see for plane stress, it is the condition of one Stress. Stress in one direction is 0. So,
we will have to use the very first equation of strain in terms of stresses and substitute this
123
particular condition. So, we will see how it will look like. So, you will have εx = 1 / E *(σx - μ
σy), εy, εz and γxy which is equal to τxy / G or γx = 2 (1 + μ) τxy / E.
Because you have replaced G by this expression we have the expression E = G (1 + μ). So, by
substituting that you have obtained in terms of E. So, here is a mere substitution of σz, τzx, τzy = 0.
So, we are left with these equations. So, here also we have to note that even those σz = 0, εz not
equal to 0, we know this, because we have the contribution of σx and σy in the other lateral
direction that strictly It will also influence the string. So, hence εz is not equal to 0. So, there are
4 strain components and 3 stress components altogether.
So, strain tensor for plane stress is denoted in this particular manner, εx εxy and we have εz as
well. Stress equations for plane stress, by inverting the previous expression, we can get σx= E/ (1
– μ2)* εx + E μ/ (1 – μ2)* εy, we are doing the inversion, it is a simple rearrangement. So, one
can always try this, you will get σx= E/ (1 – μ2)* εx + E μ/ (1 – μ2)* εy and you have σy and τxy.
So, this is stress in terms of strain. So, to write the stress tensor, you will have = E/ (1 – μ2), E μ/
(1 – μ2) to 0. So, E μ/ (1 – μ2). So, this is the expression for this is the matrix representation of
stress for plane stress condition.
124
And εz = - μ / E *(σx + σy) substituting the expression for stresses that a σx and σy we have seen in
the previous slide, if you substitute it, one can get εz = - μ / (1 - μ) *(εx + εy).
So, in this lecture, we have done the mathematical formulation for plane stress and plane strain
condition. So, plane strain 2D idealization is more prominent in geo mechanics than plane stress.
Constitutive relationship for both plane strain and plane stress discussed for a simple case of
linear elastic isotropic material. Plane strain has strain in one direction to be negligible. That is
this particular expression, εz, γxz, γyz = 0.
125
And also even though εz = 0, σz is not equal to 0 in the plane strain condition. Similarly, we have
even though σz = 0, εz not equal to 0 in the plane stress condition. So, this is all about the
mathematical formulation of plane strain and plane stress. In the next lecture, we will see the
mathematical formulation for again for the simple case with respect to axisymmetric condition.
Thank you.
126
Lecture – 15
Mathematical Formulation Axisymmetric
So, welcome back in the last lectures, we have seen the importance of 3 dimensional to 2
dimensional idealizations. We have seen the mathematical formulations related to 2D
idealization where the important 2D idealizations we have seen are plane strain, plane stress and
axisymmetric condition. Out of this the mathematical formulations related to plane strain and
plane stress we have discussed. So, in todays lecture, we will see the mathematical formulation
for axisymmetric condition.
So, here we are dealing with linear elastic isotropic condition, in the previous lectures also we
have done for the simple linear elastic models. So, here also we will see for the linear elastic
formulation. Now, in this axisymmetric condition, we have already stated that, this is very
important with respect to soil behavior, why because the triaxial samples what we discuss falls
under axisymmetric formulation.
So, we will now, in the formulation for axisymmetric condition, we will specifically focus on
triaxial sample, you can see that there is this cylindrical soil sample which is acted upon by
127
actual stress and actual stress is normally considered as the major principle stress and the radial
stress the order on stress or the radial stress is the minor principle stress. So, this is actual and
this is a radial. So, this is a situation so, if you see that if σ1 is exactly acting at the center is an
axial load.
So, it has to act at the central axis. If this is considered as the reference, then this is symmetrical
about the application of this load, it is also symmetrical with respect to its geometry. So, it is
important that these are met. Additionally, in this particular formulation or in this particular
triaxial sample, we know that σ2 = σ3 the radial stresses or all around stresses are same and to
maintain the geometric symmetry.
So, we also need to have ε2 = ε3 that is for maintaining geometric symmetry. Now, we know the
formulations which we have already seen before the formulation was let us say in general εx = σx
/ E - μ / E (σy + σz). So, this is the general equation, now, we are considering in terms of
principle stresses. So, we have ε1 is the major principle strain is equal to σ1 / E - 2 μ σ3 / E.
Why? Because here σ2 = σ3 hence, it will become - 2 μ σ3 / E. So that is denoted as equation 1.

Then we have ε3 = σ3 / E - μ (σ1 + σ3) / E. So, there is no need of ε2 because we ε2 = ε3. So, ε3 =
σ3 / E – μ (σ1 + σ3) / E and that is denoted as equation 2 and ε3 = (1 – μ) σ3 / E - μ σ1 / E.
That is if you rearrange this σ3 and - μ σ3 / E we get (1 – μ) σ3 / E - μ σ1 / E. So, based on ε1 and

ε3, one can always arrange the constitutive relationship for axisymmetric condition in matrix
form. This we have done for plane stress and plane strain as well; the same constitutive
formulation has been done for axisymmetric condition as ε as a function of stress.
So,
Now, one thing we have to note here is, if you consider this particular matrix, this is not a
symmetric matrix, you have – 2μ and – μ here. So, this is something different from what we have
seen earlier, where we had a symmetrical matrix. So, in this case it is not symmetric. Hence, the
128
inverse may not be straightforward. I mean to say if I get the equation for σ1 that will not apply
for σ3, so that we will have to see separately. So, this is the constitutive relationship in terms of
ε1 and ε3.
Now, we also need to know what is the inverse relationship that is what will be σ= E ε. So, we
also need to know this particular inverse relationship, this we have seen in the earlier lecture also
we need to do some mathematical rearrangement and simplification, so that we get the
relationship in terms of σ. So, for that, we need to do some modification for equation 1. So, 1 is
multiplied by μ which gives μ ε1 = σ1 μ / E - 2 μ2 σ3 / E and that is denoted as equation 3.
Why we are doing this? We are doing this to eliminate one by one and we are left with only 1
stress and that will give you the required equation. So, we are just eliminating it. So, then this
equation plus equation 2, equation 2 is first ε3 that will give you μ ε1 + ε3 = σ3 (1-μ) / E - 2 μ2 σ3 /
E. I again strongly urge all of you to please write it down step by step and make sure that it is all
right.
Again, if you can see with this particular operation, you can see that this is now in terms of σ3.
So, σ3/ E if we take it outside we have (1 - μ - 2 μ2) and with some adjustment that is plus μ
minus μ, you will get a proper equation. So, σ3/ E *(1 - 2μ - 2 μ2 + μ). Further, if you again
rearrange this we will get σ3/ E *[(1 +μ) - 2 μ (1 +μ)]. So, again (1 +μ) can be taken outside this
129
will give σ3 as E divided by, so if you take (1 +μ) outside here it will be σ3 (1 +μ) and you are
left with 1 - 2μ. So, σ3 will be equal to E will go on the other side E will go here. So that will
read E /(1 +μ) (1 - 2μ) * [ μ ε1 + ε3 ] which is from here.
That is the equation for σ3 in the earlier lectures we have seen that once we have got it for 1
particular stress, you can just reproduce it for other stresses as well there in that particular case,
you could see that the matrix was symmetrical where ε in terms of σ was symmetrical but in this
case, it is not symmetrical.
Hence, we cannot reproduce for σ1 we have to again find it separately. So, we will see how if
you multiply equation 1* (1 – μ)/ 2 μ. Why we are doing this? We need to find σ1 expression. So,
multiply (1 – μ)/ 2 μ. This will give (1 – μ)/ 2 μ * ε1= (1 – μ)/ 2 μ * σ1/ E - (1 – μ) * σ3 / E. So,
call it as equation 4. Now, if we again do this operation 4 + 2, we will be left with this (1 – μ)/ 2
μ * ε1 + ε3 = (1 – μ)/ 2 μ * σ1/ E – μ * σ1/ E.
So, σ3 got eliminated in this operation. So, this we are left with now σ1 now, we need to rearrange
it, if you multiply by 2 μ, we will be left with (1 – μ)* ε1 + 2 μ ε3 = 1/E [(1 – μ) * σ1 – 2μ2 * σ1].
So, σ1 /E [(1 – μ) – 2μ2]. So, this again can be further simplified which will give [(1 – μ) - 2 μ2] =
(1 + μ) *(1- 2 μ) substituted back here.
130
We can write the expression for σ1 = [E/ (1 + μ) *(1- 2 μ)] *[(1 – μ)* ε1 + 2 μ ε3 ].
So, one can again rearrange the constitutive relationship for axisymmetric condition in matrix
form that is stress in terms of strain as
Now, let us see what is the application of these constitutive relationship we will see this with a
simple example. Whereas, the same constitutive relationship can be used to solve complex
problems as well. But for understanding it better, we will see a very simple example which may
131
not need constitutive relationship as such to solve the problem but for us to understand how this
can be applied. So, how to apply constitutive relationship for plane strain and axisymmetric
condition.
I am purposely avoiding plane stress because we are in advanced soil mechanics this particular
course and specifically in geomechanics, we will have a lot of plane strain and axisymmetric
cases. So, the first example is that there is a retaining wall with outward movement, outward
movement means, we know that it is an active condition also lateral strain and vertical strain of a
soil element behind the wall are given, we need to calculate change in stresses acting on the wall
to calculate the lateral force per unit length acting on the wall.
Now, we are supposed to calculate what are the changes in the stresses, due to this movement.
Now, due to this movement outwards which is an active condition the strains are given. I am not
I purposely avoided the values here, because the same will be attempted in assignments, you just
need to understand how or what is the logic by which we can use the constitutive relationship.
So, retaining wall, what is the speciality of a retaining wall? We have already seen that the
behavior of a retaining wall can be idealized.
Where 2d behavior and that too in terms of plane strain, because the length in one direction is
long as compared to the cross section and the strain in one direction. So, we are dealing in
principle stresses and principle strain. So, the ε2 can be considered equal to 0. So, this is a
retaining wall we are talking about a given soil element whose lateral and vertical strain is given
and this retaining wall is moving outward.
So, we know since it is a plane strain problem, we know that we can apply this particular
constitutive relationship. So, here I have added Δ σ1 just to show that it is change in stress. So,
here you note that this is a symmetric matrix which is different from the asymmetric matrix that
we obtained in axisymmetric case, now, what is given .
132
Now, once we know , we can always calculate change in stresses by knowing the elastic
parameters also. One more thing we need to understand here is in this problem we are assuming
everything to be linear elastic. Now, how far it is true, how far it is close to reality that is
something different but here we are assuming the behavior to be linear elastic. So, this is an
active condition.
We have already discussed that horizontal change in stress will be Δ σ3 in active condition; we
have minor principle stress acting horizontally. So, what we are looking at is Δ σ3 now, once you
substitute the values of Δ ε1, Δ ε3, μ and E we can determine Δ σ1 and Δ σ3. Now Δ σ1 is in
vertical Δ σ3 is in horizontal because it is a moment outwards.
Now, once we know Δ σ3 we need to find out what is the lateral force per unit length. So, you are
just determining the force. Now, what is the force this can be done by integrating Δ σ3 for the
entire height of the retaining wall this will give us lateral force per unit length same problem
with its numericals we will see in the assignment.
Another case is given an oil tank founded on soil. Vertical and lateral stresses at the middle of
the soil layer are known to determine the strain and elastic vertical settlement. So, it is specified
here it is elastic vertical settlement. So, our problem becomes linear elastic. Now, cylindrical oil
tank with soil element beneath the central axis it is satisfies the axisymmetric condition, we will
133
see how, now consider this to be the cylindrical oil tank, now, we are considering a soil element
which is exactly beneath the central axis of this cylindrical oil tank.
If you are considering this then it obviously satisfy the requirements of axisymmetric condition
once we know it is axisymmetric we can always use the corresponding equation or the
constitutive relationship and
. So, this is the equation which we will be using you can see that this matrix is asymmetric.
Now, change in strain due to load from circular tank can be determined if we know the stresses
and the elastic parameters. Now, what is given vertical and lateral stresses at the middle of the
soil this is given already? So, once we know this we can always calculate what is the strain.
Now, settlement is mostly governed by the vertical strain Δ ε1. So, once we substitute this Δ ε1 is
obtained using Δ ε1. we can calculate the settlement how by integrating for the entire height. So,
settlement can be determined by integrating the strain Δ ε1 for the height of the soil layer H.
So that is what we have to do once you get this problem for assignment, then you will understand
it better. Now, we will discuss a bit on invariants for plane strain and axisymmetric condition,
what are invariants? We have already discussed which is not affected by the coordinate axis. So,
134
those invariants in terms of plane strain and axisymmetric condition we will see. So, we have
already discussed, what is the general expression for deviatoric stress which is q?
And we have also known that this particular deviatoric stress basically comes from the definition
of von Mises failure criterion. So, here
Now, what is the implication once you consider it to be plane strain and axisymmetric because
we know the conditions. So, we just need to apply that into this equation and get the
corresponding expression.
The next invariant is mean stress
Volumetric strain εv = ε1 + ε2+ ε3 and deviatoric strain
So, we need not repeat it again because this we have already discussed. Now, for plane strain, we
know ε2 = 0, there are no other shear strain. So, we do not have to discuss we are discussing in
terms of principle stresses and strains.
So, ε2 = 0 that will give this equation, we need to substitute ε2 = 0 that will give the expression
So, we have got the equation for deviatoric strain corresponding to plane strain condition. And
here in case of volumetric strain it will be simple εv = ε1 + ε3 because ε2= 0.
135
Now, for axisymmetric condition we know the conditions are σ2 = σ3 and ε2 = ε3, we just need to
substitute this into the previous equations to get
Please note this expression does the mean stress this is very important and we will be using this
quite extensively throughout this course. So, please make a note of this mean stress and this is
specifically valid for triaxial condition.
22:00
Where σ2 = σ3 and ε2 = ε3. So, we have obtained, q will also change, into this = σ1 - σ3,
you will be very familiar with this particular form of deviatoric stress. Now, when you consider
triaxial testing, we know that there is a deviatoric stress which causes shear failure and there we
have taken σd = σ1 - σ3 where it comes from, it comes from here from this general expression.
When it satisfies the conditions of access symmetry, then you get σd = σ1 - σ3 which is popular
and we know this, because we have studied triaxial test. So, it comes from the So, the genesis is
q in this particular form and εv = ε1 + 2ε3 and εq = 2 / 3 (ε1 - ε3) your mere substitution gives these
expressions. We also know bulk modulus K = p’ / εv and shear modulus G = q / 3 * εq.
Now, please make a note here if you consider bulk modulus it is mean stress upon mean strain or
volumetric stress upon volumetric strain gives bulk modulus but in the case of shear modulus G
136
it is deviatoric strain upon 3 εq where is 3 coming from again this will be dealt as an assignment
problem. But you may just look at it from where it will come. So, once we know bulk modulus
and shear modulus in this expression, we can also write the constitutive relationship in this form
where
We all already we have discussed about the decoupled behavior by which the soil is studied
which means to say the volumetric component and the shear components are separated or the
deformation components are separated. So, here this means stress deals with the volume q deals
with the deformation or the shear behavior and that K, 0, 0, 3G, 3G comes from here, because 3
goes here. So, 3G is q into εq εv εq. So, this is how it is done. The task for you is to understand
where this 3 comes from.
So, to summarize this particular lecture axisymmetric 2D idealization is predominant in soil

mechanics. This is very important we need to understand more problems relevant to
axisymmetric condition. Both geometric and loading symmetry is maintained in axisymmetric
condition. Constitutive relationship for axisymmetric condition is discussed for linear elastic
isotropic material. And the invariants are discussed for plane strain and axisymmetric condition.
137
So, with this, we are almost nearing to the end of module 1 where we have discussed some of the
basics of continuum mechanics which may help you to learn further. So that is all from this
lecture, we will see in the next lecture. Thank you.
138
Lecture – 16
Summary of Module 1
Welcome all of you, till the last lectures we were essentially discussing about module 1 and with
the mathematical formulation of axisymmetric condition, we came to the end of module 1. So,
before going to module 2, let us understand and summarize what actually we have learned in
module 1.
The module 1 was on introduction to continuum mechanics and strive to put forth a way of
perceiving stress, strain and stress strain relationship which we call as cause-effect relationship.
Now, you would have noticed the initial few lectures, I was fairly slow in delivering the lecture
and I have used repetitions in between the sentences. This was done intentionally, because some
of the participants may be new to this topics and it will be easy for them to grasp better.
So, I have made it intentionally a bit slow. And for those who already know the concept or
already know the subject, well, it will be a bit dragging for them that is fine but the whole
purpose was to deliver the concept to the maximum possible. We started of with the concept of
139
stress acting on a body or stress acting at a point in the body. And for that, we have discussed
Cauchys hypothesis, the concept of traction based on the traction and using Cauchys formula.
We have evolved Cauchys stress which is the stress acting at that point. And it was a 3 by 3
matrix formulation and 9 stress components. Later, we have seen that there are 6 independent
stress components. Then, another important aspect was introducing the concept of tensor, what is
a tensor and which is very essential when you deal with higher realms of geomechanical
problems, mostly stresses are dealt with in the form of tensors.
A tensor is a quantity with magnitude and direction we have discussed that and the stress tensor
we have understood that it is a second order tensor. Application of stress tensor for determining
stresses acting on a plane the normal and shear component of traction which we have understood
as normal stress and shear stress which are the stresses acting on a plane, we have discussed that
based on the knowledge of stress tensor.
Then a very important aspect of transformation of coordinate axis which is very important for
evolving principle stresses. So, we have understood how the given stress components can be
transformed to an altogether different axes a set of different orthogonal axes. Then comes
determination of principle stresses based on the transformation, we have discussed how to evolve
principle stresses by diagonalizing the given stress tensor.
We have discussed about characteristic equations and also understood that the Eigenvalue
problem. So, characteristic equation when you solve the parameters of the equations are called
invariants. We discussed about principle stresses and also the direction cosine matrix which we
get based on Eigen vectors. So, Eigen vectors is again by solving the matrix equation the
characteristic equation you get the normal vector.
And that normal vector when arranged in a matrix form gives direction cosine matrix for
transforming the given set of stress component to a given axis. And the concept of invariants was
discussed which is very essential in most of the geomechanical problems, the stress strain
140
relationship or the constitutive relationship when you try to solve various problems, you will
know that it is mostly in terms of these invariants essentially for soils and rocks.
Volumetric and deviatoric stress concept, we have seen that the whole of the stress is divided
into or it was decomposed into volumetric component and deviator component. So, volumetric
stress and deviatoric stress we have seen. Then, we discussed about invariants of stress tensor i 1,
i 2, i 3 and in variants of deviatoric stress tensor j 1, j 2, j 3 we have discussed majority of the
aspects and the relationship.
Having said about stresses then we discussed a bit about strain and we also told that whatever
was relevant for stresses, all of these concepts were relevant for strain and hence, we did not
spend much time discussing strain. So, we discussed about linear strain as well as shear strain,
the strain tensor all those things we have discussed. Once that was done both cause and the effect
both were discussed further, we went on to cause-effect relationship.
We discussed about very simplistic linear elastic model and the constitutive relationship. Then,
we have seen some prominent constitutive relationships which are essentially there for geo
mechanical problems. But we did not go into the details of it, some of the very relevant models
we have seen. And for any further understanding, the participants are suggested to read the
advanced books in geomechanics for constitutive relationship and understanding its formulation.
141
We have not discussed about any formulation in this course and it is out of scope of this
particular course, you can refer to any specific courses on geomechanics. Then we discussed 3D
to 2D idealization which is very important for solving problems and making it more simple in its
approach. So, then we have discussed in terms of 3D to 2D we discussed about 3 conditions, the
first one was plane stress, plane strain and axisymmetric condition.
And we insisted on the point that in geomechanics plane strain and axisymmetric condition is
very important. Then, we discussed about the mathematical formulation for plane stress, plane
strain and axisymmetric condition, considering linear elastic isotropic material because that is the
most simplistic model that we could discuss in this particular course. Most of the books referred
for this particular module specifically include the 1 authored by Singh A.K, Das B.M, Potts and
Zdravkovic and the book by Budhu M.
All these books are listed in the reference. In addition to this I have also gone through some web
materials for making the concept in a more understandable manner. So, I would suggest the
participants also to do this that apart from whatever you have gone through in this slide, you are
suggested to read through more related to this topic. So, then this is all from for module 1. Now,
we will move on to module 2 which is basically on shear strength. Thank you.
142
Lecture - 17
Basics of Shear Strength
Welcome back one and all. So, we have been dealing with module 1 on continuum
mechanics. Till now from today's lecture onwards, we will see module 2 which is shear
strength of soil. Now you may be wondering why I have chosen shear strength of soil in this
advanced level course? From the past few years of experience with the students who are
coming for M Tech, I have noticed that some of these concepts are not conceived properly,
some of these concepts related to shear strength are not conceived properly.
And this is very vital for module 3 and module 4. So, we need to have the concepts clear
about shear strength of the soil for getting a hold on module 3 and module 4. So that is the
reason why I chose to explain shear strength of soil in detail and some interpretations. So,
those who have understood this topic very well during their undergraduate, you may just
refresh it. The reason is, there are several ways by which people understand this very concept
of shear strength.
If you just read through the textbook once, you may not find it anything complicated,
everything is in place, the moment you start reading more, you are there for a lot of surprises
and some sort of complexities which induce a lot of confusion in understanding this topic.
This is only possible if you start reading more. So, shear strength understanding there are a
lot of nuances and these nuances, we need to understand it very carefully.
So that things are clear. And I would also like to make a point here that whatever we
understand is based on my viewpoint, I grasp this subject in a particular manner, another
person may grasp it in a different manner. And this we have also noted when we talk about
this subject to students, as well as our faculty colleagues. So, we have some sort of kind of
differences in what we have understood.
So, I do not say or I do not claim that whatever I have understood is the exact one, somebody
may dispute it. So, what I will be presenting here is from the majority of the facts which has
143
been reported in some of the best and the popular textbooks. So, it is very important that we
understand the shear strength of soil clearly. So, with this, let us start this lecture is about the
basics of shear strength and the following lectures will be on interpretations and certain
aspects which we need to keep in mind when we deal with module 3 and module 4.
So, soil mass subjected to unequal loading induces shear stress and causes particle slippage.
What is the importance of this sentence unequal stresses? So, if you have noted if the order
on stresses are same can you induce shear in the soil? Can it bring about failure? If you look
from the perspective of the normal loading, normal loading means not σn, the normal range of
loading that is subjected to the soil.
Possibly it will not cause any sort of failure when it is isotropic or when the order on stresses
are same, it will induce volume change. So, what causes failure in the soil? Soil is a granular
material. So, there has to be unequal stresses which results in particle movement and that
causes deformation and this is what you call it as shear stress. So, shear stress causes particle
slippage and this shear stress is caused due to unequal stresses.
For granular materials like soils, shear stress and shear strength becomes important. Stress the
limiting condition is strength. Failure of soil mass is generally associated with the formation
of a failure plane or slip plane. So, the whole definition of shear strength theory is based on
the assumption that there is a failure plane that gets formed within the soil and I do not have
to pictorially show this even though we will be having it in the subsequent slides because
these facts you are expected to know by the undergraduate learning.
144
Now a very important point is the critical combination of stresses acting on the failure plane
reaches its maximum limit. Now this particular phrase is very important, critical combination
of stresses. So, which we will appreciate better as we move further? Shear strength is defined
as the limiting value of shear stress that the soil can withstand with reference to the failure
plane, see it is always done with reference to the failure plane and it is the limiting value of
shear stress.
Shear strength is equal to the maximum shear stress only under certain conditions. Now
please read these 2 sentences together, what is that? In the first sentence we say that it is a
limiting value of shear stress which is associated with the reference failure plane. So, this
failure plane. So, it is not actually equal to the maximum shear stress in the soil but the
maximum shear stress becomes the failure shear stress under certain conditions. We will
discuss this when we discuss more about undrained shear strength.
Since there is slippage of particles, the inter particle interaction induces strength in the soil.
So, any particle which moves over one another, there will be a sort of friction which gets
developed. So, strength mobilization is mostly due to frictional resistance between the soil
particles. A criterion that decides the failure state of geo material is then necessary. Here geo
material is concerned with the soil. So, when will the soil fail, we need to mathematically
define this and that definition is what you call it as failure criterion.
So, failure criterion it indicates when the soil fails for a given condition of stresses and
material property. So, this is what the strength failure criterion gives the failure state with
respect to material properties and other stresses. So, Coulomb's friction model is the simplest
model to start with why? Because Coulomb's, Coulomb has talked about the friction. So, this
becomes the simplest model to understand what kind of strength the soil imparts.
145
So, let us go into the simplistic explanation of Coulomb's friction model. There is a block and
it has a horizontal force, T is the resistance, W is the weight, N is the reaction. So, this is R
the resultant. So, it acts at φ’. Now if you translate this to soil, we can see that we have 2
particles and there is a fictitious area at the point of contact and this is the N is the force and
T is the horizontal force. N is the normal force and T is the horizontal force at a actual area of
contact between the 2 particles.
So, we can very well write T limiting condition is equal to μ into N where μ is the coefficient
of friction and we can write T limit = N tan φ’, μ = tan φ’. Now we are just translating these 2
soil and in stress form, one can write τf = σ’f * tan φ’. Here σ’f is the, we can write σ’nf which
is normal stress. Now we need to be very specific here, it is the normal stress acting on the
failure planes.
So, τf is the shear strength or the failure shear stress which is acting on the failure plane. So,
friction model is the basis for mathematically defining shear failure of soil. That is what I told
there are different ways by which we understand shear strength. Now whatever be we need to
define this in a mathematical framework then only as engineers we can use it further. In
addition for soils there are several factors which need to be considered.
Now for a normal block model several other complications will not come into picture but for
a geo material like soils, we also need to consider various other aspects. We need to consider
load application which induces volume change. Now if you remember in module 1, we told
146
that we will study the soil in a decoupled manner, we will consider volume change separately
and we will consider the deformation separately.
Now for stress strain response of and various other continuum mechanics problem this is fine
but when you consider actually the shear strength that is the shear stress which gets induced
during loading, you also need to consider what type of volume change response it undergoes
under a given load because that is going to change this state, the volume change is going to
change the state. It is also complicated because of the voids and the pore water pressure
present in it.
Pore water pressure keeps changing. So, there are 2 conditions drained and undrained. So,
depending upon that we will come to that a bit later. Depending upon the condition of
drainage each one may have volume change or we may be left with pore water pressure.
Whatever be throughout the loading process this keeps changing. So, limiting shear stress is
governed by also strain level, rate of straining and stress history.
So, what will be the limiting shear stress or the shear stress at failure will be governed by
strain level. What percentage of strain is it subjected to? Or rather we can fix the strain level
as a reference point; we consider it to be a failure state. Rate of straining how fast it is loaded
and stress history, stress history plays a very important role in shear strength. When we
discuss further this will be clear.
147
Now, extending Coulomb's model that means, we have now consider only the frictional
aspect. Now there are certain cases where soil will have cementation for example, certain
precipitation of certain minerals can create a sort of cementation, a very good example and
some of the state of the art research that is going on these days is microbially induced calcite
precipitation which enhances the strength of granular soils. So, what it is done is?
There is a sort of cementation that happens. Now this cementation also enhances the strength
of the soil. Now we need to account that till now in the Coulomb's model that we discussed
this has not been considered. So, for cemented soil one can always write τf = c’ + σ’f * tan φ’.
Now you need to understand here we have specifically used c prime, σf prime and φ prime or
φ’, we call it in both ways. Now why this is so?
Here it needs to be specified that this model is specific to effective stress condition. And then
you may ask why this quantity does not have prime or dash because water does not have
shear strength. So, whatever be both effective and total condition will be the same. So, this is
dealt mostly in terms of effective condition and here I would also like to remember you
effective stress principle.
Now it depends on how well you have been taught this principle during your UG because this
is extremely important and it is based on this the whole of the soil mechanics is founded. And
we do not talk about effective stress condition in any other material; we talk mostly in
granular material and specifically for soils and rocks. Now what is effective stress principle?
The first statement defines what is effective stress but the second statement where we state
that the strength and compressibility of the soil is entirely governed by effective stress.
Which means to say for, it to be very specific, long term condition or long term strength of
the soil is completely governed by effective stress. So that is why this model which
specifically deals with long term condition is to be stated in effective stresses. So that is why
we have specifically used now dash. Now from here on, we have to clearly distinguish
between total stress and effective stress, c’ and φ’are the shear strength parameters.
Now it is very easy to state the strength of the soil defined in terms of c’ and φ’. And there
are several other misconceptions also associated with this which is generally not discussed
during UG. So, those are missed that it is better that you refresh that now, c’ can be true or
148
apparent cohesion. Now here c’, what is c’ and from where it comes from c’is either a
cementation characteristic.
If it is cementation we call it as a true cohesion. But it can also be due to certain

characteristics of the pore water. Pore water can induce a sort of cohesion depending upon the
state of saturation. So, you call it apparent cohesion which means, it is not going to be there.
Another possibility of this cohesion is for over consolidated state, again I will not get into the
details now, we are going to discuss that in length.
But this cohesion is dependent on certain conditions, whether it is going to be there or not is
not sure, it is condition specific. That is why it is called apparent cohesion and true cohesion
refers to the cementation which is there for sure. So that distinction one has to make
sometimes c’ becomes a mathematical artifact as well, so that is why in some of the designs,
we tend to ignore cohesion part and we consider φ’ which is more relevant for granular
materials like soils.
Now when I say this that I am sure some of you may not understand or some of you would
not agree with me. You would immediately say that a clay will have cohesion and a sand will
have friction. So, this is the manner in which we are normally taught but probably subsequent
lectures have not discussed what is the extension of that? It is because of that. So, we will see
this in detail and this is where we need to clarify our concepts clearly.
And it is to be noted that these are not fundamental properties of soil. And it will depends on
the type of test what type of test? What type of shear strength test you are doing? That will
also govern what type of values you are getting for c and φ. It will also depend upon the
manner of loading, drainage condition and the initial stress state we will not get into the
details now, we will discuss in detail and very importantly, the stress history.
So, Coulomb's friction model represents the stresses acting on the failure plane that we have
seen, if you refer to this, it talks about the stress τf and σ’f which is the stress acting on the
failure plane when I say what does that mean is a typical soil sample which is acted upon by
external stresses, σ’1 and σ’3, major and minor principle stresses. So, the loading on the soil is
represented in terms of principle stresses.
149
Now as we said the use of Coulomb's friction model assumes a failure plane within the soil
mass. Now this failure plane is at an inclination of θ with major principle axis. Now this is
the major principle axis, so it makes an angle of θ. Now the stresses acting at the time of
failure is σ’f and τf this understanding also one should have clearly. Now if the same is
represented in terms of Mohr circle, this is what it is.
So, here this particular a red line is the Coulomb's model which says c’ + σ’f * tan φ’. Now if
you draw it clearly, so this is the center, this is σ’1 , σ’3 of the Mohr circle and please makes a
note, like you need to get into the basics of Mohr circle properly and understand its
characteristics. Now here this angle will be 2 θ that also we know it is a property of the Mohr
circle representation and A C is the radius of the Mohr circle. This is the point A which
denotes σ’f and τf.
So that is what it is about extension of the Coulomb's friction model to account for the
cementation characteristics or apparent cohesion. But Coulomb's friction model if you refer
to this, it does not talk about the stress acting on the soil. Stress acting on the soil is in terms
of σ’1 and σ’3. So, Coulomb's model does not give any indication of what stress the soil is
subjected to.
So, referring to the figure again, is the same figure we can write σ’f = OC, so this point is O.
OC - BC. So, here I am talking about σ’f this is OC - BC. Now you can write σ’ what is OC
and what is BC? Now OC is from here to the center of the circle, so that we can write it as σ’1
+ σ’3/ 2 average stress minus what is BC? BC is AC cos (180 - 2 θ), this angle is 180 - 2 θ.
150
So that will give σ’f = (σ’1 + σ’3)/ 2 + (σ’1 - σ’3)/ 2 * cos (2 θ) and τf = AB and that is nothing
but τf = (σ’1 - σ’3)/ 2 * sin (2 θ). And from the geometry one can always write 2 θ is equal to
this is 90 + φ because this angle if you draw here, this angle will be φ’.
So, it is 90 + φ’ and hence θ = 45 + φ’/ 2 and what is θ? θ is the inclination of failure plane
with the major principle axis. Now if you substitute for θ that is 45 + φ / 2 you will also get
this equation σ’f = (σ’1 + σ’3)/ 2 + (σ’1 - σ’3)/ 2 * sin φ’. I have purposely written this why
because in certain textbooks it is written in this form and in this form. So, you should not get
confused why it is written in this manner. So that is why you just need to substitute it for 45 +
φ / 2 you will get this equation.
So, now, let us come to a more important or prominent failure criterion for soil which is
known as Mohr Coulomb failure criterion. Now we have already stated in the previous slide
that Coulomb's model does not talk about the stress acting on the soil. So, we just want to
make that modification and that is how the concept of Mohr circle has been integrated with
the concept of Coulomb's model.
So, one of the most powerful yield criterion or failure criterion in geo mechanics or soil
mechanics, this is elastic, perfectly plastic model. Hope you remember whatever we have
discussed about elastic perfectly plastic in module 1. So, for up to certain limit, for example,
this is stress and this is strain, so it means you have a response like this. So, up to a certain
151
limit you have elastic characteristics and once it reaches this elastic characteristic the soil
yield.
So that is why it is called elastic perfectly plastic and what is this value? This value is the
limiting condition of stress. It consider Coulomb's friction model represented by PQ. So that
is what it is, the same figure again τ versus σ, plot, PQ is the required Coulomb's friction
model which is stated and that is τ = c’ + σ’ tan φ’. Now consider a cylindrical soil sample
subjected to principle stresses again it is the same figure.
So, failure occurs when more circle touches Coulomb's failure line. Now when will the
failure occur? The failure occurs when the Mohr circle touches the failure line and Mohr
circle is, it represents the stress state of the soil. If that is a condition, if this is the condition
this is met, like at this particular point, the soil is bound to fail. So, sin φ’. this is the
inclination as φ’, sin φ’ one can write AC / PC.
And this probably you would have derived during your UG but I am just revising it sin φ’ is
AC is nothing but the radius, PC is PO + OC, OC is (σ’1 + σ’3)/ 2 and PO is c’ cot φ’.
So, if you rearrange it you can get in this manner and σ’1 = [2 c’ cos φ’/1 - sin φ’] + [σ’3 (1 +
sin φ’/ 1 - sin φ’)]and if you write
Cos
So, if you substitute this for cos phi dash then one can always get this expression,
152
Now this gives the popular Mohr Coulomb model which is
This is the popular Mohr Coulomb failure criterion. Now if you see this expression you are
representing this in terms of the stresses acting on the soil.
So, we will discuss about some facts about the Mohr Coulomb failure criterion. Soils cannot
have a state above Mohr Coulomb failure line and it is essentially an effective stress criteria.
Mohr Coulomb is independent of intermediate principle stress σ’2, we do not discuss about
σ’2. It does not consider strain at which failure occurs, there is no mention about the strain at
which the failure occurs according to Mohr Coulomb failure criteria.
But in certain cases, strain or displacement may become a governing factor which means to
say, it may not be tolerable even though the soil is not failed. So, sometimes the strain based
criteria becomes important. Failure envelope is independent of normal stress that means,
linearity and slope is constant. So, this we need to understand a bit carefully. Failure envelope
is independent of normal stress which means to say τ and σ’, this line has a constant slope. So
that is what it means.
153
So, this entire failure envelope is not dependent on normal stress rather it is a constant you
should not confuse it with failure stress because failure stress is dependent on the normal
stress but failure envelope is not because it is a constant. This may be true, this may not be
true for a wide range of normal stress. May be for a wide range, there is a possibility of non
linearity. For a moderate range, we can still approximate it to be a straight line.
Now if we ignore cohesion from the Mohr Coulomb failure envelope, one can rearrange it in
terms of this and if you take σ’3 outside it will give you
Now this term is called maximum effective stress obliquity. Now you can see that
according Mohr Coulomb failure envelope the failure occurs when maximum effective stress
obliquity is achieved and not when the maximum shear stress is achieved. You can see the
maximum shear stress is (σ’1 - σ’3)/ 2. Now what we are concerned about is maximum
effective stress obliquity.
Coupling Mohr circle with a Coulomb model define shear failure based on stress state of the
soil. Coulomb model and Mohr Coulomb is based on a slip plane along which the soil mass
slides and fail. In actual case, there are shear bands or pockets that reach failure which is
specifically pertain to dense state or it undergoes significant plastic deformation that is strain
154
hardens in the case of loose state. Now we need to read these 2 sentences together. What it
means?
The Coulomb model and more column is based on a slip plane along which the mass slides
and fail this is the assumption but in reality, you may see that there are formation of some
shear bands or pockets which reached the failure faster than the other part of the soil. And
this is a specific, this is mostly specific for dense state of the soil or it may undergo ductile
behavior or a plastic deformation for loose state of the soil. So that is all. We will see the
remaining part in the next lecture.
155
Lecture - 18
Stress Representation
Welcome back. So in the last lecture, we have started module 2 and we were discussing about
the basics of shear strength, we have discussed about Coulomb's model which is applicable
for granular materials like soil then we discussed about integrating Mohr circle with
coulomb's model and that is how we got Mohr Coulomb failure envelope. And we discussed
about some basic facts about Mohr Coulomb failure envelope and how important it is for
materials like soils in defining the failure criterion.
Now, it is a continuation of that particular lecture, we have represented the stress in terms of
shear stress and normal stress. Now there are various other ways of representing stresses for
different failure criterion. So today's, this particular lecture, we will see what are the different
representations of stresses for defining the failure of a material?
So the first one is about different stress representation, the first one is familiar to us which is τ
versus σ’ where σ’ is the effective normal stress. Now you need to note a specific point here,
I have started introducing dash or prime. Now this is very important because now, we have
started the actual discussion of strength and we know by effective stress principle strength is
completely governed by the effective stress.
156
So we have τ versus σ’ and this is a typical coulomb failure line or Mohr coulomb failure
line. We need not discuss about this anymore, we have already seen this. Another way of
stress representation is major principle stress versus minor principle stress. And we will get a
line here and that is the failure line within which the soil remains. The possible stress states of
the soil remains.
We also have discussed and we have also seen is the stress obliquity and we have also
discussed about maximum stress obliquity in the last lecture. So this is another way of
representing the stress or as a modification of this, this one, we can also represent in terms of
a more general term where it is axial stress versus radial stress and all normal stresses need to
carry this dash or prime indicating that it is effective.
So σ’a is axial stress and σ’r is radial stress. So you can also plot between σ’a verses σ’r and
again there is a failure line and the slope of the failure line is represented by kf. So in all these
representations the slope of the failure line is kf. So these are some different ways of stress
representation. Now to be very specific in terms of failure criterion, we also have some other
representations.
Which is q-p’ plot or it is known as Cambridge plot possibly due to the contribution towards
this from the research group at Cambridge. So we will see for specifically for triaxial
condition. In general we have already seen, what is q? What is p’? What is q? q is deviator
stress and p’ is mean stress for 3 dimensional state. We have discussed this in general in our
previous lectures.
Now here we will specifically confine our discussion to triaxial condition because that is
going to be a very prominent condition for soils. All our determinations in the lab confines to
mostly triaxial condition or in other cases it will be plane strain condition. Whatever be in
today's representation, we will confine to triaxial condition and q and p’ for triaxial condition
we have already evolved in the previous lecture.
Accordingly we know p’ that is the mean stress is
157
Now let me make it a point clear here. It is always better to write in general terms and when I
say general terms, I mean in terms of axial and radial stresses, why because whether σ’a that
is axial stress equal to major principle stress will depend upon certain condition, we will see
that in the subsequent slides.
So I would always suggest that we write p’ and q in terms of axial and radial. For the; time
being let us not worry; about the major, minor principle stress. So triaxial condition, what is
the triaxial condition? We have already seen I will not discuss that,
because order on stresses. So σ’2 = σ’3. So in the order on stress it is same, so 2 σ’r / 3 q = σ’a
- σ’r or it is σ’1 - σ’3 dash this may change depending upon conditions.
So that is why will confine to axial and radial. So here in compression you will see that σ’a or
the axial stress is equal to σ’1 and σ’r = σ’3. This is a typical triaxial condition that we deal
with in the lab. Now if you slap it that is σ’a = σ’3, σ’r = σ’1, it becomes a typical case of
extension, why? It gives more like a squeezing effect. So σ’1 is more, so it has a tendency to
extend so that is a typical case of extension and this is a typical case of compression.
So if you want to specifically define in terms of compression and extension then it is always
better that we define these stresses in terms of axial and radial that is the point I want to
make. So we will confine to that.
158
Now, let us see the slope of the failure line in q-p’plot. So what is the slope of τ versus σ
plot? It is nothing but φ’, so that is the slope. Similarly we want to find out what is the slope
of failure of q-p’plot. So this is q-p’plot and I have drawn 2 lines, these are failure lines and
one line corresponds to the case of compression, the other line corresponds to the case of
extension.
So depending upon whether it is compression or extension, we need to take appropriate

stresses and that we have seen in the previous slide. So kf is nothing but q/p’. Now we are
taking the case of compression. This means that σ’a = σ’1, σ’r = σ’3. Substitute for q and p’
that will give σ’a - σ’r / [(σ’a + 2 σ’r) / 3] always represent in terms of axial and radial.
So this is what it is, substitute for σa and σr that will give you this expression. Rearranging
you can take σ’3 outside, so it will become
159
Now, we have σ’1 = σ’3 tan2 (45 + φ’/ 2) this is as per the Mohr Coulomb failure envelope.
Now this is not going to change, this remains same. So we have what is the ratio? Now this is
for c’ = 0, we are not considering cohesion for the time being, so we will consider the Mohr
Coulomb failure envelope that will give σ’1 / σ’3 = tan2 (45 + φ’/ 2) = 1 + sin φ’/ 1 - sin φ’.
You substitute this in this particular expression of q / p’ that will give you (3*2 sin φ’)/ (3 -
sin φ’). So the failure envelope, the slope of the failure envelope will be q / p’ = 6 sin φ’/ (3 -
sin φ’). Remember this is for the case of compression. Now the same can be obtained for
extension as well that we will see later maybe as a part of assignment, for the time being this
is only for compression.
So what we have done? For a specific representation of stresses in terms of q and p’, we have
defined the failure line. Remember this is going to be useful throughout this course, why? We
need this in stress path; we also need this in critical state module as well. So please pay more
attention to this.
160
Now another representation of stress is t –s’ plot or it is called MIT plot or it is known as
Maximum Shear Stress plot. What is maximum shear stress? When you draw a Mohr circle,
we know that there is a maximum point that we obtain and we have seen that this maximum
point is not the stress which causes failure but a critical combination causes failure. So here
we will specifically talk about the maximum shear stress point.
If you draw a Mohr circle, I am talking about this particular point, this is the maximum shear
stress point and if you discuss it in terms of σ’a or σ’r, this point will be σ’a this is S, so σ’a +
σ’r / 2 and t which is this distance is t that will be σ’a - σ’r / 2. So again we have represented
in terms of axial and radial stress.
So here it is the typical representation of t versus s’. So t is σ’a - σ’r / 2 and s’ is σ’a + σ’r / 2
these 2 represents the maximum shear stress point. So here we are representing in this form
and there is compression line and extension line, we need to find out what is the slope of the
failure line in s –t’ plot. So s’ t remember s’ we have put prime, t we have not because this is
more like a shear stress, so u cancels off.
So you can, t is equal to t’, so s’ t is the locus of maximum shear stress point, we have already
discussed that. Now, let us consider t s’ plot, this is the Mohr circle we have represented
Mohr Coulomb failure line, we have represented. So that is the point C at which the tangent
goes. We also have this particular point through which you can draw a line. So if the Mohr
Coulomb failure envelope is given by a red line, from the same point we can also draw a line
which passes through the maximum shear stress point that is point B.
161
So, the; inclination of Mohr Coulomb failure line is φ’ and let us say the inclination of the
line passing through maximum shear stress point is α, so this inclination is α. What is this
point? That represents s’ and t point. Now this particular intercept is given by a’, the
inclination that is the line passing through maximum shear stress point makes an intercept of
a’ and we know the intercept of Mohr coulomb failure line is c’.
I have not marked it because we already know it, if I mark it; it becomes more cumbersome,
so I am not marking it. So this is how it is. We want to find out what is the inclination of this
particular line because we are talking about the maximum shear stress point. Accordingly we
can write a modified failure envelope which is different from Mohr Coulomb failure
envelope that is t, for Mohr Coulomb it is τ.
So here it is t = a’ + s’ tan α , α is the inclination. So it is more or less very identical to Mohr

Coulomb failure envelope. Only thing is the parameters changed. So what is tan α that is the
slope. So we want what is tan α, this one? So that tan α can be written as AB / AD, now that
is also equal to this smaller triangle that is a’ / DO. So this triangle and this triangle is what
we have considered.
So tan α is a slope what is needed for us. Now sin φ’ if you consider triangle DCA that is
triangle DCA, AC / DA is sin φ’, so that is what is written here. Now we also know AB and
AC are radius, so AB = AC that means, this expression and this expression are same. So, AC
= AB, hence we can write tan α which is the required inclination of the failure line of t- s’
plot = sin φ’ and φ’ we know it comes from the Mohr Coulomb failure envelope.
162
Now we also note tan φ’= c’ / DO that is tan φ’, considering the red line, the intercept is c’
divided by DO. If I write tan α / tan φ’ where tan α = a’/ DO, we can write a’ / c’. So tan α /
tan φ’= a’ / c’ from which we can write, how will we get this? So tan φ’, we have already
seen tan φ’= sin φ’, we have already proved. Substituting that here in this expression tan α =
sin φ’.
Here it will be sin φ’ that will cancel though and cos φ’ goes up. So it will be c’ = a’ / cos φ’.
So slope of the failure line of t- s’ plot, so this plot is tan α and that is equal to sin φ’. So what
we have done? We have defined the kf or slope of the failure line kf of t-s’ plot. Remember
when you do a triaxial testing, those who have not gone through this during the
undergraduate, I am just adding this detail here.
When you do the triaxial testing, what you will get is, you will get the deviator stress at
failure that is σ’1 - σ’3 at failure and you know the confining stresses σ’3. So you know what
is σ’1 and σ’3? σ’1 , σ’3 known. Now what will you do? You need to plot the Mohr circle and
get the shear strength parameters c’ and φ’ as applicable.
Now if you know this s’- t plot you will need not plot the Mohr circle. From the known
values of σ’1 and σ’3 one can always find out different points that is s’1 t1, s’2 t2, s’3 t3 and
plot this blue line, we will get maybe 3 points, corresponding to that we will get 3 points
here. In t s’ plot, one can always plot this blue line, once the blue line is plotted we can get a’
and α.
163
And once you know a’ and α, we also have the expression tan α = sin φ’. So α is known, tanα
is known, we can always determine φ’. Similarly we have this expression c’ = a’/ cos φ’, φ’is
known, a’ is known, c’ can be determined. So this, Mohr Coulomb failure parameters can be
obtained without plotting the Mohr circle once you know the concept of t s’ plot.
So that is all about that. So we will summarize this first lecture that is granular materials like
soils fails in shear. We have already told this we need to keep this in mind. Coulomb friction
model is the simplest model that represents shear strength of soil but remember Coulomb
friction model, it depends upon the development of a failure plane within the soil mass. Shear
strength parameters c’ and φ’, these are not fundamental properties of soil.
Now when I say this, there will be some for whom this is difficult to understand, why? It is
not a fundamental property because it is governed by the condition under which c’ and φ’ has
been obtained. You are not likely to get the same value all the time. But having said that,
there are cases where or there is a specific condition where it can become, it can become
means, mostly φ’ can become a fundamental parameter that we will see in the fourth module
in detail.
So for the time being, let us understand that c’ and φ’ are not fundamental properties of soil.
It could have been a fundamental property if the void ratio and or pore pressure remained
constant with normal stress. Now this is a difficult proposition to make in soils either the void
ratio because in drained condition there will be volume change and when there is volume
change, there is change in void ratio.
164
If it is undrained condition, there is no volume change only pore water pressure changes that
is why it is written void ratio and, or pore pressure. If these would have remained constant
with normal stress then c’ and φ’ could have been a fundamental property. You remember
when we discussed in module 1, we said that the volumetric component and the deviatory
component we decoupled.
Even though we said that in continuum framework during that lectures, it is very hard to
conceive soil without volume change during shearing which means to say deviatory
component also involves change in volume. So it is very difficult to decouple. So now,
subsequent lectures and all our further understanding, volume changed component is still
there during shearing. We are not actually able to decouple it, even though in the tonsorial
framework, it is decoupled.
This understanding one should have. So it should not lead to confusion. It has been said there
it is decoupled, why it is not decoupled? It is decoupled in that framework, mathematical
framework but then it is difficult to decouple to understand the behavior of soil because there
will be volume change. Now if the sharing would have happened at constant volume,
probably c’, φ’ components can be a fundamental parameter which we are not able to
achieve.
Either the volume change will happen or the co order pressure would change. So this prevents
the phi dash component to be slightly tricky and not a fundamental parameter. So this is what
I told, it is difficult to maintain this to be constant during loading. So shear strength
parameters are condition specific and it is very important to mention the condition. Now later
we will see that in under certain circumstances it becomes fundamental rather it will not
depend upon the condition.
So you underline this particular sentence here for the time being with our current
understanding, we say that it is not a fundamental parameter but under specific conditions it
can be treated as a fundamental parameter. So Coulomb model integrated with Mohr circle
gives the popular Mohr coulomb failure criterion for soils. There are different ways of stress
representation for failure criterion which has been discussed.
165
And the slope of the failure line, how to determine? That has been discussed for the important
ones. And books that may be useful for this particular module is Muniram Budhus book,
Holtz and Kovacs and Whitlow. These 2, I have already specified it in the reference but
Whitlows book is not there. It is R Whitlow, Basic Soil Mechanics, Longman in 1995. This is
also a good book to follow this. I am basically following these books for generating these
lectures.
So that is all for now, related to basics of shear strength we have splitted into 2 lectures, both
confines to basics of shear strength. In the next lecture, we will see the shear strength of
granular soil. Now what we have done is, we have splitted it into granular, granular means
basically cohesion less soil and then cohesive soil. So in the next lecture we will see, what is
the sheer strength of cohesion less soil or granular soil? Thank you.
166
Lecture - 19
Shear Strength Granular Soil - 1
Welcome all of you, in the last lecture we have discussed about basics of shear strength and
stress representation in different form. Now we will get into the shear strength of soils
specifically, while doing so, we will divide the soil into 2, the first one will be for
cohesionless soil and the other one is for cohesive soil. This is divided in this form because
the shear strength behavior or shear strain behavior of these 2 classes of soils are entirely
different.
While you know that for cohesive soil it is generally influenced by the drainage condition and
pore water pressure we will see that in detail whereas for cohesionless soil the drainage it is
almost draining soil, so the question of undrained condition does not come into effect except
for some very specific problems like liquefaction. So, we will now first get on to the shear
strength of granular soils.
Now we can see that all soils are granular in nature. What is so specific about it? But here
when we discuss about soil mechanics, we will refer granular soil to essentially cohesionless
soil which will include gravel sands or cohesionless silt size geomaterials because this will
include materials like fly ash which are essentially silt size and hence they can be considered
as geomaterials with less or no clay content.
167
So, here when we say granular soil it essentially refers to cohesionless soil. It does not
possess true cohesion that means, it is the absence of natural cementation. But some granular
soils like sand, it can exhibit what is known as apparent cohesion which is conditions
specific. It depends upon the water content or the degree of saturation of that particular sand.
You might have noticed that it is fairly comfortable walking on beaches which are partially
wet. Why this is so?
This is mainly because of the capillary action or the negative pore water pressure which we
call it as soil suction which is present in partially saturated sands. Hence it imparts more
strength and hence you are able to walk comfortably on beaches. But that is not the case with
dry sand or fully wet sand. So, this is known as the contribution or this additional strength
which the sand posses under partially saturated condition, this can be termed as apparent
cohesion.
So, we are not considering about apparent cohesion right now, we, what we understand is
cohesionless soil there is no true cohesion. So, since we understand that for cohesionless soil
there is not much influence of drainage and there is no cohesion coming into picture, the
understanding shear strength behavior is relatively simple as compared to that of cohesive
soil which we will see later.
So, we can conveniently write shear strength of granular soils, τf = σ’f tan φ’ where τf and σ’f
these are stresses acting on the failure plane which we have already seen in the previous
lecture and φ’ is the angle of internal friction. Please note that τf there is no prime, σ’f that is
for normal stress, it is always in terms of effective stress.
φ’ is mostly determined by either direct shear test or drained triaxial test because the
permeability of granular soil is fairly high. The initial state of the granular soil whether it is a
loose medium or dense state affects the shear strength. Now what is this initial state? The
initial state depends upon the initial unit weight or packing or the compaction state and is
generally defined by the term relative density which all of you knows.
So, as a guideline, a relative density less than 35% is considered to be loose, 35 to 65% is
medium and greater than 65% is considered to be dense. So, we are basically discussing
168
about 3 states of the granular soil which we define in terms of loose, medium and dense. So,
this relative density takes care of what is the kind of initial packing of the granular materials.
So, now we will see the typical stress strain response of sand which is a granular material
from a typical direct shear test wherein the x axis is represented by lateral shear displacement
and y axis is given by nominal shear stress, τ. Now we know in direct shear test, it is very
difficult to quantify the strain and the true stresses acting on the failure plane or which is
acting on a given plane.
The reason is the area of direct shear keeps changing during the test because there is a
horizontal movement of the upper box. So, it is very difficult to state the true stresses acting
and hence the term nominal shear stress, τ and instead of strain we express this in terms of
lateral shear displacement or it is written in terms of relative displacement, intentionally we
do not use the term strain there.
So, here there are 3 states we have discussed, we can see that the stress strain response is
drastically different for dense, medium and loose, at least dense and loose it is fairly different.
What is happening? We can see that for a dense soil or the dense state of the soil initially
there is an increase in shear stress reaches a maximum value and further which it starts
reducing. Now this post peak reduction is what is known as strain softening.
So, there is a maximum shear stress or peak shear stress in the case of dense packing.
Medium it reduces but for a loose state of packing, we can see that there is an increase in
169
stress but there is no reduction in stress. Somewhere at some point there will be yielding of
the loose state. So, even after yielding we can see that the stress keeps on increasing and all
of these finally reaches to a relatively constant value which is known as critical shear stress,
τcv or constant volume shear stress or critical shear stress.
So, whatever be a dense reaches peak strain softens and a loose keeps on densifying and
reaches finally a critical state. So, this is the general understanding of stress strain response of
sand or the same type of behavior is applicable for cohesive but in a different framework that
we will see when we discuss the shear strength of cohesive soil. Now what is happening is,
this is what a loose state represents.
That means soil particles are loose and when this loose state is subjected to shearing that is
the horizontal movement what will happen is when the loose state of the granular soil is
sheared, particle can move horizontally and when it moves horizontally in this direction, what
happens is, the particles moves downwards as you can see the arrow all these particles goes
and fill into the voids because it is a loose state and hence, it moves downwards.
This is very important aspect which we need to understand when a loose state is sheared
particle moves horizontally and downwards as it can shown by the arrow and hence it
undergo compression during shearing. This is an important point which we need to keep in
mind a loose state when it is sheared undergoes compression. This is very important
observation. In the case of dense state it is already in these voids.
So, there is a good interlocking of materials. Now let us say that this dense state is sheared.
Now for shearing means it has to move horizontally. Now what is the possibility of such a
movement? If it has to move horizontally then it has to undergo some movement, so that we
will see. So, the dense state of the granular soil is sheared, so it is sheared. Shearing is
restrained by interlocking because it is already occupying the spaces it is densely packed, so
there is a good interlocking.
Now during shearing, if it has to move horizontally then it has to move in the upward
direction, the particles have to move in the upward direction, this is different from the loose
soil. So, movements up the inclined plane then horizontally. So, if it has to make any sort of
170
horizontal movement, the movement of this particle is more like moving up the inclined
plane. So, it has to do additional work.
Now this causes expansion or what is known as dilation. Please remember this carefully like
when a dense state of soil is sheared for executing shear, it has to expand or it is known as
dilation and during dilation there are formations of loose pockets because a dense material
which is packed when it expands definitely it leads to the formation of loose pockets. So, this
is the change in behavior or this is the difference in behavior when a loose and dense state of
the material is sheared.
Now let us see loose state in detail. Soil particles moving horizontally can be idealized easily
by Coulomb’s friction model as what we have learnt. Because here it is loose state there is an
intergranular contact where τf, σ’f is acting, this is the predefined failure plane along with it
acts. We know the Coulomb's friction model is applicable only when there is a failure plane
attached to it, so we can easily write τf = σ’f tan φ’, no confusion in that.
So, let us see for a dense state, what is going to happen? Soil particles moving upwards
during shearing can be idealized by a block moving up the inclined plane. Now it is already
in a interlocked position, so it is more like this. So, the plane now is inclined at the point of
contact or tangent plane is inclined. So, same interpreted on an inclined plane upwards. So,
what is happening? You can be idealized by the block which is moving up the inclined plane.
171
So, you can see that there is an, this is a block here, the weight of the block acting downwards
R is the resultant, φ’ is the angle and here the slope angle is α. Now this is a very prominent
friction problem that we have learned and we can conveniently write τf = σ’f tan (φ’ + α)
because it is moving up the inclined plane. Now if it is down the inclined plane this will be
minus, in soils both will be there.
There can be upward and downward movements for particles but generally in the case of
dense soil for explaining the shear behavior, we normally refer to dilation or expansion. The
expansion or dilation in dense state increases the friction angle component, this is what it is.
So, it increases from φ’ to φ’+ α. The additional component α is called the dilation angle very
important.
Now imposing high normal stress, now if the normal stress is more, let us say, what happens
is, this particle will not be able to traverse in the upward direction, you are restricting the
movement from the upward there is a confinement. Now this confinement will not allow the
particles to move upwards freely which means to say that the amount of dilation that is going
to happen is a function of the normal stress. Now in this case normal stress is more like a
confining stress from above. So, imposing high normal stress suppress the tendency to dilate
because of which the dilation angle decreases.
Now we will see this again. There are some facts which we will enlist where dense and
medium compacted granular soil exhibit peak shear strength specifically at low confining
stress because the dilation has to happen then only the dilated behavior expansion followed
172
by the peak behavior is going to happen. So, this peak which is marked here is a function of
this dilation behavior. So, peak behavior occurs at low strain or relative displacement due to
particle interlocking and dilation.
And it is mostly a stiff response which means particles are tightly held so, it is a stiff material.
Now the peak happens at a low strain. So, it keeps on increasing, reaches a peak but this peak
behavior is exhibited at relatively lower strain not at the highest strain or it is expressed in
terms of relative displacement whatever and this happens because of the dilation or particle
interlocking. On further shearing after the peak, stress comes down to critical state or τcs or
τcv or it is known as the ultimate state.
We have already told us this particular movement downwards not movement this stress
comes down. So, this post peak reduction is called strain softening behavior. Now critical
state which is the state here or constant shear strength occurs at large strain, generally it is 10
to 20% relative displacement. Now this is again defined in different codes, it is generally 10
to 20% of strain levels you can expect critical shear strength.
Loose state does not exhibit peak and directly approaches τcv as can be seen here directly
approaches τcv. The progressive increase in stress after yielding, post yielding is called strain
hardening behavior which is denoted here.
Now that is about the stress strain behavior whatever we have discussed, let us see what
happens to the volume? Because we do not generally discuss about undrained or no volume
173
condition for granular soil there will be volume change and how it happens for dense and
loose state, we will see that. The volume of dense sand decreases initially and then dilates.
So, here on the y axis, it is vertical displacement and on the x axis it is lateral shear
displacement or relative displacement.
Now for the dense soil, we can see that there is a initial decrement in vertical displacement
which is nothing but compression. So, below is compression, no volume change and
expansion. For dense soil there is initial compression followed by an increase in volume
which we call it as dilation. So, by the time it dilates, it reaches the peak behavior. Volume
reaches constant at high strain, relative displacement. So, by the time it reaches the critical
state, the volume change also remains constant there is no more volume change.
It reaches a constant volume or a constant value at high strain condition, maybe close to
critical state. Loose sand undergoes progressive compression or it is progressive densification
till it attains critical, constant volume or critical state volume. That is also at a high strain.
Indirect shear test, the area keeps changing. Hence true stresses and strain are difficult to
determine which I have already explained. Shear stress developed on shearing surface is non
uniform for direct shear test because at the edges and towards the center it will not be same.
Now, for dense soil, the total friction angle can be written as φ’peak = φ’μ + α, we have already
discussed what is α. Now peculiarity about α is, it increases with an increase in relative
density we have seen that. Depending upon what is the relative density the dilation keeps
174
changing, so that will be reflected in α. So, alpha becomes a function of initial state or
denseness. If it is more dense α will be more.
Now φ’μ this component is due to the friction mobilized at intergranular contact there is the
particle interaction and this is somewhat a fundamental aspects of a soil, φ’μ which is
mobilized at the particle at contact. α is the effect of interlocking, dilation and hence a
function of normal stress σ’f. So, in this total friction angle of dense soil, φ’μ is a function of
particle interaction, α is a function of dilation.
Now, since α is a function of dilation we have already seen that α is a function of normal
stress, if normal stress is more it suppresses the dilation, so we will see the effect of σ’f on α
and Coulomb’s failure line, how it changes the Coulomb’s failure line? This τ versus σ which
we normally represent the shear strength behavior for a cohesion less material, this columns
failure line without dilation. So, this is φ’, so it is very easy to understand OP is the failure
line represented by Coulomb's model.
As σ’f increases, the dilation is suppressed. So, when corresponding to σ1f, let us say, it is a
dense soil, so it will have the dilation angle of α1 corresponds to a τ peak 1. So that is the first
state. Now let us say that for a, as the normal stress increases, for example, at σ2f, what is
happening? We can see that the dilation angle gets reduced. Now this is the φ’ now that is
what is denoted by φ’μ.
Now what is above that? Above φ’ is α1 that is getting added. So, this is α1, at lower normal
stress α1 is more but as σ2f it increases this α2 is less than α1 which is denoted by point R. So,
this results in a curved Coulomb's failure line which is due to dilation. If you joins through
these points then you will get a curved line. α reduces with σ’f and hence results in a curved
failure line which is represented by OQRS.
And at some σ’f that is along this x axis there is a complete suppression of dilation. So, if you
keep on increasing, so there will be another point here at some point here, there is a complete
suppression of dilation and α tends to 0 and the failure line it merges with the initial failure
line where Coulomb’s failure line without dilation is there. So, α becomes equal to 0 at this
particular point α= 0 and φ’peak = φ’μ according to this equation. So, α is 0, so φ’peak = φ’μ.
175
So, dilation angle α is the measure of change in volumetric strain with respect to change in
shear strain. And φ’μ can be termed as friction angle corresponding to critical state or
constant volume. First we have tried to understand what is φ’μ? Now what is the significance
of this φ’μ because this is the fundamental aspect or the fundamental friction behavior of the
soil. Whatever be the state, there will be some kind of friction angle which is the critical state
friction angle.
We will be able to appreciate this better when we finish the fourth module on critical state for
the timing, let us understand that there is always some value which is which remains and
which can be mobilized by a soil that is a critical state and φ’μ is termed as the friction angle
corresponding to critical state or constant volume. So, φ’cs = φ’μ. Now, whether dilation
happens or not that will determine α or the additional component that gets added to the soil.
So, shear strength corresponding to peak behavior is called peak shear strength that is τpeak
and it is expressed as τpeak = σ’f tan (φ’+ α). Strength corresponding to critical state is called
critical shear strength τcs which is given by, τcs = σ’f tan φ’cs. Now remember this is critical
strength, friction angle and φ’peak = φ’cs + α.
Earlier we have written in terms of φ’μ that is for our initial understanding. But where it boils
down to? It boils down to φ’cs. So, Coulomb’s failure line without dilation and here this
friction angle which is the friction angle which is mobilized by soil at any streak that is φ’cs.
On top of it we get α which is φ’peak. α generally ranges between 0 and 15 degrees.
176
φ’cs is the effective friction angle at critical stage shear strength, a fundamental soil
parameter. This aspect of fundamental soil behavior can be better appreciated again when we
discuss this aspect in module 4. For a plane strain case, Bolton in 1986 showed that φ’peak =
φ’cs + 0.8 α. Here we have taken plus α but there is not much of a difference.
But specifically for plane strain condition, Bolton explained this to find out what is, how the
dilation angle contributes to φ’peak. So that is all for now. We will see in the next lecture, the
remaining portions of shear strength of granular soil were in 3 aspects we will be discussing,
one is the critical void ratio then we will discuss about Taylors model and there is something
called intrinsic friction. So, these 3 aspects will be covered in the next lecture. Thank you.
177
Lecture - 20
Shear Strength Granular Soil - II
Welcome back. So, this is the continuation of the previous lecture, shear strength of granular
soils. In this we will see first what is the concept of intrinsic friction? We have already
discussed about dilation and how it gets suppressed at high normal stress, so something in
continuation with that.
This is what is known as the concept of intrinsic friction. At high normal stress particles have
less freedom of movement, we have seen that the particles are tightly packed and hence it has
less freedom of movement. Now friction is mobilized only when particles have freedom of
movement. In order to mobilize friction there should be movement, if it is tightly packed then
the surface forces activates. So, we need some sort of movement for friction to mobilize. So,
the Coulomb's model may not be applicable at high normal stress.
A kind of apparent cementation develops and the friction reduces considerably due to the
restricted movement. So, Skempton in 1961 introduced the concept of intrinsic friction, let us
see what is intrinsic friction. So, this is what it is, as the normal stress increases, it restricts
the freedom of movement and hence the mobilized friction drastically reduces which is
indicated by Ψ. So, now, it is considered different from the Coulomb's model wherein we
discussed in terms of Ψ’.
178
So, at a high normal stress, there is some deviation in its behavior and there is a kind of
apparent cohesion we cannot call it as real cohesion, a kind of apparent cementation which is
represented by the term k. So, this basically is k and the angle is Ψ which is considerably less
than φ’. So, this equation τf = k + σf tan Ψ where Ψ is called the intrinsic friction angle. And
for quartz k is found to be close to 9 x 105 kilo Pascal and Ψ = 13 degrees as per Skempton,
1961.
So that is all about intrinsic friction, just to add to the discussion of dilation and the
suppression of dilation. Next we will see Taylors model. Taylor in 1948 proposed an energy
model to define the shear strength of soil. Now this model is very much similar to that of
Coulomb's model but marginally different in some assumptions. We will see what actually it
is? Assumed that shear strength of soil is due to sliding friction and interlocking of soil
particles, in Coulomb’s model also it is due to sliding friction.
But the sliding happens along a failure plane whereas, in the case of Taylors model, there is
no assumption of failure plane but the shear strength has 2 contributions, one is due to the
sliding friction of the soil particles in movement and the other one is due to interlocking of
soil particles. When I say particles in movement it is on a micro scale. So, it is not actually
moving there is a sort of relative movement because obviously the friction gets mobilized.
So, what has been done in Taylor's model is? It has splitted the 2 effects whereas, in the case
of Coulomb’s model, this was embedded in the friction angle itself, it was an additive
179
component. Now simple shear means shearing at constant volume. So, Taylors model
actually refers to simple shear. Now what do you mean by simple shear? Simple shear means
those shearing that happens at constant volume.
You remember when we discussed the first module, we said that the effects are decoupled.
We are studying a material by decoupling the volume change and shearing or the deviatory
component separately. But this will not work all the time fine for soils because we need to
understand what is the sort of volume change that happens during shearing, otherwise it is not
meaningful. So, there is a sort of deviation from whatever we have learnt initially in module
1.
We need to discuss about pore pressure or volume change during shearing and this is very
specific for cohesive soils. So, it is important to understand the volume change during
shearing for soils. So, there is a sort of deviation from our simple shear concept, the soil
element is subject to shear stress τ under constant vertical stress σ’z and this is the soil
element it is subjected to τ which is the shear stress and σ’z which is the constant confining
stress or the constant confining vertical effective stress σ’z.
And there is a kind of rotation and it is not actually rotation there is a displacement y Δx in
the x direction and the compression expansion which is denoted by Δz in the z direction. So
that is the expansion compression during shearing. And volumetric strain εz can be written as
Δz / H. This H is the initial height and this is close to a plane strain condition where εx = εy =
0.
So, shear strain γzx = Δx / H. So, this is the concept of a linear strain and the shear strain is
discussed in module 1, so that it is easy for you to understand. So, Δx / H is the shear strain,
so incremental shear strain is dσ and dεz.
180
According to Taylors model external energy because it is a energy based concept is stress
into compatible strain. Now what causes shear that is τ*dγ. So, this is the compatible strain or
the shear strain. An internal energy is essentially associated with work done by friction
specifically sliding friction and work done by soil movement against the vertical effective
stress. So, we can write μ *σ’z becomes the shear or the sliding friction into dγ.
So, this becomes the shear strain plus σ’z * dεz that is σ’z is the vertical stress dεz is the
vertical strain. So, for equilibrium we can equate internal and external energy as shown here.
Now dividing by σ’z dγ we will get τ / σ’z = μ + dεz / dγ or shear strain. Now we know that at
critical state μ becomes equal to tan φ’cs.
That is, there is no dilation, whole of the sliding friction component is associated with the
critical state friction angle and dilation angle tanα= 0. But for a peak shear strength condition
tanα = dεz / dγ or vertical strain to shear strain. So, Taylors model is then the summation of
these 2 components. So, τ / σ’z which is written here is equal to tan φ’cs which is the
component of critical state or the fundamental aspect plus the dilation aspect. Now what has
happened? We have splitted these 2 instead of φ’cs + α in the Coulomb's model.
181
So, Taylor model does not need the assumption of plane of failure like Coulomb's model. It
can be applied at every stage of loading because the relative aspect of dilation and the critical,
the basic friction angle is taken care of by the summation. Whereas, in the case of Coulomb's
model it is within the friction angle together, the total friction angle. So, it is applicable for
homogeneous soil which deform under plane strain condition or a simple shear. The model is
not applicable for soils with predefined failure plane like joints or interface between 2 soils.
So, here there is no consideration of plain of failure.
So, now, we will come to the next concept what is known as critical void ratio. It is the void
ratio at critical state shear strength and it is called critical void ratio CVR or ec. Now what is
this critical void ratio? This concept was introduced by Casagrande in 1938 to study
liquefaction of soil. What it states is that when the soil mass with an initial void ratio less
182
than ec, ec is the critical void ratio that is a dense state when it is sheared, it dilates and attains
ec close to failure.
So, when a dense state of the soil if it is sheared then the void ratio will increase because of
the dilation and finally it reaches what is known as a constant volume or a constant void ratio
at close to critical state. So, it is called critical void ratio or constant volume void ratio.
So, this is what it is, this void ratio on y axis relative displacement or strain on x axis, this is
the point. So, initial void ratio is less than ec and this is a dense state of the sand or soil and
when it is sheared it dilates and it reaches a constant void ratio which is known as critical
void ratio at this point. Similarly when a soil mass with initial e which is greater than ec that
is for a loose state when it is sheared it densify and attains easy close to failure, failure means
here we refer to critical state or critical void ratio where the volume becomes constant.
Once the void ratio attains ec during shearing, it remains constant. So, for all practical
purpose critical void ratio remains constant, once it reaches that particular state. ec may not be
achieved for the entire soil but it may be for the shearing zone which means the entire soil
may not reach the state of ec basically where the failure zone initiates that particular zone ec
will be achieved.
183
So, Casagrande noted that ec is related to effective confining pressure σ’c. So, depending
upon the confining pressure, ec will change, why? The amount of dilation effect also changes
depending upon the confining pressure. So, accordingly whatever ec it achieves that also
changes. The locus of ec as a function of effective confining pressure σ’c is called critical void
ratio line, CVR line, as can be seen here. It is e versus σ’c.
Now this blue line represents the locus of critical void ratio, we can see that as σ’c increases,
ec reduces. Now this blue line is specifically ec, variation with σ’c. CVR line, this particular
line marks the boundary between loose state and dense state. Let us consider a drained test
now, we are discussing this in general. For a drained test a loose sample, this is the loose
sample, densify that is e reduces for a drained test. What is meant by drained test?
There is no pore water pressure and hence the volume changes. For a loose state it reduces till
it reaches the critical void ratio and dense soil, this is the dense soil, as the shearing happens,
it increases in a drained test because drained test volume change happens. So, it increases and
reaches the critical void ratio. Now for an undrained test for the same case, let us say the case
of loose state. What happens in the undrained test?
In the undrained test, there will be development of positive pore water pressure. We will be
discussing this in detail in the next lecture on shear strength of cohesive soil. So, for the time
being, we will understand that for undrained test there will be development of pore water
pressure. Now in loose material, it will be positive pore water pressure. Now because of
184
positive pore water pressure, the confining pressure that is, σc that keeps on reducing because
the effective confining pressure keeps reducing.
So, σ’c keeps reducing till it reaches the critical void ratio line or till it achieves critical void
ratio, similarly in the case of dense state, we can see that as shearing happens under
undrained condition, there will be negative pore water pressure and till it approaches the
critical void ratio at close to failure. Now what is happening here? In the case of dense
material which has a tendency to dilate the pore water pressure development will be
essentially negative, keeps on reducing, so it becomes negative.
Now because of this negative pore water pressure, σ’c that is the effective confining pressure
keeps increasing, this we know effective stress keeps increasing for negative pore water
pressure till it reaches the constant void ratio line.
So, CVR line describe the state towards which soil specimen would change at large strain.
Now by the time it reaches the critical void ratio line, it would have undergone large strain.
So, we are talking about a critical state of that particular material which mostly happens at 10
to 20% strain, so that is what it means. So, CVR line is the line towards which soil state
moves during shearing and finally it reaches that particular constant volume and then remains
constant.
So, CVR line it is the locus of the line which towards which the soil specimen would change
at large strain, by volume change in drain condition which we have seen, by effective
confining pressure in undrained condition which is the manifestation of pore water pressure
change, this is decreasing or this is increasing and it may be a combination of both in a
partially drained specimen where there will be an interplay of both pore water pressure and
volume change. So that is the concept of critical void ratio.
185
Now we will summarize the shear strength of granular soil where the dilation characteristics
of soil is governed by both initial state of compaction and normal stress which is the
confinement that we have discussed. Higher the relative density, I mean the initial
compaction state, higher is the dilation because interlocking is more and hence the dilation
will be more. Higher the normal stress, lower is the dilation because there is a kind of
suppression that happens.
A complete suppression of dilation happens at high normal stress. At this condition, friction
angle is equal to critical friction angle. Peak shear strength, τpeak, is only applicable for
dilating soil. The shear strength of all soils at larger relative displacement or strain mostly
greater than 10% is critical shear strength τcs. So, this is the kind of minimum guarantee shear
strength for a given material.
So, whatever be the state we will see that this soil of a particular type possess critical shear
strength which is τcs. At critical state dilation 0, dilation angle is 0 and hence τpeak = τcs. φ'cs is
a more like a fundamental soil parameter whereas, phi dash peak is not a fundamental
parameter because φ'peak depends on whether dilation happens or not. Whereas, φ'cs is not a
function of that.
186
Coulomb's equation gives shear strength when slip is initiated along a plane within soils. So,
there is an inherent assumption of plane of failure developed within the soil for Coulomb's
equation to be valid. It does not give any information on the strain at which that slip or failure
happens. We have discussed this aspect before also that is, all these models does not talk
about the strain at which the slip or failure happens.
The friction angle obtained from Coulomb's model relates to sliding friction. This is true for
Taylors model as well. And Taylors model is applicable for homogeneous soil with plane
strain condition and does not need any assumption of failure plane within the soil mass. The
void ratio at critical state is called critical void ratio that is the constant void ratio which a soil
mass attains in towards failure when it is sheared.
A loose and dense sand approach the state of CVR during shearing. So, CVR is used to
define liquefaction susceptibility of granular soil. So that is all about shear strength of
granular soils. In the next lecture we will see shear strength of cohesive soil which is even
more complex due to the fact that, it is highly influenced by pore water pressure and drainage
condition, so that we will see in the next lecture. That is all for now. Thank you.
187
Shear strength of cohesive soil
Indian Institute of Technology, Guwathi, Assam
Lecture - 21
Shear Strength of Cohesive Soil
Welcome back in the last lecture, we have seen shear strength of granular soils. So, in this
lecture we will see, we will begin with shear strength of cohesive soils. Now, you may be
wondering why such a distinction in fact, the response that you obtain when I say response it
may be shear strain response or the manner in which you interpret shear strength the output
may be similar for these 2 soils.
But the conditions leading to that output or the initial conditions of soil may be different and
the factors contributing to that final output may be different in both. So, in this lecture, we
will see shear strength of cohesive soil.
Now, let us start with this introductory remark that shear strength of cohesive soil is highly
complex as compared to cohesionless soils. Now, is it really true or not? We will have to see
as we proceed, I would also like to add the statements from a popular book by Taylor
fundamentals of soil mechanics were in his states. In fact, no physical property of cohesive
soil is more complex than this shearing strength.
This property depends on many factors and the individual factors are themselves
complicated. But, in addition, they are interrelated to such a degree that it is extremely
188
difficult to understand their combined action. What more we need to know to understand the
complexity of shear strength of cohesive soils. You remember I told this sentence in the very
beginning of this module, when we introduce this module.
That interpretation of shear strength of soil to some extent lies with the reader himself. So, I
am reading certain concepts, I have my own perception of conceiving how or what a shear
strength is all about. So, there will be some differences that is mainly because of the
complexity associated with so, for cohesive soil, the complexities are more because there are
more factors which governs its final shear strength.
Also associating the relevance of shear strength determined the lab for a real life scenario is a
complex problem. Now, where is this complexity coming from, we will see what are the
factors that governs the shear strength of cohesive soil. Before going to the factors, we need
to understand the sentence like we try to simulate the actual field conditions in the lab but
many times we may not be able to achieve due to some sort of oversightness.
Hence, how much our laboratory test results represent the actual field condition. There is
something difficult to tell and you will be able to understand a bit of it as we proceed in the
lecture because we normally perform to axial compression tests. Now, whether this particular
test is suitable for all the conditions in the field, we will see this as we proceed in the lecture.
Also, in the case of cohesionless soil, we have seen that the drainage is instant. There are very
very few cases where pore water pressure becomes an issue but that is not the case with
cohesive soil. So, let us move on to the factors affecting the shear strength of cohesive soil.
So, the first factor is drainage condition as I have told and the controlling or maybe the
inherent drainage property results in pore water pressure.
If the drainage condition is not good, then it would result in pore water pressure. Now, this
dynamics itself is a big issue and that will depend upon the soil flow properties. The second is
stress history. Now, if you consider the shear strength of undisturbed soils, then stress history
becomes its initial state like what the soil has been subjected to in the past. I hope we all
agree to the fact that soil carries certain memory along with which gets erased and rewritten
every time when we load it.
189
This aspect also will become clear as we proceed. So, stress history becomes very important
loading condition and rate of loading well that does not need any explanation initial
condition, initial compaction state for example, what is the water content or the dry unit
weight specifically in case of remoulded samples, the type of test this also we have discussed
a bit like depending upon the type of test the shear strength characteristics also changes and
that is the reason why we need to state categorically what type of test we have adopted.
But that is on a general terms as you proceed again, by the time you finish the fourth module,
you will understand that there is a sort of fundamental characteristics of soil shear strength
characteristics of soil which you can very well capitalized and use it as a fundamental
property. Now, just like direct shear tests that we used for explaining shear strength of
cohesionless soil we will use typical triaxial conditions are to actual testing for cohesive soil.
Because you need control of 2, 3 factors specifically drainage border pressure measurement,
so on because these are important factors as you could see.
So, let us first discuss a bit on drainage condition and volume change some of these aspects
will be quite explicit and some of you would have understood this very well during your
undergraduate. Now, this refreshing of these portions is only for those who have not
understood it well and there are certain editions also which you may find in these
explanations which is important for our subsequent modules.
So, difference between static and excess pore water pressure, why soil mechanics is different
from other mechanics, it is because of the presence of voids and the relative content of water
190
and air in it. Now, we are talking about only 2 phase system saturated systems. So, voids are
filled with water. Now, when there is soil on its own, there is no external loading acting on it
whatever is the pore of pressure, the hydrostatic pressure we call it as static pore water
pressure.
So, this is what it is, we have a soil in place which has saturated soil Z is the height. So, u
static or the static port a pressure is Z into γw understands this is an equilibrium condition and
pore water pressure is merely under the self-weight of the soil, there is no external loading
acting on it and this is a typical equilibrium condition that means flow has taken place.
Soil is not changing and whatever is the condition of the soil and the water within this is
under equilibrium condition. Now to this equilibrium condition, let us say that there is a load
which is acting Δσ, the external load or increase in total stress. Now when we say total stress,
we understand this is some additional or extra load acting on the soil. So, external pressure,
Δσ applied on the soil mass, let us say triaxial loading.
Now at time t = 0 there is no drainage that can happen because this is first of all cohesive soil.
So, at time t = 0 there is no drainage now, what is the result of no drainage. So, there is a built
up of pore water pressure and there is a disequilibrium condition. So, Δσ is entirely taken up
by the water. So, pore water pressure increases by the same amount of Δσ at time t = 0.
So, the pore pressure from loading is called excess pore water pressure you excess now, this
excess pore water pressure makes all the difference in soil behaviour with respect to its stress
strain as well as shear strength behaviour. So, we need to understand the difference between
what sort of pore water pressure causes changes in soil and what sort of pore water pressure
is going to influence the shear strength behaviour or the mechanical behaviour of the soil.
It is essentially the excess pore water pressure and u excess to start with when time t = 0 = Δσ
and total pore water pressure will be Z into γw that is static component + this additional
component. So, u excess what happens to when something is not in equilibrium. Nature if
you observe if something is not in equilibrium, it always tried to achieve equilibrium by some
means, in the same manner.
191
Because there is disequilibrium under applied load condition u excess will not exist as it is. It
is a transient quantity and there is always a tendency for the soil to dissipate this and reach to
a new equilibrium condition. So, u excess will get dissipated over a period of time. Now, the
important question is how this will get dissipated. In what manner it will get dissipate this
will depend upon again the soil property.
Which is the hydraulic behaviour of soil property, hydraulic conductivity, how fast the water
can move out of the soil also, what are the drainage conditions which are possible whether
there is a pathway for water to move out all these factors determine how you access is going
to dissipate. So that is why it is written it is it will depend upon the loading rate and drainage
condition.
Drainage condition hydraulic conductivity of the soil on one hand and on the other hand, the
loading rate, the manner in which the load is applied, let us say it is an earthquake loading;
the whole of the loading is going to impose in this very small amount of time. So, the built up
of pore water pressure will be quite high. Whereas, if it is a very staged loading slowly, then
pore water pressure may even not develop, because it will have sufficient time to dissipate.
So, this is the 2 different extremes which we need to keep in mind and this is going to
influence the mechanical behaviour of cohesive soil.
So, now, let us start with undrained condition. Now, what is meant by undrained condition,
the hydraulic conductivity of cohesive soil is relatively low as compared to cohesionless soil
and it is soil and compaction conditions specific it will depend upon what is the type of soil a
192
clay soil will be having extremely low hydraulic conductivity as compared to a silty one.
Similarly, for the same soil the compaction state may alter the pore structure.
So, all this is going to govern what kind of hydraulic conductivity it has depending on the
loading rate and hydraulic conductivity of the saturated soil dissipation of u excess may not
happen adequately and instantly like cohesionless soil which means to say u excess is going
to remain there. Because of this stiffness water attracts the load and it holds it. Now, only
when the water is allowed to move out to expend the energy then only it will come back to its
original state.
And when it comes back to when it tried to achieve its new equilibrium state. There will be
some changes that will happen to soil and that is how we say the total stress is transferred to
the soil as in the granular stress or effective stress. Now, depending upon fast loading rate and
low hydraulic conductivity, also the access for the water to move out these 3 conditions can
lead to undrained conditions in soil.
So, any volume change now, the question associated with undrained condition is yes we
understand what is not allowed to move out u excess will remain there. So, what, what is the
net effect of undrained kind of when we say it is an untrained condition. What should we
understand from it now, let us say from volume change perspective of the soil now, for
during undrained condition.
We know that water is not going to move out so, fast then any volume change other than
movement of water will be associated with the volume change due to compressibility of the
water and or compressibility of the soil solids. So, under undrained condition the stiffness of
water is high and it attracts the external load. Since it is undrained, it is not able to move out.
So, naturally the stiffness of the water will be high and it has the tendency to attract all the
load to itself at the same time, stiffness of soil solids are also high.
So, what does this mean water is not allowed to move out both water and soil compressibility
there is no response or there is no volume change associated with well compressibility of
water and compressibility of soil solids which means to say that volumetric strain will be 0
and volume change will be 0 for undrained condition and Poisson’s ratio is equal to 0.5.
193
Now, when whenever we hear the term undrained condition, the first thought that should
come to us is it is going to be a no volume change condition.
And we have already seen this in the previous lecture, when the volumetric strain is 0; the
poisons ratio goes turns out to be 0.5. So, undrained condition, no volume change Poisson’s
ratio equal to 0.5. So, these are some of the facts which we need to keep in mind why I am
repeating this, why we need to understand this, you will understand its significance as we
move forward in the lecture, especially in module 3 and module 4. So, basically we are laying
foundation for a better understanding to deal with module 3 and module 4. So, pore water
pressure changes with loading. So that we have already discussed.
Now, some thoughts about drained condition because these facts are important before we
actually start understanding the shear strength behaviour of cohesive soil. So, if the hydraulic
conductivity of soil is high then drained condition exist with loading or if there is proper
access or if there is proper venting or pathway for the water to move out of the cohesive soil.
If the loading rate is very low and the hydraulic conductivity is fairly good or high then the u
excess whatever got developed may get dissipated fast.
Now, such a condition is called drained condition where drainage of pore water is possible
the loading rate is less than there is enough time for the dissipation of u excess even for a low
hydraulic conductivity case. Now, these conditions can lead to drain condition in soils. As the
u excess is dissipate, what happens as the u excess is dissipate. Now, this is another important
194
aspect which we need to understand clearly as the excess water pressure dissipate. What is
the net result?
When the water moves out of the voids the voids will try to come closer the packing is better.
So, whatever stress which the water has carried before dissipating that will get transferred to
the soil solids. Along its point of contact and that is what is intergranular stress any loss
importer pressure is the gain in intergranular or effective stress. So, it comes closer by
reducing the void space. Now, this is very important aspect.
Such a process is time dependent. Now that adds complexity to the behaviour of cohesive
soil. Whereas in the case of cohesionless soil, it is an instant response and drainage is also
more or less instant. So, you do not have to wait to see its final response. But that is not the
case with cohesive soil. It is time dependent and what time in what manner we do not have
any clue at the beginning. So, it depends entirely on the soil characteristics.
This causes consolidation which we know consolidation leads to progressive densification

and this is known as yielding which is nothing but plastic adjustments or readjustments. You
can call it in an initially loose saturated soil deposit, I have added this specifically loose
saturated soil deposit because only loose soil will densify so that is very important. So, you
need to understand that it undergoes consolidation; consolidation is a process of densification
for loose samples.
And when it happens, there is an elasto plastic response there will be a kind of yielding that
will happen to the soil and plastic adjustment of particles. Now, densification is applicable
only if the u excess is positive yes only positive pressure leads to densification if it is not
what is the other case u excess can be negative, we will see that later. But densification is
possible when it is loose, loose soil pore water pressure excess pore water pressure will be
positive and that dissipation causes densification.
For drain condition, pore water pressure is 0 and volume change occurs. So, now, this is
another important fact which we need to keep in mind when the condition is assumed to be
drained, we understand that there is no pore water pressure development and volume change
happens when we know it is undrained pore water pressure develops volume change is 0. So,
195
this are the 2 important aspects which we need to keep in mind and this influences the soil
behaviour.
But in actual field condition, this will be a combination like we cannot exactly have a clear
drain condition or very specific undrained condition. This aspect we are talking with respect
to the laboratory test we are conducting because we have proper control of drainage but in
field which will not be and there will be a combination of both conditions. Now that is where
we started off saying that a laboratory tests how good it translates to actual field condition.
So, volume change characteristics needs some more discussion let us see, volume change in
cohesive soil is a time dependent process because dissipation of u excess is governed by
drainage outlet and hydraulic conductivity. Volume change, characteristics of cohesive soil is
represented in terms of void ratio and log of consolidation pressure which is the net result of
consolidation test.
So, this is what it is there is a response between void ratio e and log sigma dash which is the
consolidation pressure is the idealized consolidation behaviour. So, volume change
characteristics can be described by one 1-D consolidation this is 1 dimensional consolidation
result or it can be due to isotopic consolidation which we can carry out in a triaxial testing
under the consolidation that happens under all round stress condition.
In both the cases shear stress development is negligible. Hence we are talking only about
volume change behaviour in both the cases because of infinite rigidity or in the case of
196
isotropic consolidation all round specifics are same, there is no possibility of development of
shear stress. So, we are basically understanding volume change behaviour specifically the
way we discussed in module 1 like no shearing only volume change
For soil consolidation stresses are always in terms of effective normal stress, second
statement of effective stress principle which says that all compressibility behaviour is entirely
governed by the effective stress. Let us say a freshly deposited soil is subjected to
consolidation as shown.
So, in this line AH is known as virgin consolidation line or normal consolidation line, NCL
and it have got a slope of Cc which is which is known as compressibility index. Now, why
this is called virgin consolidation line I hope you would have understood during your UG
because any stress on this AH it is for the first time the soil is experiencing that is the reason
why it is called virgin consolidation line and very popularly known as normal consolidation
line.
So, this whole response, the consolidation response indicates this status of stress history.
Now, let us say that you have a soil from the field, it is subjected to the stress of
consolidation, how it gives its response will depend upon what is the stress history it has been
subjected to. Now, history means, what is the kind of memory which the soil holds whether it
has been subjected to a higher stress or a lower stress, this we will understand from this
response.
197
Normal consolidation indicates that soil has not been subjected to this stress in the past we
have already seen that the volume change behaviour is steep and considerable plastic
rearrangement happens for a so, now, this is another important fact which we have to keep in
mind and normally consolidated soil undergoes high volume change, because it is
experiencing that stress for the first time.
When it is overloaded before it will have better stiffness and this is very important
philosophy when you experience more stress in the past it makes you more strong let us the
same philosophy that the soil also exhibit. So, when it is experiencing for the first time it will
have less resilience to it and that is why it undergoes a very high plastic rearrangement.
Densification happens for every incremental consolidation stress along AH so, we are talking
only about the NCL for NC soil volume change will be high the response AH is elasto-
plastic.
So, this much we need to keep in mind. So, a normally consolidated soil undergoes elasto-
plastic response. Now, let us say that at B the soil is unloaded and it is unloaded in steps BC
is the unloading line and CD is the reloading line. So, it is an unloading reloading line the
slope of this unloading reloading line or it is called recompression line it is mild and it is
represented by Cr.
Which is much less than Cc now, the gain in void ratio now, whatever wide ratio it gains, it is
also less because this much of densification has already happened. Now, there will be some
amount which it gains and that is the elastic component which we have already seen in
198
module 1. Cr is less than Cc now, URL BC this particular stretch of response this represents
the over consolidated state of soil.
Now, why it is over consolidated at this particular point it has been unloaded. Now, if you
consider any stress state along this line BC, this stress state is less than what has been
subjected to in the past that is σ’B is the stress which it has been subjected now, because of
this BC is more stiff till it reaches σ’B because it has already subjected to in the past to σ’B
and it knows the soil knows that I can withstand σ’B.
So, up to σ’B any stress which is less than that, soil can take it without any issue. So, it
exhibits a kind of elastic response. So, void ratio changes are less so, if you load along CB
the change in void it is not going to step down steeply like this. So, it is not going to go if you
start loading from C it will not take this route, it will still go along CB. So, why? Because it
has already reached a stiff state and hence it will exhibit an elastic response.
So, void changes are very less. So, consolidation stress at B σ’B is then known as the
preconsolidation pressure for the soil state along BC and this we have seen during UG
classes. What additional we need to understand about preconsolidation pressure is that σ’B
can also be considered as the yield stress for soil state along BC.
So, when the soil is loaded from C to D, it will exhibit elastic behaviour till D because the
yield stress is at σ’B. So that is an important aspect which is very important for module 4. So,
up to σ’B since the yielding has not occurred, so, up to σ’B it will exhibit and kind of elastic
199
response, the moment it crosses point D once the reloading reaches D again the yielding
happens and soil will further undergo elasto-plastic response along DH.
So, you can see that along DH, it will again show the same elasto-plastic response. If the soil
is unloaded at E, now, there is a new yield stress the preconsolidation stress then
consolidation stress σ’E becomes the new yield stress for the soil. So, then the question arises
what happened to the previous yield stress σ’B. So that is what we told soil keeps things in
memory but it gets erased also.
When the new loading comes, so that is what has happened from D to E again some plastic
rearrangement happened. So, from E if you unload the new yield stress for that particular soil
is σ’E. So, yield stress is not constant and changes depending on the loading that is stress
history. So, this we need to keep in mind unlike other materials yield stress in soil keeps
changing and it depends upon the previous loading or the stress history, the stress state on EF
is over consolidated and it behaves elastically similar to BCD.
To summarize, there are a lot of factors that affects the shear strength of cohesive soil as
compared to cohesionless soil. Presence of pore water pressure its dissipation effects shear
strength and are governed by its soil hydraulic conductivity. For undrained condition volume
changes 0 and pore water pressure excess. Poisson’s ratio for undrained conditions 0.5 for
drain condition pore the pressure is 0 and volume change exists.
200
Volume change governs the stress history of soil normal and overall consolidated condition.
So, volume change it governs the stress history of soil to how much it has been subjected to
in the past. Normally consolidated state undergoes high volume change and exhibit elasto-
plastic behaviour and an over consolidated state exhibits minimal volume change and exhibit
elastic response.
So that is why in certain construction if we know that the soil is heavily over consolidated
then we do not have to bother too much about its settlement. Because it will be mostly in
elastic limit the moment the load which is acting on that soil exceeds its preconsolidation
pressure yes then the settlement will be more.
Over consolidated soil has already undergone high stress in the past which is plastic
rearrangement. So, over consolidated soil exhibit less volume change till it reaches the
maximum past stress which is the preconsolidation stress. Over consolidated soil exhibit
elastic behaviour due to less volume change. The maximum stress the soil is subjected to in
the past is called preconsolidation pressure.
Preconsolidation pressure is also known as yield stress. Yield stress is not fixed for a given
soil and yield stress changes and is equal to the maximum stress the soil is subjected to in the
past. So that is all about this lecture, we have discussed basically about 2 important aspects
that governs the shear strength of the soil that is drained and undrained characteristics. A lot
more is going to come in the similar lines related to undrained characteristics as well as drain
characteristics.
201
We will see that as we proceed in the lecture. So, in the next lecture, we will see because of
this how the stress-strain response of cohesive soil. So, we have completed the first part of
shear stress of cohesive soil. We will see now the stress strain response and the shear strength
the way we have seen for granular soil. So that is all for now. Thank you.
202
Lecture – 22
Shear Strength Cohesive Soil - Stress Strain
Welcome back. So, we have seen we have started shear strength of cohesive soil. Now, in this
lecture, we will see the typical stress strain characteristics of cohesive soil and how it influences
the shear strength of cohesive soil. Now, we have already seen a kind of stress strain response
for loose and dense state of the soil when we discussed granular soils. So, more or less the
lectures will be in the similar lines, but the initial condition what we stated as dense and loose,
there is a slight difference when it comes to cohesive soil.
So, we will see what is that difference? The typical stress-strain response of cohesive soil it is
similar to granular soil which means to say the end response remains same. In granular soil the
initial compaction state is defined in terms of loose medium and dense we have seen that the
stress strain response is dependent on the initial compaction state. For cohesive soil, the initial
state is defined in terms of stress history normally consolidated or over consolidated.
Now, possibly you will appreciate the point why we discussed about the volume change
characteristics of the soil it is basically for this the stress history dictates the kind of stress-
strain response of cohesive soil. The stress-strain response of NC is similar to loose state. The
stress strain response of OC is similar to dense state. So, stress strain response is in terms of
deviator stress and axial strain because we are referring basically to triaxial tests.
203
In general shear strength of course, if soil is 𝜏𝑓 = 𝑐 ′ + 𝜎𝑓′ tan 𝜙 ′ , this is a general expression.
Now, when will c’ be there, when it is not there, that is a typical thing which we need to see.
Now, let us not go by the judgment that a cohesive soil is always have cohesion, this is not
correct. As we move in the lecture, as we progress in this particular module, we will see that it
is not so under what circumstance we have cohesion we will see later, but in general we can
consider the Coulomb’s model 𝜏 = 𝑐 ′ + 𝜎 ′ 𝑡𝑎𝑛𝜙 ′ and this is for effective stress a long term
condition.
So, typical stress-strain response of cohesive soil: Now, more or less things remain same the y
axis, x axis have changed, but the response remains more or less similar. For an NC soil that is
NC, state when it is sheared, we can see that it is a loose state, why it is a loose state, because
it has not been stressed too much in the past. So, NC, state of cohesive soil is sheared particles
move horizontally and downwards because it is in a loose state undergo compression during
shearing.
So, it is it is getting packed more how because in the spaces it gets readjusted significant plastic
rearrangement of particles happen and that will result in a kind of response which you can see
here similar to that of loose state. Now, for OC state, when it is sheared shearing is restrained
by interlocking. Why? Because the present stress which we are applying the soil has already
experienced in the past.
So, it is stiff enough to respond to that particular stress. So, since the packing is done in a stiff
204
manner, the particles have to move upwards. Now, this causes expansion or dilation, the way
it we explained it for dense state of this soil. Formation of loose back pockets when it dilates
and that is known as shear bands. So, because of which there is a dilating response and there is
a strain softening which happens.
So, this is mainly because of the upward moment of the particle which causes dilation. Now,
there is an intermediate state a stress state, which is called lightly over consolidated or LOC.
Now, whether an LOC will exhibit a peak behavior or not, it depends upon the amount of over
consolidation it has been subjected to. So, it need not always exhibit peak.
So, OC soil exhibit peak shear strength at low confining stress so, this is what it is at low
confining stress when the dilation is possible it will exhibit peak and peak behavior essentially
occurs at low actual strain. So, that it can show a stiff response at low axial string because of
dilation that is what has been shown by this response on further shearing after peak after
reaching peak the stress comes down to critical state the same manner what we have discussed
for dense state of granular soil.
So, it reaches ultimate state or by strain softening behavior. So, this is a strain softening
behavior and critical state essentially occurs at 10 to 20% instead of here it is not relative
displacement it is also strain. So, normally consolidated state does not exhibit peak and directly
approaches 𝜏𝑐𝑣 or the critical shear strength the progressive increase in stress post yielding. So,
somewhere it yields so, it goes and meets the 𝜏𝑐𝑣 by strain hardening behavior. So, this is a
strain hardening behavior.
205
Now, we have discussed about the typical stress strain response, now, we will see what happens
to a drained volumetric change response. So, volume of OC that is over consolidated decreases
and then it dilates, so, this dilation happens. So, increase in strain increase in the vertical strain
happens. So, volume reaches constant at high strain at high strain condition, what happens is
the concept of critical void ratio comes into picture. So, the volume essentially becomes same.
Normally consolidated undergoes progressive densification as you can see here, till it attains
constant volume state, which is the critical state.
Now, to know the initial state of cohesive soil, it seems does now I am going to introduce
another scenario where the cohesive soil is at the same stress, but a different void ratios and
OCR. So, if you see an NCL behavior and there is unloading, let us say at 3 different points
and on this unloading, like let us say, let us drop a vertical line, so, here, 𝜎′ is same, but the
206
void ratios are different 𝑒𝐴 , 𝑒𝐵 , 𝑒𝐶 𝑎𝑛𝑑 𝑒. If you see it is over consolidation ratio, that is also
different for this point B, it is not that over consolidated.
We can call it as likely over consolidate, it is a relatively less stiff than other over consolidation.
See, for this much it got unloaded. So, it is a heavily over consolidated sample. D is more
heavily over consolidated then point C, here it is same normal stress, but different void ratios.
Now, because of this, even though the is at the same stress condition, the stress strain response
will be totally different for all these 4 points a anyway it is normally consolidated, but for B, C
and D it will be totally different.
How it is different? It has been marked here. Now B A is a normally consolidated plot. So we
know it strain hardens and reaches tau CB. B is likely over consolidated as I told in the
beginning, it may or may not exhibit peak depending upon its over consolidation ratio, let us
say B is not that over consolidated. So, it is also more or less similar to A but slightly different
from that. So it undergoes densification only there is no dilation.
But for C there is a kind of dilation that happens and D that dilation is more which is heavily
over consolidated in the same manner. So, there at large strain, all the samples have same shear
strength. So, this is what it means, now, even though the conditions are different, where is the
actual difference that is coming that is along this peak response at peak response the shear
stresses are different.
But all of them comes out to be the same value later that is at 𝜏𝑐𝑣 , it will also achieve the same
207
void ratio at tau cv, that is a concept of critical void ratio at a critical stage, it ultimately reaches
a critical void ratio, but towards the peak response, the void ratio will be different at this
particular point, if you see the void ratio, then it will be different for all these 4, but in the end,
when it reaches that critical state, the void ratio also becomes comparable.
At peak state shear stress is different, I have already discussed that void ratio will be different.
So, in the volume in the volumetric response also, we can see that A and B undergoes
progressive densification and C and D exhibits initial compression and further dilation.
So, Coulomb’s failure line; how these ABCD translates to columns failure line, this is the
critical straight line fi dash cs and depending upon the amount of dilation, there will be a kind
of peak angle which gets added is a dilation angle which gets added to fi dash critical state and
that results in fi dash peak. So, we can see that C and D falls on the big failure line. So, it also
depends on the kind of confinement it has.
So, we really do not know where actually C and D will fall, but C and D will exhibit peak
behavior with dilation angle whereas, points A and B it falls on the critical straight line why
there is no dilation. So, this stress remains same and similar to density which we discussed
before. There will be a kind of dilation separation at a higher normal stress or higher confining
stress.
208
So, that is all about the shear stress response of the cohesive soil during shear strength test. So,
let us summarize the important points. The stress strain response of cohesive soil is determined
by its stress history. Now, that is a very important point which we need to stress. When we
discuss about granular soil we talk in terms of loose and dense when we talk in terms of
cohesive soil we understand this in terms of stress history, rather NC or OC.
Stress strain response of NC is similar to loose state which is strain hardening. Stress rate
response of OC is similar to dense state which has strain softening. For the same normal stress,
NC, LOC and HOC can have different initial void ratio. At peak state, all the shear stresses are
different for NC, LOC and HOC. But at large strain shear stress the same for these conditions
and that is referred to as critical shear strength.
Reason for exhibiting peak is due to the dilation in the case of OC. The explanation of dilation
remains same as that of dense state of granular soil. So that is all about shear strength of
cohesive soil. We have just begun. Now we will discuss some of the factors in detail. For
example, in the next lecture, we will continue with the discussion on importance of pore water
pressure and it is prediction what we call it as poor order pressure parameters and then we will
continue with the interpretation of triaxial test results.
So we need a background for those interpretations and that is what we have achieved till now.
So, that is all for this lecture. Thank you.
209
Department of Civil Engineer
Lecture – 23
Pore Water Pressure and Skemptons Equation
Welcome all of you, so we are now currently discussing shear strength of soils, in the last
lecture we have seen shear strength of cohesive soils and we have noticed that there are 2 3
important aspects which we should be knowing very well to understand the shear strength
behaviour of cohesive soils properly. One is the drainage condition which results in drained
and undrained conditions and the other aspect is with respect to stress history.
Now with respect to drainage condition, let us say when we have a drain undrained condition
existing the net result is the pore water pressure and as we have already seen in the past and
when we have understood soil mechanics for the first time we know that voids the pore water
present in it and the pore water pressure makes soil mechanics different from the other solid
mechanics.
So, in today is lecture we will discuss a bit on pore water pressure now in advanced soil
mechanics why we should bring the subject again? So, in undergraduate we have learned about
pore pressure and the method of calculating effective stresses but in this we will try to
understand a bit more in detail like if you need to predict pore water pressure normally it is
measured.
Now in certain circumstances for solving some problems numerically you also need to have
what is the undrained condition, what sort of total stress that gets developed and what is the
kind of effective stress which is responsible for the soil mechanical behaviour. If that is the
case, we should also need to have the knowledge of pore water pressure. So, we will get into
some details of pore water pressure the method of prediction of pore water pressure and certain
other aspects which are very relevant for module 3 and module 4.
Again, we are building up the platforms in all these subsequent lectures what we are doing is
we are building platforms for a better understanding of module 3 and module 4.
210
So, let us start with pore water pressure, so importance of pore water pressure, so when total
stress. Now total stress we say it is the external load which is acting this also we have seen
when we say total stress that is the stress acting on the medium. When total stress is applied on
a saturated soil mass there is an instant increase in pore water pressure nothing new, the rate at
which total stress is applied and the rate at which pore water pressure dissipate are critical
factors that governs the mechanical behaviour of cohesive soil.
So, whenever we say mechanical behaviour we know that it is effective stress condition which
is important but undrained condition do exists. Now the rate of loading governs what type of
undrained condition it has and how much of the pore water pressure gets developed. Now the
hydraulic conductivity of the soil determines how fast the pore water pressure dissipates. Now
these 2 jointly determines what will be the effective stress which is gained in this soil at a given
time.
And what will be the equilibrium effective stress and when it is going to reach because all these
factors, it influences the mechanical behaviour or the stress strain behaviour or the strength of
the soil. Let us see a typical construction sequence now for comparison have placed the drained
response as well where the soil drains fast and there is no possibility of development of water
pressure.
So, the first figure represents how total stress changes that means let us say there is a
construction that is happening there is a foundation and there is a building load which gets
added up or maybe an embankment load which gets added up. Now, what happens? The total
211
stress increases with time. So, this is the pattern in which the total stresses increases now what
is the instant response when soil masses saturated soil is loaded there will be pore water
pressure.
Now since this is a completely drained response you can see that there is no development of
pore water pressure, so it is 0. So, this can be treated as 0 now it remains same. Now what will
be the change in effective stress since this is 0 whatever is the total stress that will be the gain
in effective stress as you can see here. So, 𝛥𝜎 ′ = 𝛥𝜎 this is equilibrium but any point along
the curve also this is satisfied because it said drained response.
Similarly there is a volume change has happened and this is a typical compression case where
it compresses and it follows the same sequence as that of the loading. Now let us see the
undrained response let us see that this is the increment of total stress this is the loading and
after this particular point it remains constant. So, let us see what happens for pore water
pressure in the case of pore water pressure this is the instant response now this is an undrained
response where water is not allowed to drain off fast.
So, there will be a mounting of pore water pressure since it is saturated there is an instant
capture of Δ 𝜎 in terms of Δ u. So, whatever is the change in total stress that remains same in
the case of pore water pressure up to this particular point t. Now for us as a layman, there is no
further change in total stress that is there is no further change in loading. So, as a layman one
would always understand we have already completed the loading on the soil mass till now
nothing has happened to the system. So, there is nothing to worry.
Now this aspect gets defeated when you know or when you learn soil mechanics the reason is,
whatever has to happen will happen after the completion of construction or after the external
load gets completed. So that is what you can see here. Now after the construction got completed
beyond that point the pore water pressure will slowly get disappeared and the excess pore water
pressure it tried to achieve or it becomes close to 0. I am cautiously using the term close to 0
because there will be some differences.
So, whatever be beyond the construction period the pore water pressure starts reducing what is
the net effect? You can see that the effective stress remains 0 till the construction got over
because that is a saturated system and whatever loading it is all captured by pore water pressure
212
and beyond the construction stage, you can see that the effective stress which starts increasing
now effective stress starts increasing you can see that beyond certain time period.
Whatever is the total stress applied what is the load applied effective stress becomes equal to
that. So, change in effective stress is equal to change in total stress after some time now what
is the net effect of this as the intergranular stress goes on increasing as the effective stress goes
on increasing this will induce some sort of volume change. So that volume change you can see
till the construction got over there is no volume change which is happening.
Now all the volume change happens after that now this particular information is very vital to
understand what is the importance of pore water pressure? So, at the end of the construction, it
is important for us to assess what will be the kind of pore water pressure either it is measured
or it is estimated. Now it is not possible to measure always because instrumentation for us is
pretty costly. Only very very important projects instrumentations are done and it is monitored.
But in normal cases to understand the effect of pore water pressure it is not always possible to
measure it rather it is estimated and that is the whole crux of today is lecture that we will try to
see how we can estimate pore water pressure. This is what I am summing up there is no increase
in external loading you excess that is excess pore pressure start dissipating and there is an
increase in effective stress which results in volume change.
Now volume change is nothing but settlement as a function of time imagine in the case of an
embankment or a building the settlement happens once the construction phase gets over. So, if
we do not anticipate the kind of pore water pressure, if we do not estimate the kind of pore
pressure that we anticipate in the field which is dependent on the soil characteristics then we
may end up in trouble that is why there are lots of failures that has happened in terms of
settlement.
213
So, with that topic very clear that pore water pressure need to be estimated let us see how we
can go for prediction of pore water pressure. So, it is apparent that the knowledge of pore water
pressure is mandatory for undrained loading change in pore water pressure Δ u with change in
total stress is necessary for different projects like water retaining structures where it is bound
to be undrained and pore water pressure keeps changing.
Predictions of pore water pressure with different loading conditions are necessary for what to
define the effective stress and then effective stress dictates what is the mechanical behaviour
of soils Loading conditions are in terms of total stress change that we have to keep in mind,
whenever we say loading on a soil that is always expressed in terms of total stress. Now total
stress change can be in terms of hydrostatic which causes volume change and deviatoric which
causes shear failures both are important.
Pore pressure develops for each of the loading stages the concept of pore pressure parameters
were introduced for predicting excess pore water pressure under different loading conditions.
So, the bottom line is we will be introducing something known as pore pressure parameters for
predicting the excess pore water pressure corresponding to whatever is the loading condition.
Now those who have understood this concept of pore pressure parameters still I would suggest
that you go through this in detail because this chapter and this discussion is important in stress
pat as well.
214
So, having known that pore water pressure parameters are important, let us see Skempton pore
pressure parameters defined A and B and this is age old paper in 1954 it is published by
Skempton the pore pressure coefficients A and B in geotechnique in 1954. But over these all
these periods, I think this still remains a fundamental aspect in many of the soil mechanics
problem. So, it is important for us to know what was the genesis or where it all started with.
So, Skempton pore pressure parameter A and B becomes very important there are different
variants of this but the concept remains the same. An increase in principle stresses which is
defined in terms of that is total stress and external load whatever B that is defined in terms of
principle stresses. So, any increase in principle stresses results in Δu under undrained condition.
Dissipation of Δu causes volume change we have known this.
Now for deriving this A and B we need to do some indirect approach, let us see what it is, let
V be the initial volume of soil Skempton when we say it undergoes volume change or there is
sport a pressure it is all associated with some sort of strain. We know the elastic volumetric
strain
ΔV 1 − 2 μ
𝜖𝑉 = = [Δσ1′ + Δσ′2 + Δσ′3 ].
V E
This is a drained condition characteristic where effective stress characteristics are used, so
ΔV
volumetric strain is always associated with drained condition. So that is why 𝜖𝑉 = . So, this
V
is the expression and this we have seen already in the previous lecture. Now, this we know is
215
drained conditioned response consider undrained condition with no drainage now this is what
we actually need.
Assume some sort of volume change occur under fully undrained condition now, this is not a
possible condition because under undrained conditions only pore water pressure develops but
we are adopting an indirect method for obtaining this pore water pressure. So, what we are
doing is we are trying to assume make an assumption that there is some volume change that
has happened in an undrained condition.
Now what could be the possibility of such a volume change now this is possible only if volume
change occur due to the compressibility of water which is associated with Δu now there is Δu
water, let us assume that water is compressible and because of this compressibility property
the volume change can happen during undrained condition now water is incompressible it does
not undergo volume change.
So, it that concept remains same and it is very well valid but here we want to understand we
want to determine what is Δu, so that is why we are making such an assumption. So, we
associate that is assumed volume change in undrained condition to compressibility of water.
So, mathematically change in volume under undrained condition due to Δu which is associated
with compressibility of water can be expressed 𝑎𝑠 − Δ𝑉 = 𝑉𝑤 𝐶𝑤 𝛥 𝑢.
216
-ΔV is the water volume change that has occurred under undrained condition which is
associated with compressibility of water now, compressibility of water is 𝐶𝑤 and 𝑉𝑤 is the
volume of water in the voids.
Since it is saturated volume of voids remains equal to volume of water Δ u is the pressure which
creates this volume change. Now you may be wondering from where this expression has come,
this is similar to consolidation settlement equation you might be remembering this equation
very well. This is consolidation settlement 𝑆𝑐 = 𝐻 𝑚𝑣 𝛥𝜎′. Δ𝜎′ is the one which causes this
settlement.
What is it settlement? It is nothing but the volume change here it is one dimensional so, it is
ΔH, instead of ΔH here it is ΔV. So that is C corresponds to ΔV here H is the original height
here it is what is the original volume and here we are associating the volume change to only
the compressibility of water, so where is water water is in voids. So, volume of water which is
equal to volume of voids becomes the original volume.
So, H corresponds to volume of water here, 𝑚𝑣 is coefficient of volume compressibility in

consolidation, the same property which represents compressibility of the soil skeleton here it
is 𝐶𝑤 which is exclusively related to compressibility of water and what causes this settlement
that is Δ𝜎′ and consolidation here the volume change is caused by Δu please remember this is
assumption.
Now we can also write Vw we can replace Vw by total volume how because n is porosity, n is
𝑉𝑣 𝑉𝑤
the porosity, 𝐶𝑤 is the compressibility of water and 𝑛 = = . So, you can always write
𝑉 𝑉
ΔV
𝑛𝑉 = 𝑉𝑤 . So that is what has been replaced here −ΔV = nVCw Δ u . So, we can write − =
V
nVCw Δ u Now this is already obtained in the previous slide.
So, if we equate these two we get

1−2μ
𝑛𝐶𝑤 𝛥 𝑢 = [Δσ1′ + Δσ′2 + Δσ′3 ]
E
𝑛𝐶𝑤 𝛥 𝑢 is the volume change that has occurred we are now trying to equate these two. In
essence, like it does not make much sense because this is not going to happen but then we are
assuming that it is going to happen and that volume change is associated with the volume
217
change that would have happened under drained condition. So, Δ u is that term what we need
we need to estimate what is the excess pore pressure that gets generated.
So,
1−2μ
𝛥𝑢 = [Δσ1′ + Δσ′2 + Δσ′3 ]
𝑛𝐶𝑤 E
Now triaxial condition is a very prominent condition in the lab. So, let us now specifically
discuss about triaxial condition, what is the triaxial condition? We have 𝛥𝜎2 ′ = 𝛥𝜎3 ′. So, if
you put that we have
1−2μ
𝛥𝑢 = [Δσ1′ + 2Δσ′3 ]
𝑛𝐶𝑤 E
We can replace the effective stresses by total stresses

So,
1−2μ
𝛥𝑢 = [(Δ𝜎1 − 𝛥 𝑢) + 2(Δ𝜎3 − 𝛥 𝑢)]
𝑛𝐶𝑤 E
1−2μ
𝛥𝑢 = [Δ𝜎1 + 2Δ𝜎3 − 3Δ𝑢]
𝑛𝐶𝑤 E
Since,
[Δ𝜎1 + 2Δ𝜎3 ] = [Δ𝜎1 − Δ𝜎3 + Δ𝜎3 + 2Δ𝜎3 ]
We have,
3(1 − 2 μ) 1−2μ
[1 + ]𝛥 𝑢 = [Δ𝜎1 − Δ𝜎3 + 3Δ𝜎3 ]
𝑛𝐶𝑤 E 𝑛𝐶𝑤 E
218
[𝑛E𝐶𝑤 + 3(1 − 2 μ)]𝛥 𝑢 = (1 − 2 μ)[Δ𝜎1 − Δ𝜎3 + 3Δ𝜎3 ]
Simplifying, finally we get the expression,
3(1 − 2 μ) 1
𝛥𝑢 = [Δ𝜎3 + (Δ𝜎1 − Δ𝜎3 )]
[𝑛E𝐶𝑤 + 3(1 − 2 μ)] 3
1 1
𝛥𝑢= [Δ𝜎3 + (Δ𝜎1 − Δ𝜎3 )]
𝑛E𝐶𝑤 3
[1 + ]
3(1 − 2 μ)
1 1
𝛥𝑢 = [Δ𝜎3 + (Δ𝜎1 − Δ𝜎3 )]
𝑛𝐶 3
[1 + C 𝑤 ]
s
3(1−2 μ)
Now, 𝐶𝑠 is known as compressibility of soil skeleton 𝐶𝑠 = . So, the expression for Δ u
𝐸
is with an assumption that the saturated soil first of all it is for saturated soil it behaves as
perfectly elastic material.
And this expression corresponds to that this expression whatever we have obtained it
corresponds to first of all saturated soil this corresponds to saturated soil it corresponds to
triaxial condition and it also goes by the assumption of perfectly elastic. So, these are some of
the conditions which are pertaining to this particular expression, so Δu is due to change in
confining pressure because 𝛥𝜎3 is related to confining pressure interaction testing and change
in deviator stress 𝛥𝜎1 − 𝛥𝜎3 .
219
So, there are 2 components one due to confinement the other due to the deviatoric component
both results in Δu. So, the equation for Δu can be written as the
𝛥𝑢 = 𝐵 [𝛥𝜎3 + 𝐴(𝛥𝜎1 − 𝛥𝜎3 )]
1 1
Now 3 is replaced by A, why because corresponds to strictly a perfectly elastic condition.
3
Now we will not expect soil to behave as elastic like any other materials.
So, for generality, so to deviate from elastic behaviour to give that option we have introduced
Skempton introduced A parameter. So, this equation was proposed by Skempton in 1954. So,
A and B are very popularly known as Skemptons pore water pressure parameters the use of A
and B becomes prominent in the subsequent chapters, so, it is important to understand this.
So, now the same equation this equation is written here
𝛥 𝑢 = 𝐵 [𝛥𝜎3 + 𝐴(𝛥𝜎1 − 𝛥𝜎3 )]
𝛥 𝑢 = 𝐵 𝛥𝜎3 + 𝐴𝐵(𝛥𝜎1 − 𝛥𝜎3 )]
This is an alternate way of representing pore water pressure with Ᾱ
𝑢 = 𝐵 [𝛥𝜎3 + Ᾱ(𝛥𝜎1 − 𝛥𝜎3 )]
Now we will discuss a bit about both these parameters what are its physical significance, the
1
first one is pore water pressure parameter 𝐵 = 𝑛𝐶𝑤 . Now let us say the situation is for
[1+ ]
Cs
completely saturated case, if it is completely saturated now we have assumed water to be

compressible but in fact water is not compressible. So, 𝐶𝑤 for a completely saturated state with
state when there is only water 𝐶𝑤 tends to 0.
220
Even though we made an assumption that it is not, so that whole assumption was to correlate
for obtaining Δu, so if there is no volume change which is going to happen. Then what is going
to happen, so that was the kind of logic which was used. So, when for fully saturated case
𝐶𝑤 approaches 0 and hence the value will be equal to 1. So, this clearly indicates for a fully
saturated system B = 1.
Now what is happening to partially saturated soil, so the same expression instead of 𝐶𝑤 it
becomes 𝐶𝑓 where 𝐶𝑓 is called compressibility of pore fluid which includes both water and air
now, it becomes more compressible it is not equal to 0 because the contribution of air is there
1 1
which is highly compressible. So, 𝐵 = 𝑛𝐶 = 𝑛𝐶𝑓 . Now what is happening this
[1+ 𝑤 ] [1+ ]
Cs Cs
component is no more 0 it is an appreciable amount.
So, in partially saturated this is what it is 𝑢𝑎 pressure is more 𝑢𝑤 which is held in the pore
pressure is negative, it is u is 0 and 𝑢𝑤 is negative that we call it as negative pore pressure or
soil suction and 𝑢𝑎 is greater than 𝑢𝑤 the compressibility of the whole system is now more
than one and it is appreciable. So, 𝐶𝑓 is very much greater than 0 due to high compressibility
of air, what is the net result?
Because of this, for example it is a dry state completely dry state 𝐶𝑓 is only due to the
contribution of air this will be exceptionally high quantity. Now if this is very high, the whole
of the B comes down drastically and it will be very much close to 0. So that is what is written
here for dry states B value 0 because see if this denominator becomes very high. Now as
saturation increases from the dry state as the saturation increases the B value also increases by
the time it reaches fully saturated B = 1.
So, the range of value of B is in between 0 and 1, where 0 indicates dry state, 1 indicates
saturated state B parameter hence can be used to represent soil saturation, so this parameter B
is indication of saturation. Now as the confining stress increases air expels or it gets dissolved
in water saturation increases and B value increases. So, this is a very effective way of using B
parameter for ensuring the saturation of triaxial soil sample.
221
We all know and we have seen triaxial testing the first procedure is after mounting the soil is
to get it saturated because conventional triaxial testing we discussed only about saturated soil.
So, you need to saturate the soil first and we need to know whether it has actually got saturate
now be parameter which comes from Skempton pore pressure parameter is used to check
whether the soil is saturated or not, this will become clear in the subsequent slide.
1
Now let us talk about pore pressure parameter A we have already seen 𝐴 = 3 for perfectly
elastic case and triaxial loading. Note this is also important whatever we have discussed as
Skempton equation is with respect to triaxial loading but we know soil is not perfectly elastic,
1
so 3 is not valid always, so that is why it is replaced by A. A parameter changes with loading
or with increase in deviator stress that means.
There is a variation of A with loading which gets which takes place. So, there will be initial A
so, before the starting and during the loading A changes and finally there will be at the failure
condition there will be a value of A which is denoted as 𝐴𝑓 . So, 𝐴𝑓 is the A parameter at failure.
Now, the value of A is also influenced by stress history depending upon whether it is OC or
NC. Normally, 𝐴𝑓 varies between - 0.5 and 1.5 that is what has been observed in the lab.
So, this is typical A f values which is given by Skempton in his paper in 1954 where we can
see that for example, normally consolidated clay it varies from 0.5 to 1 and for a heavily over
consolidated clay it is from - 0.5 to close to 0. This clearly indicates that the stress history of
the soil influences the parameter A and with loading also the A will change because it is A is
associated with deviator stress and deviator stress keep changing. So, the A value also changes.
222
Now Δ u for different stages of loading in attraction compression testing. So, we have discussed
about the actual condition here. So, let us see how Δ u develops for different stages of loading.
So, the first stage is confining stage soon after saturation of the soil sample in triaxial testing
we go for confinement application of confining stress. Now if it is an undrained condition, how
pore water pressure develops that is what we will see you know.
So, this is a typical traction sample and 𝛥𝜎𝑟 is the confining stress where 𝛥𝜎𝑟 = 𝛥𝜎1 = 𝛥𝜎3 , it
is an all-round stress condition only volume change happens. Now, if it is an undrained
condition volume change will not happen but pore water pressure increases the excess pore
pressure for confining stage is represented by 𝛥 𝑢𝑐 .
So, substituting this in the equation of Skempton equation we can write 𝛥 𝑢𝑐 = 𝐵[𝛥 𝜎3 + 0] is
very much there that is the order on pressure plus 𝐴 [ 𝛥𝜎1 − 𝛥𝜎3 ]. Now this is equal, so it
𝛥𝑢
becomes 0. So, 𝐵 = 𝛥𝜎𝑐 . Now this is where I told it can be used to check whether triaxial
3
sample is saturated or not when will it be saturated when b value is equal to 1.
So, if you have done saturation, if you have done the triaxial testing in the lab you would have
definitely looked for B value. Now, how B value the presence one now if for a fully saturated
condition, what happens is whatever is the increase in 𝛥𝜎3 that will be fully reflected by 𝛥𝑢𝑐 if
there is some amount of air present in the sample whatever is the load that is applied that is
𝛥𝜎3 what is it total stress which is applied 𝛥𝜎3 .
223
It will be partitioned between pore water pressure 𝛥𝑢𝑐 and some volume change that happens
because of the expulsion of air. So, it is not 1 is to 1 hence, 𝛥𝑢𝑐 will be always less 𝛥𝜎3. So, B
value will be less than 1 and that is how it is, so if you improve the saturation and the whole of
the pore are filled with water, whatever 𝛥𝜎3 is applied that will be reflected as Δ u but it is very
hard to get the value 1.
But it will be close to 1 anything about point 9.95 can still be considered a saturated sample,
so after completing the confining stage then we go for the deviator stress that is the loading
stage. Now, here please be careful this is the soil mass to which 𝛥𝜎3 was already applied. So,
the order on stress is there now, you start increasing the axial load which is 𝛥𝜎1 . Now here in
the loading stage 𝛥𝜎3 remains constant or 𝜎3 whatever has been applied in the beginning is not
changing in the loading stage.
So, 𝛥𝜎3 there is no change in the order on stress, so whatever actual load you are applying that
itself will be 𝛥𝜎𝑑 or 𝛥𝜎1 but the concept of 𝛥𝜎𝑑 is what causes failure in triaxial soil sample.
Now what stress causes failure whatever actual load that has been applied 𝛥𝜎3 was already
existing before. So, whatever is causing failure is the total load minus whatever has been
applied on to the sample minus this whatever was there initially that is 𝛥𝜎3 .
So that is how it becomes 𝛥𝜎1 - 𝛥𝜎3 this 𝜎𝑑 is very much valid for the failure condition but
here in this case we should know that whatever is the actual stress that we are applying that is
𝛥𝜎1 and that remains 𝛥𝜎1 1 because there is no increment in 𝛥𝜎3 . 𝛥𝜎3 = 0 and hence,
𝛥 𝑢𝑑 = 𝐵[0 + 𝐴𝛥𝜎1 ].
So that here in this case it is saturated, so B = 1. So, you can write 𝐴𝑓 that is 𝑡ℎ𝑒 A parameter
𝛥𝑢
at failure 𝐴𝑓 = 𝛥𝜎𝑑𝑓 . So, this both corresponds to at failure condition what is the 𝛥𝜎1 at failure
1𝑓
and what is the pore pressure at failure. So, 𝐴𝑓 is defined. So, the final stage is the summation
of the two which is 𝛥 𝑢𝑓 = 𝛥 𝑢𝑐 + 𝛥 𝑢𝑑𝑓 .
And total pore pressure at the end of the stage is equal to the one which comes from the
confining stage and the one which comes from the loading stage. So, the parameters A and B
are determined in the lab for the conditions simulated close to field problems. So, if you want
to apply this for estimating the pore pressure corresponding to any given geotechnical problem,
224
we need to understand what is the A and B parameter. B is not important A is important, what
is the kind of a parameter corresponding to that soil for that particular field condition.
So, this is a typical variation of A parameter for with respect to stress history this we have
already discussed what will be the kind of stress strain response this is a strain hardening and
this is strain softening for NC and OC for corresponding to NC for an undrained condition, the
pore water pressure will be positive. So, this is positive whereas for OC there will be initial
compression for OC in the volume change.
That corresponds to a positive pore water pressure in the beginning and then it comes to
negative for OC because OC tends to dilate. So, those which tends to dilate will try to suck the
pore water inside, so that is negative pore water pressure, so that is given here. So, pore water
pressure will be negative for OC or it keeps on reducing. So, correspondingly A parameter for
NC you can see that it will reach close to one at the end of the test for NC and for OC it will
become progressively negative. So, you can see this close to -0.5 for OC.
225
So, to summarize this chapter, it is pore water pressure prediction is important for studying the
undrained behaviour of cohesive soil. Skempton equation with parameters A and B can be used
to predict the pore water pressure as a function of change in total stresses. This pore water
pressure equation corresponds to triaxial condition that we have already seen and it is estimated
in 2 stages one corresponding to confining stage the other one is loading stage.
Pore pressure parameter B represents soil saturation for fully saturated B = 1 completely dry B
= 0. So, B parameter helps to assess saturation of actual soil sample parameter A changes with
loading and A value at failure is considered as 𝐴𝑓 . A parameter is influenced by stress history
1
of soil for triaxial loading and considering soil to be perfectly elastic the value of 𝐴 = 3 . So
that is all about pore water pressure estimation using Skempton equation.
In the next lecture we will see that how to use this same equation rearranged in a different
format and how that becomes important for certain field problems that is all for now. Thank
you.
226
Department of Civil Engineer
Lecture – 24
Overall Pore Water Pressure Parameter
Welcome all, so this is the continuation of our previous lecture which is on pore water pressure
parameters. We have already seen Skempton’s pore water pressure parameter which
corresponds to triaxial condition. In this lecture we will see a different way of presenting the
pore water pressure parameter and what is known as overall pore pressure parameter B.
So we will start with the lecture which is overall pore water pressure parameter 𝐵̅ . There are
different ways by which pore water pressure equation can be expressed. Now before going to
𝐵̅ we will see a typical representation of pore water pressure equation this expression of 𝛥 𝑢
𝛥 𝑢 = 𝐵[Δ𝜎3 + A(Δ𝜎1 − Δ𝜎3 )]
is what we have already seen, and understood. 𝛥 𝑢 expression can also be written in this
manner where
1 3𝐴−1 1
𝛥 𝑢 = 𝐵 [3 (Δ𝜎1 + 2Δ𝜎3 ) + (Δ𝜎1 − Δ𝜎3 )] (Δ𝜎1 + 2Δ𝜎3 )
3 3
This expression (Δ𝜎1 − Δ𝜎3 ) we all know it is deviator a stress q which we have seen in the
1
previous lecture. And what is this expression? (Δ𝜎1 + 2Δ𝜎3 ) this is nothing but mean stress
3
for triaxial condition. So, the Skempton pore pressure equation can be rewritten in this manner
to express it in terms of mean stress and deviator stress.
227
1 3𝐴−1
Now if soil behaves elastically we know that 𝐴 = 3. So, if you substitute it in then the
3
3𝐴−1
term (Δ𝜎1 − Δ𝜎3 ) becomes 0. So, pore pressure depends solely on the main principle
3
stress only in those circumstances where the soil behaves elastically, deviator stress is not there
1
is no contribution from deviator stress. Now, for all those conditions where 𝐴 ≠ 3deviator
stress has significant influence on Δ𝑢.

Now, let us see what is overall pore pressure parameter 𝐵̅ which relates Δ𝑢 with only Δ𝜎1. 𝛥𝜎1
is normally considered in terms of loading. So, what is the total stress which is acting because
of some load external load acting on the soil. So, Bishop in 1954; in this again a classical paper
which deals with pore pressure coefficients the use of pore pressure coefficients in practice
which is published in geotechnique.
The relationship this 𝐵̅ this relationship is useful while dealing with practical geotechnical
problems. Knowledge of initial Δu as a result of loading is necessary to estimate the stability
of foundation soil of an earth dam maybe for drained condition after a pretty long time, what
will be the kind of state of the soil for that we need to know what will be the kind of pore
depression that gets developed due to loading which is very essential for situations like
foundation of soil often earth dam.
Stability and settlement of impervious fill of earth dam due to superimposed layer in staged
construction. Increase in stress due to rapid drawdown it is not only that you are adding water
228
to it but when there is immediate release of water from the soil, the effective stress increases.
Now, due to this effective stress increase there will be a kind of settlement, so that issue is
relevant with rapid drawdown. Now pressure change due to removal of weight from overlying
soil.
So, what will be the pore pressure change due to release that is we are removing the weight.
So, what happens during that time to pore pressure again we start with the same equation
Skempton equation we are doing against some mathematical rearrangement
𝛥𝑢 = 𝐵[Δ𝜎3 + A(Δ𝜎1 − Δ𝜎3 )]
Δ𝑢 = 𝐵[Δ𝜎3 + Δ𝜎1 − Δ𝜎1 + 𝐴(Δ𝜎1 − Δ𝜎3 )]
Δ𝑢 = 𝐵[Δ𝜎1 − (Δ𝜎1 − Δ𝜎3 ) + 𝐴(Δ𝜎1 − Δ𝜎3 )]
Δ𝑢 = 𝐵[Δ𝜎1 − (1 − 𝐴) (Δ𝜎1 − Δ𝜎3 )]
Δ𝜎
Δ𝑢 = 𝐵Δ𝜎1 [1 − (1 − 𝐴) (1 − Δ𝜎3)]
1
Δ𝑢 Δ𝜎3
= 𝐵 [1 − (1 − 𝐴) (1 − Δ𝜎 )]
Δ𝜎1 1
Δ𝜎
𝐵̅ = 𝐵 [1 − (1 − 𝐴) (1 − Δ𝜎3)]
1
𝛥𝑢
Now, 𝛥𝜎 is a function of 𝐵̅ . So once you know 𝐵̅ we can always calculate what will happen
1
to Δu as a function of 𝛥𝜎1.
229
For that to explain this, let us take an example of an earth embankment which is shown here.
Now, there is a soil element here which is at a depth of h. So, assume this 𝛥𝜎1 it is not assume
delta sigma is very much equal to gamma integration where gamma is the unit weight of soil.
𝛥𝑢
Now, what we are assuming is that for about equation = 𝐵̅ makes it convenient to
𝛥𝜎1
determine the variation of pore water pressure with increase in height of the embankment
during stage construction. Now, what we are assuming is that the entire loading and the
influence that is happening at this particular point is due to the major principle stress 𝛥𝜎1.
So, 𝛥𝜎1 so, whatever is the change that is happening at this point is entirely due to its geostatic
stress which is 𝛾ℎ . And that is true also because there is no load acting whatever is the
superimposed load which is acting at this particular point that can be assumed as the major
principle stress 𝛥𝜎1 which means to say any kind of loading is represented here in terms of
𝛥𝜎1 .
So, this assumption is important because then only you can determine the pore pressure as a
function of 𝛥𝜎1 . So, we are concerned about other principle stresses here we are correlating or
we are assuming whatever stress change happens that is 𝛥𝜎1 only. So, once we know that then
the above equation, it makes it very convenient to determine the variation of border pressure
happening with loading. Loading means increase in height of the embankment during stage
construction.
So that is the beauty of this particular equation pore pressure at any instant of loading that
means staged construction can be determined.
230
Δ𝜎3
Now but we have seen that in this particular equation Δu also is a function of stress ratio Δ𝜎1
so that also becomes important so, Δu also depends upon the principle stress ratio. Now we
Δ𝜎′
know coefficient of lateral earth pressure K is equal to Δ𝜎3′ so, this is known to us. So K is equal
1
Δ𝜎 −Δ𝑢
to Δ𝜎 3−Δ𝑢 you are just transferring the store total stress condition.
1
𝐾Δ𝜎1- 𝐾Δ𝑢= Δ𝜎3 − Δ𝑢
𝐾 Δ𝜎1 + Δ𝑢(1 − 𝐾) = Δ𝜎3
Δ𝑢 Δ𝜎3
𝐾+ (1 − 𝐾) =
Δ𝜎1 Δ𝜎1
231
Δ𝜎
So, if you substitute this for Δ𝜎3. The equation takes the forms like this
1
Δ𝑢 Δ𝑢
= 𝐵 [1 − (1 − 𝐴) (1 − 𝐾 + Δ𝜎 (1 − 𝐾))]
Δ𝜎1 1
Δ𝑢 ΔU
[𝐵 − B(1 − 𝐴)1 − K)]=[(1-K)(1 − A) Δσ )]
Δ𝜎1 1
Δ𝑈 B [1 −(1−A) (1−K)]
=
Δ𝜎1 [1−B(1−A) (1−K)]
B [1 −(1−A) (1−K)]
𝐵̅ = [1 −B(1−A) (1−K)]
So, again it is the same overall pore pressure parameter but in a different form in terms of
coefficient of lateral pressure.
So, in the embankment problem if the horizontal layers are of infinite extent it is reasonable to
assume that the lateral deformation is negligible. If the lateral deformation is negligible, then
we can always say this coefficient of earth pressure K is equal to coefficient of lateral earth
pressure at rest which is 𝐾𝑜 and 𝐵̅ can be determined from laboratory test by maintaining the
Δ𝜎′
appropriate constant ratio 𝐾 = Δ𝜎3′ which basically simulates the field condition.
1
So, one thing is if it is of infinite extent then it can be treated as 𝐾𝑜 or any other value of K
which resembles the field condition. Now if the slope is steep and in a condition of limit
equilibrium that is almost the failure condition then K will take a minimum value and which is
close to active condition which is denoted as 𝐾𝑓 so that 𝐾𝑓 can also be simulated. Now, for any
slope with factor of safety which is greater than one which is not at failure K value will be in
between these 2 bounds that is in between 𝐾𝑓 and 𝐾𝑜 .
So once we know the appropriate K value that can be determined put into this particular
equation and that gives us what will be the pore pressure for any change in loading.
232
So the summary of overall pore pressure parameter is the use of this parameter 𝐵̅ is a
convenient way of estimating pore pressure as a function of loading in a practical situation. Δu
depends on 𝛥𝜎1 and the principle stress ratio delta sigma 3 upon 𝛥𝜎1 . 𝐵̅ is expressed in terms
1
of coefficient of lateral earth pressure which comes from this. If soil behaves elastically 𝐴 = 3
then Δu will depend only on the main principles stress.
1
For 𝐴 ≠ 3 then deviator stress part has a significant influence on Δu. So that is all for today is
discussion on pore water pressure. There is one more section left to be discussed in this which
we will see in the next lecture. Thank you.
233
Lecture - 25
Pore Water Pressure Plane Strain Effect of Sampling
Welcome back all of you, in the last lecture we have seen the importance of pore water pressure
in soils and the need for its estimation we discussed about Skempton pore water pressure equations
and A and B parameters which are known as Skempton pore pressure parameters. So this lectures
continuation of the last lecture and we will basically deal with pore water pressure for plane strain
conditions, Henkels pore water pressure equation and the effect of sampling on pore water pressure
which occurs in the field. So these 3 aspects we will cover in today's lecture.
So for specific geotechnical problems we have already seen plane strain condition is more realistic
than triaxial condition, we have already seen this. So we will try to have the derivation for Δu
corresponding to plain strain. Now the condition for plane strain let us say we have if you deal
with a principle stresses and principle strains. Let us consider intermediate principle strain 𝜖2 = 0
and that can be expressed as
𝜎2′ 𝜇
𝜀2 = 0 = 𝐸
− 𝐸 (𝜎1′ + 𝜎3′ )
234
This gives
𝜎2′ 𝜇 ′
= (𝜎1 + 𝜎3′ )
𝐸 𝐸
In terms of incremental stresses, we can write
Δ𝜎2′ = 𝜇(Δ𝜎1′ + Δ𝜎3′ )
Now how did we derive the pore water pressure equation we have associated or we have already
allocated. We saying that there will be some volume change that can occur in undrained condition
then we associated this volume change to the compressibility of water. Then we have derived the
equation for pore water pressure doing the same thing for plane strain. So we will have to start
from the same point whatever is the volume change that is due to compressibility of water doing
that associating the volume change under undrained condition to compressibility of pore water.
We can write this is the starting
(1 − 2𝜇)
𝑛𝐶𝑤 Δ𝑢 = [Δ𝜎1′ + Δ𝜎2′ + Δ𝜎3′ ]
𝐸
associating the undrained volume change that could have happened due to Δu to the actual volume
change that would have happened under undrained condition.
Then delta Δ𝜎2′ can be replaced by this equation Δ𝜎2′ = 𝜇(Δ𝜎1′ + Δ𝜎3′ ). So, when you replace
that you get this equation.
(1 − 2𝜇)
𝑛𝐶𝑤 Δ𝑢 = [Δ𝜎1′ + 𝜇(Δ𝜎1′ + Δ𝜎3′ ) + Δ𝜎3′ ]
𝐸
235
So rearranging the terms you get
(1 − 2𝜇)
𝑛𝐶𝑤 Δ𝑢 = [(1 + 𝜇)Δ𝜎1′ + (1 + 𝜇)Δ𝜎3′ ]
𝐸
(1 + 𝜇)(1 − 2𝜇)
= [Δ𝜎1′ + Δ𝜎3′ ]
𝐸
(1 + 𝜇)(1 − 2𝜇)
= [Δ𝜎1′ + Δ𝜎3′ + Δ𝜎3′ − Δ𝜎3′ ]
𝐸
(1 − 2𝜇)(1 + 𝜇)
= [2Δ𝜎3′ + (Δ𝜎1′ − Δ𝜎3′ )]
𝐸
Now this equation is similar to the equation that we have seen during the derivation of Skempton
pore water pressure equation, so this has become very similar to that. Now if we take 2 outside.
And we have we are expressing this effective stress in total stress.
2(1 − 2𝜇)(1 + 𝜇) 1
𝑛𝐶𝑤 Δ𝑢 = [Δ𝜎3 + (Δ𝜎1 − Δ𝜎3 ) − Δ𝑢]
𝐸 2
Now if we take this Δu on the other side.
2(1 − 2𝜇)(1 + 𝜇) 2(1 − 2𝜇)(1 + 𝜇) 1
[𝑛𝐶𝑤 + ] Δ𝑢 = [Δ𝜎3 + (Δ𝜎1 − Δ𝜎3 )]
𝐸 𝐸 2
Now if you notice this is now very much similar to the Skempton pore water pressure equation
except the fact that there it was 1 / 3 and now it is half.
236
Again rearranging.
2(1 − 2𝜇)(1 + 𝜇)
𝐸 1
Δ𝑢 = [Δ𝜎3 + (Δ𝜎1 − Δ𝜎3 )]
2(1 − 2𝜇)(1 + 𝜇) 2
[𝑛𝐶𝑤 + ]
𝐸
1 1
Δ𝑢 = [Δ𝜎3 + (Δ𝜎1 − Δ𝜎3 )]
𝐸𝑛𝐶𝑤 2
[1 + ]
2(1 − 2𝜇)(1 + 𝜇)
So the final pore at a appreciate equation will be
1 1
Δ𝑢 = [Δ𝜎3 + (Δ𝜎1 − Δ𝜎3 )]
𝑛𝐶 2
[1 + 𝐶 𝑤 ]
𝑠
Again rearranging you can get is the same way we have done composability of soil skeleton the
same thing we have done before also and E comes down. So 𝐶𝑠 is the compressibility of soil
skeleton for specifically form plane strain condition and that is
2(1 − 2𝜇)(1 + 𝜇)
𝐶𝑠 =
𝐸
So what is the essential difference between plane strain and triaxial condition. So A parameter for
1 1 1
triaxial actual condition is 3 whereas for plane strain condition 𝐴 = 2 remember this values and
3
1
corresponds to elastic behaviour only.
2
237
So that is about the derivation of equation for plane strain condition. Now let us see Henkel pore
pressure equation in terms of octahedral stresses. Henkel 1960 proposed new equation for Δu in
terms of octahedral stresses I hope you remember what is octahedral stresses? Which, we have
seen in the earlier lecture. So according to Henkel it is
Δ𝑢 = 𝛽[Δ𝜎𝑜𝑐𝑡 + 3𝛼𝜏𝑜𝑐𝑡 ]
So both 𝜎𝑜𝑐𝑡 and 𝜏𝑜𝑐𝑡 has been accounted this is very much similar to our previous 6 Skempton
equation b is replaced by 𝛽 a is replaced by 3 into 𝜎 because 𝜎𝑜𝑐𝑡 mostly it represents condition
similar to that of deviatoric stress which is 𝜎1 − 𝜎3 for triaxial condition so it is one at the same.
Now 𝜎𝑜𝑐𝑡 just to revise let us write 𝜎𝑜𝑐𝑡 is equal to mean principle stresses or mean stress p.
I am not intentionally putting prime because now it corresponds to total stress condition because
pored pressure is there it is an undrained conditions. So I am referring it to t mean stress and
√2
t 𝜏𝑜𝑐𝑡 = Δ𝑞 all these expressions we have seen in the previous lectures. Now if you rearrange
3
in terms of p and q one can write

Δ𝑢 = 𝛽[Δ𝑝 + 𝛼√2Δ𝑞]
So we are just replacing it in terms of p and q. Here 𝛼 and 𝛽 are called pore water pressure
parameters just like a and b and here also 𝛽 = 1 for fully saturated samples. So this is about
Henkels pore water pressure equation we need to understand that this equation is in terms of
octahedral stresses. So in those models which adopt octahedral stresses it is quite handy to see how
the pore water pressure changes with corresponding changes in the octahedral stresses.
238
Β=1
Δ𝑢 = 𝐵[Δ𝜎3 + 𝐴(Δ𝜎1 − Δ𝜎3 )]
= 𝐴Δ𝜎1
Now let us try to find out the relationship between 𝛼 and A, because these 2 parameters are
similar A one is in henkel the other one is in skempton pore pressure equation for constant
confining pressure. So there is there are conditions given for constant confining pressure during
compression of saturated clay. So there are certain thing which comes out from here there is no
change in confining pressure that is constant confining pressure.
So 𝛥𝜎2 = 𝛥𝜎3 = 0 since it is saturated 𝛽 = 1 and B =1. So let us first start with Skempton’s pore
pressure equation. So this is skempton’s pore water pressure equation.
Δ𝑢 = 𝐵[Δ𝜎3 + 𝐴(Δ𝜎1 − Δ𝜎3 )]
So here we know that 𝛥𝜎2 = 𝛥𝜎3 = 0 substituting that we will get
𝛥𝑢 = 𝐴𝛥𝜎1
Now this is Δ𝑢 = 𝛽[Δ𝜎𝑜𝑐𝑡 + 3𝛼𝜏𝑜𝑐𝑡 ] the Henckel’s pore pressure equation which we have just
seen,
We know
𝛥𝜎1 + 2 𝛥𝜎3 𝛥𝜎1
𝜎𝑜𝑐𝑡 = Δ𝑝 = =
3 3
And
239
√2 √2 √2
𝜏𝑜𝑐𝑡 = Δ𝑞 = (𝛥𝜎1 − 𝛥𝜎3 ) = 𝛥𝜎1
3 3 3
For corresponding to a typical triaxial condition we know that 𝑞 = 𝜎1 − 𝜎3 . So Δ𝑞 =
√2
(𝛥𝜎1 − 𝛥𝜎3 ) so finally 𝜏𝑜𝑐𝑡 = 𝛥𝜎1 . We will substitute this back into the equation
3
Δ𝑢 = 𝛽[Δ𝜎𝑜𝑐𝑡 + 3𝛼𝜏𝑜𝑐𝑡 ], this will give

Δ𝜎1
𝛥𝑢 = + √2𝛼Δ𝜎1
3
Now 𝛥𝑢 we already know that 𝛥𝑢 = 𝐴𝛥𝜎1 . So substituting that
𝜎1
𝐴Δ𝜎1 = Δ + √2𝛼Δ𝜎1
3
now you will take this Δσ1 on the other side so that will give us
1
[𝐴 − ] = √2𝛼
3
So this gives the relationship between alpha and A where after rearranging you can get
3𝐴 − 1
𝛼=
3√2
This is the relationship between alpha and A but for specific condition where the confining
pressure remains constant and the soil is fully saturated. So this about henkel’s pore water
pressure equation and its relationship with skempton’s pore pressure parameter A.
𝜎1 = 𝜎𝑣
𝜎2 = 𝜎3 = 𝜎ℎ
𝜎𝑣 = 𝛾ℎ + 𝛾𝑠𝑎𝑡 (𝑍 − ℎ)
𝑢𝑖 = (𝑍 − ℎ)𝛾𝑤
𝜎𝑣′ = 𝜎𝑣 − 𝑢𝑖
So next let us see the effect of sampling in the field on pore water pressure. Many a times we need
to sample soils from the field specifically undisturbed soils sample this is very much important for
studying the compressibility of the soil as well as the shear strength characteristics. If you want to
plan a foundation at a particular level we need to know what is the bearing capacity of that
particular soil So we need to have the undisturbed sample with us.
Now during sampling what happens let us say it is a saturated soil mass and when you do the
sampling whether there is any change that can happen to the soil mass in terms of its pore water
240
pressure. So that is what we will see just now and this part of the lecture I have taken from head
manual of soil laboratory testing volume 3 effective stress tests published by john Wiley. Let us
consider a soil element A in the field.
This is the depiction of a given soil you can see that the ground surface is here, the water table is
given and we know the soil beneath the water table it is saturated. So accordingly the unit weight
here will be bulk unit weight or total unit weight. So here please remember the soil above the water
table is not a saturated soil mass it is a partially saturated. So we will consider this to be bulk unit
weight.
We will consider a soil element A which is shown. Now if you put a piezometer what we will get
the head causing what is the head of water. Here in this case it is a static pore water pressure static
head. So here the water will rise in the piezometer up to the groundwater table it is an unconfined
case. We have vertical stress and horizontal stress acting.
The thickness considered as z and the thickness of the partially saturated portion that is thickness
about the water table is h. So here it is Z - h. So showing the kind of stresses which is acting on
the element which is 𝜎1 , 𝜎2 𝑎𝑛𝑑 𝜎3 . Soil is saturated and it is in equilibrium. So this given soil it
is in saturated state and it is also in equilibrium please note that we are considering the element A.
241
So with respect to that the soil is saturated it is in equilibrium means what? The pore water pressure
has reached to its equilibrium value that is what it means so here we have static pore water pressure
and that is denoted by u static pore pressure u. Let us see the total stress 𝜎1 = 𝜎𝑣 . So let us we are
considering the vertical stress to be major principle stress and 𝜎2 = 𝜎3 = 𝜎ℎ .
Now 𝜎𝑣 what will be 𝜎𝑣 from the figure it will be

𝜎𝑣 = 𝛾ℎ + 𝛾𝑠𝑎𝑡 (𝑍 − ℎ)
Initial pore water pressure 𝑢𝑖 = (𝑍 − ℎ)𝛾𝑤 .
So one can write 𝜎𝑣′ = 𝜎𝑣 − 𝑢𝑖 .
𝜎ℎ = 𝜎ℎ′ + 𝑢𝑖
𝜎ℎ = 𝑘𝑜 𝜎𝑣′ + 𝑢𝑖
Δ𝑢 = 𝐵[Δ𝜎3 + 𝐴(Δ𝜎1 − Δ𝜎3 )]
𝐵=1
So that will give
𝜎𝑣′ = 𝛾ℎ + 𝛾 ′ (𝑍 − ℎ)
and
𝜎ℎ′ = 𝑘𝑜 𝜎𝑣′
𝑘𝑜 coefficient of lateral earth pressure at rest. So 𝜎ℎ can be written as 𝜎ℎ = 𝜎ℎ′ + 𝑢𝑖
242
Now why did we write this, why can we write? 𝜎ℎ = 𝑘𝑜 𝜎𝑣 . Why can we write that? This is not
possible because all these coefficient of lateral pressure active passive case all of them are in
terms of failure condition and hence it is not only failure condition the lateral earth pressure is
always in terms of effective stress why?
Because we are considering the effect of soil or the lateral pressure of soil acting on something
and this is mostly dictated by the effective stress condition. So the equation 𝜎ℎ = 𝑘𝑜 𝜎𝑣 . It is always
in terms of effective stress that is why we have first found out what is the total vertical stress acting
then found out affective component then applied 𝑘𝑜 for getting 𝜎ℎ′ .
Now if you want to obtain 𝜎ℎ then we just need to add pore water pressure to it. So that is what
has been done here which is 𝜎ℎ = 𝜎ℎ′ + 𝑢𝑖 and that gives k naught into 𝜎ℎ = 𝑘𝑜 𝜎𝑣′ + 𝑢𝑖 . Now let
us consider the soil sample is taken from the field that is the process of sampling when soil element
is sampled from the subsurface the confining stresses are relieved. So confinement now reduced
to 0 that is true.
So when you are trying to take this particular soil maybe in a bore hole what we are doing is we
are making a hole here. So this is relieved and when you take out the sample this is also relieved.
So change in stress conditions induce change in pore water pressure now just because there is a
release of stresses or the stresses are relieved, we will have changes in pore water pressure because
it is under undrained condition and it is for a smaller period of time. So both combined there will
be changes in pore water pressure.
So let us start with the Skempton pore water pressure equation Δu is equal and then we need to
find out what is the changes in 𝛥𝜎3 𝑎𝑛𝑑 𝛥𝜎1?. Soil is saturated and mass, so B can be considered
equal to 1.
243
Δ𝜎3 = 𝜎3𝑓 − 𝜎3𝑖
= 0 − 𝜎ℎ
= −𝜎ℎ
Δ𝜎1 = 𝜎1𝑓 − 𝜎1𝑖
= −𝜎𝑣
Δ𝑢 = 𝐵[Δ𝜎3 + 𝐴(Δ𝜎1 − Δ𝜎3 )]
Now Δ𝜎3 is what the changes in confining stress Δ𝜎3 = 𝜎3𝑓 − 𝜎3𝑖 now 𝜎3𝑓 we know it is 0
because the confinement is lost. And 𝜎3𝑖 is nothing but the lateral pressure 𝜎ℎ . So this will give
Δ𝜎3 = −𝜎ℎ . Now what is Δ𝜎1? it is Δ𝜎1 = 𝜎1𝑓 − 𝜎1𝑖 , again 𝜎1𝑓 is also 0 because it is sampled
Δ𝜎1 = 0 − 𝜎𝑣 .
If we substitute this in the equation
Δ𝑢 = 𝐵[Δ𝜎3 + 𝐴(Δ𝜎1 − Δ𝜎3 )]
you will get
Δ𝑢 = [−𝜎ℎ + 𝐴(Δ𝜎1 − Δ𝜎3 )]
Δ𝑢 = −𝐴𝜎𝑣 − (1 − 𝐴)𝜎ℎ
244
Now we have Δu what is the change in pore water pressure which is
Δ𝑢 = 𝑢𝑓 − 𝑢𝑖
So uf = 𝑢𝑖 + Δ𝑢. So we are interested to find out what is that particular pore water pressure after
sampling that is nothing but 𝑢𝑓 which is equal to what was existing in the beginning 𝑢𝑖 + Δu.
𝑢𝑓 = 𝑢𝑖 − 𝐴𝜎𝑣 − (1 − 𝐴)𝜎ℎ
So this is what we get for 𝑢𝑓 when you substitute Δu that is a change in pore water pressure.
Now let us say that the given soil element is a typical NC soil. So A will be positive and close to
0.5. So what is the implication A is positive and this close to 0.5 which means that this portion
𝐴𝜎𝑣 − (1 − 𝐴)𝜎ℎ in the equation 𝑢𝑓 = 𝑢𝑖 − 𝐴𝜎𝑣 − (1 − 𝐴)𝜎ℎ will be positive that means𝑢𝑖
minus some quantity. So when it is u i is minus some quantity we know that the final border
pressure will be always less than the initial pore water pressure.
So it is apparent that final pore water pressure will be less than initial pore water pressure only
under the condition when A is positive. Now this can impart a negative pore water pressure because
the pressure or the confinement is getting released. So the tendency of the soil is to suck more
water and that is during volume change condition but that is during drained condition. But since
the time is less here he has it we can consider this to be a typical undrained condition.
245
If that is the case then the pore water pressure tends to negative or suction. So this induces a kind
of apparent cohesion in the soil. This results in an internal effective stress which is instrumental in
holding the soil structure together without collapsing. So this apparent cohesion as a result of this
negative pore water pressure helps to hold the soil and it will not collapse and fall. So when
normally consolidated soil is sampled it becomes slightly over consolidated because of this release
and that is the reason why it can stand on its own.
So that is all about the effect of sampling. So what happens new to release there will be a kind of
a negative pore water pressure which helps to hold the soil intact. So that is why when the
undisturbed sample is taken from the field it can stand on its own that is basically because of the
effect of pore water pressure changes. So let us summarize this lecture pore water pressure equation
is derived for plane strain condition.
1 1
If the soil behaves elastically 𝐴 = 2 for plane strain as against 𝐴 = 3 in triaxial condition which
we have seen henckels pore pressure equation is in terms of octahedral stresses the relationship
between henckels border pressure parameter 𝛼 and Skempton’s pore water pressure parameter A
for constant confining pressure has been derived. A saturated soil sample from the field undergoes
pour water pressure changes.
246
And we have estimated it the final pore water pressure after sampling has been derived. For a
typical saturated NC soil sampling causes release of confinement. So release of confinement
induces negative pore water pressure that holds the sampled soil without collapse. So these are all
the things which we have seen as part of pore water pressure estimation. So in the next lecture we
will just work out one small problem just to give us a feel of how pore water pressure is determined
and how practically it is significant. So that is all for now. Thank you.
247
Lecture – 26
Pore Water Pressure Estimation
In this lecture, we will try to work out a problem related to water pressure equation. So here the;
problem given us a soil is isotropically consolidated at a cell pressure of 600 kPa. So there is cell
pressure of 𝜎3 = 600 kPa please notes that it is consolidated. So if it is consolidated there will be
no pore water pressure, pore water pressure will be equal to 0 we will understand this when we
discuss more about the triaxial testing which we will see in the next lecture.
With drainage valve closed, now we have closed the drainage cell pressure is increased to 720 kPa
in a pore water pressure and resulting in a pore water pressure of 115 kPa please remember the
soil is close to saturation. So there will be changes in pore water pressure and hence after the
drainage valve was closed when you increase the cell pressure to 720 kPa’s. So then 𝜎31 that is
the next stage it is 720 kPa.
And this results in pore water pressure u31 = 115 kPa the sample was then subjected to loading.
So without opening the drainage valve the soil is subjected to loading to give a deviator stress of
248
550 kPa's. So deviator stress 𝜎𝑑 = 550 𝑘𝑃𝑎, what is deviator stress? The stress in addition to 𝜎3
that is deviator stress which is applied axially the pore water pressure at the end of the test was
262 it is a total pore pressure at the end of the test.
So uf = 262 𝑘𝑃𝑎 please make a note here you need to understand the question very carefully we
need to know whether the pore water pressure is in excess of the confining or right from the
beginning to the end. So here there is no mention and this clear that pore water pressure at the end
of the test means pore water pressure it is total pore water pressure that has accumulated,
cumulative pore water pressure or in the soil sample.
So we are asked to determine the pore water pressure parameters B, A, Ᾱ and 𝐵̅ and this is what
we do actually in the laboratory for a given situation or for a given soil at a given depth we will
find out using triaxial testing what is it pore water pressure parameters, this pore water pressure
parameter values can be used for any other numerical modeling or whatever. So this value can be
input.
And when I say parameters mostly I am associating it to the parameter A because B if it is saturated
we know it is 1 or depending upon the partial saturation we will know what the value is. So here
change in confining stress 𝛥𝜎3 = 720 − 600 that is what it is here it is equal to 120 kPa and
change in pore water pressure for the confining stage please note this is confining stage is 115 kPa
because initially it was 0. Now we can use the pore water pressure equation
𝛥𝑢 = 𝐵[𝛥𝜎3 + 𝐴(𝛥𝜎1 − 𝛥𝜎3 )]
Now please make a note here nothing has been told about the saturation of the soil samples. So we
cannot really assume B = 1 we need to determine it and we have whatever 𝛥𝜎3is there and 𝛥𝜎1 =
𝛥𝜎3 in the confining stage. So first let us see the confining stage and this condition we know.
𝛥𝑢 115
So 𝐵 = 𝛥𝜎𝑐 and both are known. So substituting 𝐵 = 120 will get B = 0.958 which is very close
3
to saturation but not equal to 1 and practically in the lab it is difficult to obtain the value 1 it will
be close to 1 maybe something from varying from 0.9 of words is all expected for depending upon
the type of clays the odd maybe if it is silty, so as it becomes more and more clay it is difficult to
achieve B = 1. So anything around close to 1 is considered to be saturated.
249
Next is the loading stage. So we have determined B so that B value we have to use B ≠ 1 here,
𝛥𝜎𝑑 is equal to now this is the loading stage. So we have 𝛥𝜎𝑑 = 𝛥𝜎1 − 𝛥𝜎3 here 𝛥𝜎3 in the
loading stage is 0. So because during the application of deviatoric stress there is no change in the
confinement it remains same. So 𝛥𝜎3 = 0 that gives 𝛥𝜎𝑑 = 𝛥𝜎1 .
𝛥𝑢 = 𝐵[𝛥𝜎3 + 𝐴(𝛥𝜎1 − 𝛥𝜎3 )]
262 − 115
Ᾱ = 𝐴𝐵 = = 0.267
550
Again substituting it back we know the value of B 𝛥𝜎3 is 0. So that will give 𝛥𝑢 = 262 − 115
please make a note here we are considering only loading states, so we cannot take the cumulative
pore water pressure. So we need to detect the confining stage pore water pressure. Now 115 is
𝛥𝑢𝑐 = 115. So we are trying to minus this from the total pore water pressure, so that you get the
pore water pressure corresponding to the deviator stress application
𝛥𝑢 = 𝐵[𝛥𝜎3 + 𝐴(𝛥𝜎1 − 𝛥𝜎3 )]
262 − 115 = 0.958[0 + 𝐴(550 − 𝛥𝜎3 )].
Because whatever is the change in deviator stress that is equal to 𝛥𝜎1 so that we have seen here
550 − 𝛥𝜎3 which is nothing but 0. So here it will be B into A first that is A bar we have seen
we already know that A bar = A into B. So here B into A that is Ᾱ that is equal to
262 − 115
Ᾱ = 𝐴𝐵 = = 0.267
550
250
And please remember this is due to the understanding that 𝛥𝜎3 = 0 𝑎𝑛𝑑 𝛥𝜎𝑑 = 𝛥𝜎1. So Ᾱ =
0.267; so A can be found out by dividing it by B which is equal to
0.267
𝐴= = 0.279
0.958
So we have determined B, Ᾱ A what is left is 𝐵̅.
Overall pore pressure parameter 𝐵̅:

𝛥𝜎3
𝐵̅ = 𝐵[1 − (1 − 𝐴)(1 − ( )]
𝛥𝜎1
𝛥𝜎
This is the stress ratio (𝛥𝜎3 ). Now here please do not go by just like that fast stress ratio stress ratio
1
indicates what is the ratio between the total the vertical stress that is major principle stress by or it
𝛥𝜎
is between 𝛥𝜎3 . So it is minor principle stress upon major principle stress.
1
So do not go by delta here because it is the overall stress applied on to the soil in the previous slide
we have seen 𝛥𝜎𝑑 = 𝛥𝜎1 that is only for that deviatoric stress change that is for the loading stage
here the stress ratio corresponds to the overall stress that is applied during the whole course of the
test. So here we will have 𝜎3 and 𝜎1 . So what is 𝜎3 ? 𝜎3 is 120 kPa because that is what has been
applied.
251
And 𝜎1 = 670 kPa in the previous slide it was 𝛥𝜎1 = 𝛥𝜎𝑑 . So that is why it was 550 here it is the
overall stress that is acting which is 670. So that is which comes from here, from the figure
representing the forces on the soil element, so 𝜎3 is there on the soil sample and seen 𝜎𝑑 = 𝜎1 −
𝜎3 . So 𝜎1 = 𝜎𝑑 − 𝜎3. So that is the overall stress acting on the soil which is used for determining
the stress ratio.
So 𝜎1 − 𝜎3 = 550, so 𝜎1 = 550 + 120 = 670 so I have just forgotten the delta for the time
being just for you to understand that it is the overall stress which is acting on the soil sample. So
here it is 𝛥𝜎1 = 670 which comes from here. So substituting this we will get
120
𝐵̅ = 0.958[1 − (1 − 0.279)(1 − ( )]
670
So, 𝐵̅ = 0.391. Alternately you can also determine
𝛥𝑢 262
𝐵̅ = = = 0.391
𝛥𝜎1 670
So that is the cumulative change in pore water pressure. So in the previous slide we have splitted
it into confining and loading stage but overall pore pressure parameter we will consider the overall
pore pressure change which is equal to 262 upon the 𝛥𝜎1 here is the total 𝛥𝜎1 which is equal to
670. So that will also give the same result 0.391. So either of these equations we can use. So that
is all related to pore water pressure equations and it is estimation.
So, this problem we have discussed just to give you a feel of how the estimation of pore water
pressure happens and how this can be used further please make a note that this information is very
important when you deal with module 3 and module 4. So before going to module 3 and module
4 please revise these portions properly, so that we can use these information in the subsequent
modules.
So, we have now completed the discussion on pore water pressure. And in the next lecture onwards
we will see the shear stress of soil very specifically which will be more like a bit of reputation
from your undergraduate portion but with more interpretation. So that is all for now and we will
see in the next lecture. Thank you.
252
Lecture - 27
Triaxial Testing
Welcome back all of you, so in the last lecture we have seen a few aspects related to pore water
pressure and its estimation. Next we will move on to shear strength determination with specific
focus on triaxial testing, many of you would have already studied triaxial testing properly during
your undergraduate. So, please take this as a revision of triaxial test because in the subsequent
lectures we will rely a lot on triaxial test.
And its interpretation which is mandatory for understanding stress path and critical state soil
mechanics. So, I cannot forego this topic even though it is start in UG and in addition we will be
discussing a few additional points which is actually needed for this particular course. So, we will
have a brief on what axial test is and what are its possibilities and then in the subsequent lecture
we will see its interpretation.
So, this is all about triaixal test. So, triaxial test is the most versatile shear strength test for soils
due to control of drainage condition and pore water pressure measurement. Now in the field you
have different possibilities a low permeable soil has the drainage capacity quite low which means
253
to say when a load is acted upon on a low permeable soil you are likely to have undrained condition
for substantial amount of time.
Similarly there will be development of pore water pressure. So, there will be undrained condition
pore water pressure with time pore water pressure dissipation happens and at very large time what
happens is and a completely drained condition would exist. So, we have the strength of the soil at
its short term that is undrained condition what will be its strength and for a long term what will be
its strength.
Now it is very difficult to define something called undrained shear strength because we know that
based on effective stress principle the strength is entirely dependent on effective stress condition
however soil will be possessing some amount of effective stress even under undrained condition
and that is why we need to have the idea of pore water pressure and most of the effective stress
condition which is relevant for the soil it is there right at the beginning.
Which is mostly associated with its initial state from where the soil has been sampled. So, for
simulating different conditions the most versatile test that we have is a triaxial test. So, we will see
quickly what traxial test apparatus looks like because that is needed to understand this particular
aspect how it becomes versatile. So, this is the soil sample which is mounted in the triaxial cell.
And on top of it you have porous disc which permits drainage of water from soil outwards when
the loading happens.
And there are 2 caps one is the bottom cap and the top cap for delegating the load equally to the
soil. Now we have the provision for applying cell pressure then there is a provision for recording
pore water you are now by this time I hope all of you are familiar with u if I write only u It means
pore water pressure, so pore water pressure can be measured and there are a provision for drainage
of the soil sample.
And there is a provision for back pressure which is more or less in together with drainage condition.
So, we have different outlets then the soil is housed inside a membrane, so that the soil is not able
to interact with the water outside because cell pressure is delegated there will be water within this
254
chamber. This is the ramp through which the actual load is delegated to the sample and there is an
air bleed valve in order to allow the air to escape from this water.
So, when you fill this chamber with water for applying cell pressure the any entrapped air bubbles
or dissolved air bubbles is forced further out through this air bleed valve. So, this is the overall
setup of traxial test. So, when the stress is applied or when the load is applied on to the soil sample
referring the figure. First is the confinement that you give 𝜎3 and then is the actual load which
comes the top.
So that is denoted as 𝜎1 now in triaxial testing it is important to understand what causes failure.
So, initially there was a confining stress 𝜎3 all round this stress will not partake in the process of
failure because there is an isotropic stress condition as a mean stress condition where all round
stresses are equal. Now this creates volume change but not shear stress within the soil this we have
already discussed then what causes failure.
So, when you apply the load that is the ramb this load which is delegated to the actual this system.
So that causes failure that induces shear within the soil this also we have discussed before. But
what is exactly that is causing failure that is the deviator stress component which causes which
induces a shear and that causes failure that is given by this. So, total stress which is the actual stress
as well plus the confining stress which is acting on the top of the sample that is together the major
principle stress.
Hence the deviator stress is nothing but 𝜎1 − 𝜎3 this is further explained here𝜎1 = 𝜎3 + the force
which is applied through actual force that is applied through this ramp divided by area of cross
section of the sample. So, 𝜎1 = 𝜎3 + 𝜎𝑑 . Hence 𝜎𝑑 = 𝜎1 - 𝜎3 . So, F is the axial load applied through
loading arm that causes failure. So, area of cross section of soil sample and 𝜎1 is the major principle
stress 𝜎3 is the minor principle stress and 𝜎𝑑 is the deviator stress.
255
Now what are the different tests performed in conventional triaxial compression test setup. Now
why I am using this term conventional means I am talking about the basic triaxial test equipment.
Nowadays there are different variants of triaxial testing possible you have cyclic triaxial testing
you have unsaturated the soil to triaxial testing there are stress but for traxial testing. So, there are
different provisions which got added up for specific situations.
But I am talking about the very basic triaxial compression test. So, these are the 3 possible tests
which one can do in triaxial compression testing which is unconsolidated undrained denoted by
UU, consolidated undrained CU and consolidated drained CD now all the information are there in
the name itself. So, there are 2 stages essentially with respect to loading. First is the initial
consolidation phase which is acted upon by cell pressure.
So, initial consolidation under cells pressure and there is drainage and there is loading stage or
shearing stage now whether there is initial consolidation under cell pressure for UU test no and
whether drainage happens under loading. Loading means both cell pressure as well as the triaxial
stress or the deviator stress, so there is no drainage which is happening. So, as it is mounted it may
be saturated and then the load is applied without any drainage and consolidation.
So, consolidated undrained CU there is initial consolidation which means that there is some
changes that is happening to this soil sample during the consolidation phase that is confinement is
256
applied and then the drainage valve is open, so that water moves out and consolidation can happen
whereas in the case of loading that is in the shearing stage it is not the drainage wall was closed
and hence it is a complete undrained condition.
Then consolidation also happens that is drainage valve is not closed during confinement also it is
open and during shearing also it is open and hence the there is volume change which is happening
during both the stages. Now accordingly the first test is very quick because nothing you have to
do you have to mount the sample saturate and then load it and hence the failure happens relatively
faster.
CU it is relatively slow whereas CD is extremely slow by this is relatively slow and slow we will
come to know a bit later as we discuss this test. So, conventional triaxial tests are only applicable
for saturated soil. Now this aspect one need to understand very carefully some undergraduate
students they do not appreciate this fact properly but we need to understand that for conventional
triaxial test is only applicable for saturated soil alone.
The 3 important stages in triaxial testing the first one is the saturation stage we need to a certain
that the soil samples are fully saturated if the samples are not fully saturated we need to perform
the saturation. So, the first aspect is we draw the sample from the field for determination of shear
strength and mostly these samples will be undrained. Now we cannot guarantee that the particular
soil sample that we get from the field is saturated.
So, first we will check whether the soil is saturated if it is not then the first stage is to execute or
to induce saturation in the soil sample then comes the confining stage or the consolidation stage
and finally the loading or the shearing stage. So, these are the 3 important stages in to triaxial
testing.
257
So, let us see the features of unconsolidated undrained test we are not going to discuss any results
in this lecture we will understand clearly what are the possibilities of these tests, no drainage and
consolidation during confining and loading states we have seen this now pore water pressure
measurement is not useful. This is an undrained test what else is going to happen other than pore
water pressure. But still one can always measure the pore water pressure in UU test.
But that is not going to be useful why? We will see this when we discuss about the results of UU
test primarily a total stress analysis required for determining undrained shear strength. So, this is
purely a total stress analysis and the second statement and third statement they are complimentary
to each other. Because if there is no pore water pressure measurement use there if the pore water
pressure measurement is not going to be useful.
Then it is apparent that we are not going to have the results in terms of effective stress. So, it is a
total stress analysis and only useful for undrained shear strength confining pressure is applied and
the specimen is sheared there is no waiting period because consolidation is not there. Rate of
loading is fast and results in quick failure. No volume change test, so this is specifically and no
volume change test if the soil is fully saturated.
258
So, these minute points one need to keep in mind why should I write if soil is fully saturated? A
clue you will get when because you if you have gone through the lectures of pore water pressure
what is the effect of partial saturation on B parameter. So, same thing is applicable here if there
are air present in the sample then air will get displaced in the process. So, what is going to happen
is if it is fully saturated whatever is the initial effective stress that remains same.
But if you allow any sort of volume change and that will alter the initial effective stress and once
it is altered the strength also gets affected. So, that is the reason, so it is understood that there is no
volume change which takes place. So, the stage 1 is saturation the stage 2 is confinement. So, we
have the pore water pressure which is developed which is corresponding to the confinement sigma
3 you may choose to measure or you may choose not to measure it is not mandatory. But you see
even if it is measured is not going to be useful in this test.
So, the loading stage is you are applying the deviator load through the axial provision which is
there in the triaxial testing and hence the cumulative pore water pressure and the failure is 𝑢𝑓 , so
that is how it is.
Next we will see consolidated undrained test, so we have drainage allowed during confinement
stage resulting in isotropic consolidation yes we need to note here that the consolidation happens
under isotropic condition after consolidation stage u = 0 or u = backpressure please hold on for
259
some time we will discuss at length what is meant by back pressure ideally during consolidation
the drainage valve is kept open. So, there will be no development of pore water pressure.
Hence at the end of consolidation stage pore water pressure will be 0 provided the drainage line is
kept open to atmospheric pressure. Now there is some arrangement for applying back pressure
which we will see soon and if it is, so then at the end of consolidation stage this particular pore
water pressure in terms of back pressure do exist these are again some minor points but very
important points you need to keep in mind for appreciating or understanding the development of
pore water pressure.
How you determine the effective stress all these things are dependent on your understanding on
what is pore water pressure what is back pressure how it is applied. So, these things become
important. So, please pay specific attention on these points even though it may look very minor.
So, after consolidation the drainage valve always closed axial load is increased and pore water
pressure is measured till failure.
Now you can see here in CU test it is invariably necessary to obtain the pore water pressure only
if you know the pore water pressure you can give the shear strength or you can represent the stress
strain in terms of effective stress. And here it is extremely important that we need to have the
effective stress strength parameters he was that is more relevant. Now due to end restraints of
loading cap we have already seen in the initial figure there are 2 loading caps top and bottom.
And hence during the undrained loading pore water pressure generated may not be uniform within
the soil sample. Just to avoid this non uniformity it is always ideal to have loading at a slower rate.
Now you will appreciate why this particular test is relatively slow than UU test this is mainly
because that the loading rate has to be slow. So that the pore water pressure development is there
is a height applicable for the sample.
It is not like a consolidation sample there is an appreciable height. So, when the load is delegated
it may not instantly equilibrate the pore water pressure may not instantly equilibrate. So, there will
be a certain time lake that is what is known as u may not be uniform and after keeping for some
260
time it will achieve a uniform value it is a hydrostatic stress, so it has to be equalized. So, stage 1
is saturation same as before stage 2 consolidation there is confinement and isotropic consolidation.
So, it is a drained stage then the drainage valve is closed and the loading stage happens is the same
as before but there will be development of pore water pressure that is 𝑢𝑓 during the undrained stage
and this 𝑢𝑓 need to be known.
As I told in the beginning we will discuss a bit on what is back pressure in triaxial testing and this
backpressure is applicable for all the 3 testing because back pressure is useful for saturation and
saturation is mandatory for all the 3 types of tests, all the 3 types of triaxial testing. Now what is
back pressure? It is the water pressure applied to the soil pore water. So, it this pressure goes within
the soil sample to force the air present in the voids if any.
So, if it is fully saturated then most of the air voids will be filled with water. But in case it is not
then it is the water pressure which will help the air bubbles which are present in the voids to
dissolve in the water already there is water in the pose. So, the air is displaced from the voids and
it dissolves in the water and in the process some amount of minor amount of water gets in and fills
the pose.
261
Please make a note this process is without changing the effective stress condition or the initial state
of the sample only what is happening is the air which is present dissolves into the water and will
get displaced by water. So, water displaces the air and hence it achieves complete saturation which
is a condition mandatory before consolidation or loading application of backpressure is needed to
facilitate full saturation and saturation should be ensured.
So, it is the raising of pore water pressure without altering the initial effective stress condition of
the soil. This is very important the moment it changes the effective stress then it is going to have
its effect on strength. Now you will probably understand the importance of discussing pore water
pressure. Now just before discussing this lectures we had on pore water pressure we discussed
about A and B parameter there when we discussed we specifically told that B parameter is used
for assessing soil saturation.
So that application will come here, so if B parameter is less than 1, so you check the B parameter
what do you do for checking the B parameter you need to increase the cell pressure by some
amount that is not substantially some small increment of cell pressure you give and measure the
𝛥𝑢
port water pressure and B is we know that it is that is the during the confinement.
𝛥𝜎3
So, you measure both if it is substantial value then if it is much less than 1 it indicates that it is not
saturated. If it is close to 1 it indicates it is saturated. So, both the if the parameter is much less
than 1 which means that it is not saturated then both the cell pressure and the back pressure is
increased in equal increments. Now back pressure is mostly along with the drainage line from
where the pressure is delegated into the soil sample and that raises the pressure.
Now if cell pressure is not raised simultaneously there will be a kind of blowing out effect and that
will destabilize the soil sample. And always we have to keep in mind that back pressure application
should not be substantial. It also has to be done in small incremental stages. So, if the B parameter
𝛥𝑢
is less than 1 both cell pressure and back pressure is increased in equal increments. So, 𝐵 = 𝛥𝜎𝑐.
3
262
And we need to keep in mind that always the cell pressure is kept more than back pressure else
there will be case of negative conditions. So, always the cell pressure is greater than back pressure.
So, the procedure for back pressure application will not alter the initial effective stress condition
of the soil, if the initial effective stress condition is affected this will influence the strength and
determination of the soil. So, this is the condition when the confinement is applied.
′
So, the initial effective stress condition is 𝜎3𝑖 = 𝜎3 − 𝑢. So, this is the initial condition of the soil
now we will make an assessment of B parameter let us say it is substantially less than 1. So, you
need to saturate the sample this much is understood. So, you apply the back pressure. So, when
you apply the back pressure that is in terms of delta u that is increase there is an increase in cell
pressure 𝛥𝜎3 and 𝛥𝜎3 = 𝛥𝑢, so it is equal increment.
So, back pressure is increased by 𝛥𝑢, so the cell pressure is also increased by 𝛥𝜎3 , so that the
effectively the condition remains the same. So, 𝛥𝜎3 = 𝛥𝑢. So, what is the effective stress
condition for this backpressure application is
′
𝜎3𝐵𝑃 = (𝜎3 + 𝛥𝜎3 ) − (u + 𝛥𝑢)
= 𝜎3 − 𝑢
′ ′
𝜎3𝐵𝑃 = 𝜎3𝑖
263
Uses of back pressure once the back pressure is applied the saturation will not happen instantly.
This also we need to understand when a load is acted upon into the soil we say that immediately
the water captures all the load yes this is true but there is a bit of time lake by the time it delegates
to the entire soil sample. So that the back pressure is applied the saturation will not happen
instantly.
It is not an instant response just because of this each increment should be small enough and
equilibrium time should be long enough. So, we need to wait for some time and the increment has
to be smart let us say that ultimately the backpressure application is around 200 kPa. So, always
start with a small increment let us say 50 kPa and then wait for some time till it equilibrate.
How do you know it will equilibrate when you keep measuring the pore water pressure it becomes
constant. So that indicates that the pore water pressure has almost equilibrated if it is still rising
we have to wait for some time. So, this is what we need to keep in mind back pressure application
is in small increments and that small increment has to be kept for sufficient time and that is known
as the equilibrium time.
So, equilibrium time permits air displacement and water saturation the change in pore water
pressure after the application of back pressure need to be monitored till it reaches the constant
value the B parameters need to be checked at the end of equilibration time. So, once before the
pressure equilibrate we know what is the incremental pore water pressure that has happened for an
incremental rise in cell pressure or the back pressure back pressure application.
And cell pressure applications simultaneously we will measure 𝛥𝑢 and 𝛥𝜎3 check the B value if B
value is not satisfactory normally it is as a rule of thumb it is considered as 0.95. So, anything
above 0.95 you can treat it to be saturated but if it is less than 0.95 is an indication that the saturation
has to be made the process of back pressure increment is repeated. So, with 50 kPa it is not getting
saturated and other 50 is added.
So, total back pressure now becomes 100 kPa is simultaneously the cell pressure is also increased.
So, this process goes on the monitoring goes on the whole of the saturation process goes in an
264
incremental manner till the B value achieved is around more than 0.95. So, the back pressure is
maintained throughout the test and may not be released once the saturation is achieved. There is
another important point like once the back pressure is applied.
So, It kept as it is once the B value is achieved it is not that we close it. So, this backpressure would
remain or it may remain in the soil sample till the test is completed. So, back pressure is more like
a static pore water pressure we have seen what is a static pore water pressure only the excess pore
water pressure brings about changes or volume change in the soil mass or densification in the soil
must this we have seen.
So, here back pressure will not change any such changes will not bring about any such changes.
So, it is more like a static pore water pressure. So, in case of back pressure application, u at the
end of consolidation stage will be 𝑢𝐵𝑃 that is there will be some amount of pressure already
existing in the sample. If back pressure is not applied for saturation then u at the end of
consolidation stage will be 0.
Stage 2 is the consolidation stage after saturation stage the cell pressure is increased to the required
confinement value which is already pre decided by the person who is conducting the test the
drainage valve is opened for consolidation to happen. Now none of this was relevant in UU test,
265
so that we will possibly understand why this test is taking a lot of time consolidation is also not an
instant response it takes time.
So, for under a given pre decided confining stress the consolidation is carried out and for that
drainage valve is opened during consolidation volume change is measured as a function of time
volume change is equal to volume of water expelled out of the soil for fully saturated soil again
now you will understand why soil has to be fully saturated whatever volume change that happens
during consolidation we need to capture that.
And for capturing that it is mandatory that the whole of the soil is fully saturated. So, whatever is
the volume of water that gets expelled out that itself is equal to the volume change that is happening
during consolidation the volume of water flowing out of the soil is recorded and plotted as a
function of time as like given here. So, this is volume of water expelled and this is the time we can
see that initially there is an increase in volume of water expelled.
And it slowly becomes constant for that particular amount of confinement which is applied. So,
once it becomes constant we understand that the consolidation phase got completed the plot is used
to verify whether the consolidation phases complete. So, if the soil sample is not fully saturated
then the about this plot will not represent the actual volume change during consolidation phase
why?
Because the volume change occurring due to expulsion of air cannot be recorded. Now let us say
that there is some amount of air also present because of poor saturation. Now what will happen
when the load is applied when the confinement load is applied for conducting the consolidation
the air also displays and there will be a volume change associated with the displacement of air that
volume change we cannot capture.
We can capture only from the volume of water which is getting expelled out which means to say
that whatever has been represented by this curve is not complete if the soil is not fully saturated
the volume of air we cannot measure that is the reason it is important that we need to have complete
saturation and this particular curve is very important to decide upon the shearing stage like when
266
it comes next. So, especially that will become important for drained shearing. So, this particular
information becomes important.
So, next is the shearing stage of CU test where drainage valve is close before the application of
axial load and then axial load is increased which is the deviator load, u measurement is mandatory
we have seen they will be end restraints due to loading cap whatever this particular aspect we have
already discussed u may not be uniform throughout the soil sample to avoid the non uniformity of
u loading should be at a slower rate.
267
So that is about CU test and next is CD test. Now CD test is very much similar to that of CU test
everything remains the same. Only thing is the drained loading happens saturation and confining
stage. The same as that as CU test essential that soil is saturated this all same volume of water
expelled with time during consolidation state is important that information is important volume
change data is used to estimate the rate of loading for shearing stage that is the loading stage.
All the cells are in terms of effective stress because there is no development of pore water pressure
during this test. Now this also induces the time required for conducting this test why because rate
of shearing should be low such that pore water pressure is not developed throughout the test we
have already seen in the previous lectures that if the rate of loading is higher then also there can
be pore water pressure even though the permeability of the sample is more.
Here we need to ensure that there is no pore water pressure that gets generated during loading.
Hence the shearing rate should be low. So that it does not allow any pore water pressure to develop.
So, this makes the CD test extremely slow CD test is suitable for freely draining soil, so that u is
not developed especially for Sandy type of soils where generally CD test is conducted. So that is
all about the different types of tests that is conducted in triaxial those who understand these tests
properly you can think of skipping this lecture.
But what are the different possibilities and during interpretation of triaxial testing we should know
what are the different provisions of these tests. So that is why this lecture is important.
268
So, the triaxial test can be used to simulate different drainage and pore pressure condition thereby
simulating realistic condition in the field. In general there are 3 types of triaxial test UU, CU and
CD. There are 3 stages that is required in the triaxial testing which is the saturation, confinement
and loading or shearing. Full saturation of the soil sample need to be ensured before consolidation
and shearing stage.
B parameter is used to assess saturation and practically B greater than 0.95 represents saturation.
Back pressure is used to increase saturation of the soil sample. In case it is found it is not fully
saturated u measurement is not required in UU test whereas it is mandatory in CU test. The volume
of water expelled with time during consolidation stage can be used to decide the rate of shearing
in shearing stage.
This will prevent the development of u during shearing in CD test. So that is all about the
discussion on different tests that we can conduct in triaxial testing. Next lecture onwards we will
see the interpretation of different types of these tests which we have discussed today. And that
interpretation is extremely important because from what test what shear strength you get is
extremely important. And the results will be used further when we discuss module 3 and module
4. So that is all for now thank you.
269
Lecture - 28
Interpretation Triaxial Test - UU UCS
Welcome back all of you, in the last lecture, we have seen the different possibilities of triaxial
testing and what are the different type of tests that we can execute in a triaxial cell. In today's
lecture, we will learn how to interpret these test results. As we have seen there are 3 major type
of triaxial testing, UU unconsolidated undrained, CU consolidated undrained and CD
consolidated drained in today's lecture, we will see the interpretation of unconsolidated
undrained test.
So, it iSuU test we have already discussed that UU test is a quick test and it is completed mostly
by in 10 to 20 minutes you may refer to Holtz and Kovacs book because much of these
discussions I am referring to that it is also known as quick test. Let us first see, what are the
different the stress condition in the soil sample for UU test the initial condition pertains to the
soil sample once it is sampled from the field.
Referring the figure, let us consider the sample soil which needs to be tested for UU we have
already discussed about the effect of sampling in our previous lecture. And what did we discuss
there what we have told is that during sampling the confinement and the overburden stresses
gets released and we have tried to compute what is the pore water pressure that would be there
in the soil specimen now, that information becomes quite handy for this particular lecture.
270
Because, when you sample the stresses are released and hence, we have the lateral as well as
actual stresses to be 0 and this is immediately after sampling. Now, what is the pore water
pressure whether pore water pressure will be 0 or it will be something else? We have already
seen for a typical normally consolidated case with A value of 0.5 in our previous lecture, we
have seen the final pore water pressure will be always less than the pore water pressure when
it was there in the field that means it tends to become negative.
So, here I am considering it to be negative because it is lesser than the initial condition due to
release of the confinement the pore water pressure tends to be negative. Now, this is represented
as -ur as negative pressure after sampling and thiSur is known as residual pore pressure. So,
thiSur or the - ur is instrumental in preserving the sample or holding the sample intact this also
we have discussed in the previous lecture.
Now, thiSur what will be its value? It will be a function of initial state and initial state when I
say it also pertains to w that is the gravimetric water content and e which is the void ratio. It is
also a function of a third quantity, which we have already seen, which is stress history, because
A value is dependent on stress history. So, what will be the effective stress total stress minus
′
pore water pressure. So, we have 𝜎30 that is all round stress will be equal to ur. So, this is the
initial condition of the sample.
271
Now, this sample need to be saturated first so, the very first stage in triaxial testing is saturation.
And we use the B parameter we have seen all these things. Let us say that the sample is fully
saturated now. So, the next stage is confinement so, what will be the stress changes, that
happens to the sample? So, to the sample we apply the cell pressure 𝜎3 . Now, because of this
what happens?
The pore water pressure will change because in UU test it iSunder undrained condition. So,
pore water pressure will be −𝑢𝑟 was there the residual pressure pore water pressure which was
already there in the sample, it was already there, so, −𝑢𝑟 + 𝑢𝑐 now, +𝑢𝑐 comes from where
now, this +𝛥𝑢𝑐 is comes from because of the cell pressure application.
So, the pore water pressure after the application of confinement is −𝑢𝑟 + 𝛥𝑢𝑐 , now, 𝛥𝑢𝑐 comes
from where it is because it is associated with the application of cell pressure. So, we need to
use the pore water pressure equation, so, it will give the same value, what is the pore water
pressure equation? That is
𝛥𝑢 = 𝐵[𝛥𝜎3 + 𝐴(𝛥𝜎1 − 𝛥𝜎3 )]
𝛥𝑢 = 𝐵𝛥𝜎3
𝛥𝑢 = 𝛥𝜎3
So, I am writing this just to make you understand how handy the pore water pressure equation
is now. So, that is how it comes here −𝑢𝑟 + 𝑢𝑐 , so it becomes −𝑢𝑟 + 𝛥𝜎3 and 𝛥𝜎3 is the
confinement applied. So, what will be the effective pressure?
Effective pressure will be 𝜎3 − 𝑢𝑟 + 𝜎3 = 𝑢𝑟 . So, the effective stress is 𝑢𝑟 now, so, then
comes the shearing stage. So, without opening the drainage wall stress is applied or the actual
load or the actual stress is applied and that is equal to 𝛥𝜎 which is the deviatoric stress
component that causes failure and what will be the pore water pressure? So, it will be − 𝑢𝑟 +
𝜎3 ± 𝛥𝑢. Now, this 𝛥u is because of the result of 𝛥𝜎 now, why it is ±𝛥𝑢?
Because, we have already seen that for depending upon the stress history, whether it is a
normally consolidated or an over consolidated soil in terms of volume change during shearing
over consolidated soil will tend to dilate and normally consolidated soil tend to compress now,
it is an undrained condition. So, we have seen that accordingly the pore water pressure would
change.
272
So, dilating soil will have negative pore water or pore water pressure will start reducing and
tend towards negative whereas, compressing soil like NC that will have positive pore water
pressure. So, that is the reason why we have plus minus 𝛥𝑢. So, it is simple then the effective
stress will be this 𝜎3 + 𝛥𝜎 + 𝑢𝑟 − 𝜎3 ± 𝛥𝑢.
So, ultimately here it will be 𝜎3 gets cancelled off and the final expression will be 𝛥𝜎 + 𝑢𝑟 ±
𝛥𝑢. So, this is the different stages of UU test and what will be the kind of stresses in the soil
sample.
Now, how do we do the UU test we need to have at least 3 identical samples, for example, we
want to find out what is the undrained strength at a particular depth. Now, from that particular
depth we take it we sample or during subsurface investigation we collect undisturbed samples;
which corresponds to that particular point. So, at least 3 identical samples are needed so, that
iSused for determining the undrained shear strength. Now, all the 3 samples are assumed to
have initial state same that is dry unit weight water content and hence the void ratio.
Now, this is not really practically possible to have exactly same, but we assume that these 3
identical samples are in fact identical in terms of its initial state, because this assumption is
very important for understanding the strength the moment it changes, the strength aspect also
changes. Now, this is the condition let us say these are the 3 identical samples to which we are
applying 3 cell pressure 𝜎31 first, then 𝜎32 , 𝜎33 , where 𝜎31 < 𝜎32 < 𝜎33.
273
Now, at failure, I am not going to the details of it after the test is conducted, we have the failure
stresses now, which is deviatoric stress, which is 𝜎11 − 𝜎31 𝑜𝑟 𝜎13 − 𝜎33 so, this is the
deviatoric stress at failure. So, this is at failure condition and we have already discussed that
there is no need of measuring pore water pressure, we will come to know in a few slides like
why this is valid, but still I am marking that the pore water pressure is 𝑢1 , 𝑢2 𝑎𝑛𝑑 𝑢3 .
So, what happens like if you plot these results like 𝜎1 − 𝜎3 versus axial strain 𝜖𝑎 , there is
something interesting what comes out that is the final deviatoric stress at failure, it comes out
to be more or less same. So, for saturated soil, even though the confining stresses are different,
we have the same almost identical magnitudes of deviatoric stress at failure. Now, this has
some important significance, there is no change in the initial effective stress condition in the
confining stage.
Now, why this is so? Why we are having such a result? Like there is no change in deviatoric
stress, even though the cell pressure is increased, we know that we are not consolidating the
sample, it is an unconsolidated test. Hence, there is no change in the initial state of the sample
only when you consolidate, there will be changes in void ratio or water content. And hence,
the initial effective stress condition changes and we know that the strength of the sample is
dependent on effective stress conditions.
So, here there is no such change and hence, the deviatoric stress at failure is more or less same.
The soil sample is mounted and total stress applied under undrained condition. So, there is no
scope for consolidation in UU test.
274
So, what is the net result of this? What do you mean by same deviatoric stress and how does it
translate to failure envelope so, that is what we will see here is a failure envelope for UU test
here it is 𝜏 versus σ plot. And now, we will see that sets of experimental results we have
plotted in terms of Mohr circles. Now deviatoric stress strain is same means it means 𝜎11 −
𝜎31 is same so, you have same diameter Mohr circle.
So, this is the first one, this is a second one and this is third one, all of them have same diameter.
So, this means what is changing then? 𝜎31 does the initial point is changing, because there is
a Mohr circle. Now, 𝜎32 has been increased now, it starts from here, but the deviatoric stress
or the diameter of the Mohr circle remains same, same with the third case.
Now, for the first one, we have the pore water pressure measurement, if you find out the
effective stress circle, how will you do just minus the pore water pressure we will get the
effective stress circle what is happening there are 3 total stress small circles, but there is only
1 effective stress circle. So, that is what we see from this particular figure. So, what will be the
failure envelope, the failure envelope is basically tangential to the Mohr circle. Now here the
only possibility of tangential this thing is a horizontal line.
So, this horizontal line represents the Mohr failure envelope, but this is very specific to
undrained condition. So, this is purely a total stress analysis, there is no effective stress
envelope there is effective stress circle, but there is no effective stress envelope when will you
say that you have an effective stress envelope? Only when we have 𝜏 versus σ and if there is
275
multiple effective stress then I can draw a failure envelope but here 3 total stress circles have
resulted in only 1 effective stress circle.
So, if you cannot draw a Mohr circle failure envelope, we can draw Mohr failure envelope only
if we know what is the angle of internal friction? So, there is effectively no effective stress
envelope possible from UU test and some of you would have understood this, concepts very
well during your UG but it is normally observed that these aspects are not fully appreciated
during your UG learning.
So, it is very important to understand what is the significance of UU? So, that is why I am
taking time and explaining it in detail for some this may not be needed. The Mohr circle at
failure have same diameter I have already explained only 1 effective the Mohr circle at failure
I have explained what is the reason because there is a corresponding increase in the pore water
pressure for σ31 there iSu1, σ32 proportionally it is increasing.
So, by u2, σ33 proportionally it is increasing to u3. So, there is no scope for any sort of changes
not possible to get Mohr column failure envelope in terms of effective stress. So, there is no
use of u measurement in UU test now, possibly you will appreciate this point why there is no
need of pore water pressure measurement, even if you measure you are not going to get any
effective stress envelope.
This test giveSundrained shear strength exclusively the value of Su now, there are several
textbooks in which you will find it is denoted aSundrained cohesion. In fact, if you understand
this aspect, that is that from UU test what you get iSundrained shear strength, there is no harm
even if you use the term untrained cohesion, because it is very difficult to comment on the kind
of cohesion that exists.
So, here it is, in general, it is better to use the terminology undrained shear strength and which
is represented by Su and this Su is what you obtain from UU test and which iSused for specific
design applications. So, Su is what? Su = qu which is equal to σ1 - σ3 / 2. So, this we need to get
the radius of the Mohr circle and that is the Mohr failure envelope.
So, σ1 - σ3 / 2 effective stress at failure is independent of total cell pressure applied we have
seen there is no effect of σ31 σ32 σ33 these statements denotes the same thing, but why I am
276
stressing these statements again and again is that these are different ways by which you tell the
same thing so, that you need to understand it carefully. Since it is completely undrained the
friction cannot be mobilized.
So, there is no concept of 𝜙 coming into picture so, total stress envelope is 𝜙 = 0. Hence, this
test is also known as 𝜙 = 0 test. So, undrained shear strength is independent of normal stress
which is constant limiting shear strength, thiSundrained shear strength Su is known as constant
limiting shear strength now, what do you mean by this statement at various normal stresses,
you have the same value of Su so, that is why it is independent of normal stress.
Now, this part many of you would have understood, but what is the significance of this 𝜙 = 0
failure envelope. Now, this condition iSused by Tresca failure criterion or it is known as
maximum shear stress theory because you are considering the maximum shear stress point. In
fact, the soil when you consider the effective stress conditions or various other triaxial test
results you will see that it does not fail at maximum shear stress conditions only for undrained
condition.
The failure is associated with maximum shear stress that emerges in this soil. So, Tresca yield
criteria for soil is based on the undrained shear strength Su. So, σ1 - σ3 = 2 Su so, when the stress
condition approaches Su it is found that this soil would yield for dealing with any undrained
shear problem in soil mechanics, the Tresca criterion iSused. This equation is valid for total
stresses even when the shear strength is governed by effective stresses why? What is the
importance of this statement?
277
This equation means σ1 - σ3 we know that even though there is pore water pressure or not,
whether it is effective stress condition or not σ1 - σ3 = 2 Su will be valid, we know that shear
strength of the soil is exclusively determined by its effective stress condition. So, even if I write
𝜎1 − 𝜎3 = 𝜎1 ’ − 𝜎3 ‘ so, it is not violating in any form that when whatever our understanding
is that the shear strength of the soil is exclusively dependent on effective stress conditions.
So, this does not violate this for an expected drain undrained condition that is a short term
response shear strength of the soil is equated to Su what we need to understand is that, if we
know that we are doing a short term analysis and the conditions are undrained it is that we
equate the strength of the soil to this undrained shear strength Su this is more relevant for end
of construction phase where undrained condition may be relevant.
Su is dependent on initial condition we have already seen this void ratio and water content
hence, it is not a fundamental soil property. It exhibits a wide variability you can see that it is
approximately 0 for soft clays, to MPa range for stiff soils depending on the initial state. Now,
if you considering the actual field problems, then Su at a site is normalized with respect to
effective overburden pressure, which is equal to 𝜎𝑣0 ′ it is effective overburden pressure why?
Because, as the overburden pressure increases, we know that the void ratio or the water content
would change. Now, we get we will get multiple values of Su which was influenced by the
overburden conditions. So, that is why it is normalized with respect to effective overburden
pressures, so, that comparison will be easier. So, this normalization with respect to 𝜎𝑣0 ′ helps
in better comparison of data.
278
Now, we have talked it is quite understood that whatever we are discussing now is with respect
to saturated soil. Now, what would happen if there is improper saturation for the soil samples
so, that is what we will see in this slide. Partial saturation results in air present in the volume.
So, this is also known to us when σ3 is applied now, when there is air present what will happen
as you apply the σ3 air tends to dissolve into the water and hence the void tends to close.
So, what we are doing for an expected undrained test there is going to be a volume change now,
this is against what we want, because for a truly undrained condition, we expect that there is
no volume change, but if there is presence of air then volume change is bound to happen. So,
if you see that this is the cell pressure application, so, if air is present, what will happen there
will be volume change.
So, there is an unexpected or hidden volume change in an undrained test where we assume that
there is no volume change. So, volume change occurs under undrained conditions, unlike the
saturated case, this alters the initial state of the soil. Now, once the initial state alters, then that
will impact the strength of the soil and the initial effective stress condition why? Because, there
has already been a volume change now, the deviatoric stress which was same at failure for a
saturated system.
Now, it will increase with σ3 because as σ3 increases, the amount of volume closure also
increases till the entire air voids is removed. So, that will influence the deviatoric stress at
failure which keeps increasing. So, saturation will increase with decrease in e as the voids close
279
when the air gets dissolved into the water or expelled out what will happen the saturation also
will increase.
Now, as the volume change increases, saturation also increases and at some σ3 saturation
approaches the value of 1 and the results will be then close to a typical UU test. So, what will
happen as the confinement goes on increasing the amount of wide closure also increases
saturation increases at some point of time saturation approaches one. So, then the results what
you will be getting will be identical to the saturated UU test.
So, that is the impact of saturation let us see what is happening now, this is σ31, where some of
the voids gets closed because of the expulsion of air σ32 again some volume closes and you can
see that the deviatoric stress keeps on increasing now deviatoric stress keeps on increasing
means diameter of the Mohr circle keeps on increasing now, up to σ34, there is a increase in the
Mohr circle now, beyond that, the Mohr circle diameter, more or less remains constant.
So, what would have happened now, this is a typical test of UU, but the soil sample was not
initially saturated. So, instead of getting a horizontal failure envelope, you are getting a curved
failure envelope and at some confining stress, it becomes same as ϕ𝑢 = 0. So, this is the
value of the undrained shear strength. So, what is this initial portion, this initial portion is
mainly due to the partial saturation or improper saturation.
So, here beyond this line, it is Sr less than 100 percentage and for this it is Sr = 100%. So, from
here on, it is 100% saturation and hence, we get the same failure envelope ϕ𝑢 = 0.. So, σ31
280
to σ34 causes volume change due to confinement beyond σ34 results are identical to typical UU
test σd changes with σ3 and beyond σ34, σd remains same.
So, this is again I have redrawn it here to show you important aspect which we have to keep in
mind, when we know that we are conducting UU test. So, if one gets such a pattern of Mohr
circle or maybe most of the time, what we do is we have only 3 or 4 repetitions or identical
samples tested. So, if it is 3 or 4, then what happens is 1, 2, 3 we will be getting a curved failure
envelope, but we know that it is a UU test result.
So, if we know it is UU test, then we know that we are not expected to get a curved failure
envelope it has to be a horizontal failure envelope. So, unless we appreciate this point, when
you go through a test report, then it is difficult to make out this point. So, always keep in mind
that if it is a UU test, then 𝜙 = 0. If you find that there is a curve failure envelope which is
tangential to the Mohr circle, then the problem lies with the saturation.
So, this is what I told curve failure envelope is not expected it may be due to improper
saturation. So, we need to discard the initial portion, either repeat the test or discard the initial
portion or take an average value of Su and definitely if you are considering the initial portion
where it is curved, this particular portion if you are averaging it out, definitely you will be
having a lesser value of Su. So, it is always better to investigate why and what has caused
curved failure envelope in UU test.
281
Now comparison of UU test for NC and OC state. So, you can see that 𝜏 versus σ is a total
stress circle 𝜙 = 0. Now, because of positive pore water pressure, you will have effectives to
circle shifted towards the left and this is a typical case of NC state that is normally consolidated
soil normally consolidated soil during shearing will have positive pore water pressure.
So this will shift the effective stress circle towards left and the same for over consolidated soil
u will tend to be negative and hence the effective stress circle shifts rightward. So, these are
the 2 comparisons how the stress history would influence the UU test results.
Now we have almost finished the UU test discussion and its interpretation we will see what
happens in unconfined compressive strength. Now this is a special case of UU test that is why
I am discussing along with UU test. So it is a special case where σ3 = 0. So this is the condition
where σ3 = 0 and σ1 is the unconfined compressive strength, this is what I meant.
So, instead of starting from somewhere here, the starting point is at σ3 = 0 and this diameter is
σ 1 - 0 that gives the unconfined compressive strength, but everything else remains the same,
we have the same undrained shear strength. So, UCS = σ1 - σ3 and Su = σ1 - σ3 / 2 rather σ1 / 2
in this particular case, specimen to be homogeneous and intact soil must be fine grained and
negative u exist I mean to say the negative residual pore water pressure is there otherwise this
soil will not remain intact.
So, it is applicable UCS is applicable for these situations loading should be quick to maintain
undrained condition because this considered to be an undrained test. So, Su obtained from UCS
282
and UU are same for identical soil samples and conditions. So, there are cases where you may
not get identical test results of UCS and UU depending upon specific conditions, but in practice
or in general, we always consider UCS test and UU test for this same initial state of the sample
will be more or less identical.
Stress condition in UCS same as that of UU test let us see that this is the initial condition ur the
same as that of whatever we have discussed. Now, axial loading state there is no confinement.
So, it is directly axial loading stage we apply Δσ and that is nothing but σ1. So, what will be
−𝑢𝑟 ± Δ𝑢 which is coming from delta σ why this is considered to be an untrained test?
So, please do not forget that particular aspect, you will be wondering where is the pore water
pressure coming from, but it is there because it is a quick loading test. So, what is the effective
stress condition it is −𝑢𝑟 ± Δ𝑢, you can see that if you compare this effective stress with that
of UU you will see that this is same. So, that means that we are going to get the same strength.
283
Now, what is the relevance of undrained shear strength is relevant for cases where loading is
rapid and undrained condition prevails Su which is the undrained shear strength to change in
total stress during construction, it is not going to influence the whatever is the strength that is
available and of construction is most vulnerable phase where undrained condition and u will
be highest.
So, UU shear strength is more relevant for such cases when there is end of construction phase
this is very easy to understand, because we have already discussed this total stress variation
with time in one of our previous lectures. So, this is their quick construction of embankment
over a soft clay it gives Su now, this Su is there for a particular state of the soil sample. And
that is not going to change with the construction sequence that remains the same only aspect is
if it is a quick construction.
Then the undrained condition prevails and Su becomes active or Su becomes more prominent
short term load capacity of piles is also based on the total stress analysis where Su is relevant.
I hope you remember in foundation design for piles, there is a short term analysis, where we
consider that the soil exhibit and undrained behavior in that case, what we have is the undrained
shear strength Su so, for designing of piles or determining the pile load capacity one need to
know the value of Su.
So, that is what is shown here undrained modulus now if you want to determine the elastic
settlement or the cube settlement, what that happens immediately after loading we need to have
284
what is known as undrained modulus it may be initial or secant modulus it is obtained from
UU test and use for determining the immediate settlement.
So that is all about the interpretation of UU and UCS. So let us try to summarize what we have
learned in this lecture. UU test is a quick test pore water pressure is not monitored and
measured. The reasons we have stated sampling of soil in the field induce negative residual
pore water pressure that holds the soil intact that is minus ur for soil samples with same initial
state.
Different confining stress results in the same magnitude of deviator stress at failure for identical
samples there is no change in the initial effective stress condition in the confining stage. So,
even if you apply σ3 that is not going to affect the initial effective stress condition and that
results in the same deviatoric stress. UU and UCS are total stress analysis which gives
undrained shear strength Su.
Su is independent of normal stress and that is what the Tresca criteria. Tresca criteria is based
on Su, Su exhibits wide variability, it depends on initial condition and it is not considered as a
fundamental property of soil.
285
Impact of partial saturation and what is the influence of stress history on UU test result is
discussed. UCS is a subset of UU is a specific case of UU under identical conditions and soil
samples UU and UCS is expected to give same Su effective stress at failure is same for UU and
UCS for the same or the identical soil sample and of construction phase, which is essentially a
short term undrained and short term analysis of pile load capacity, request the knowledge of
Su.
So, that is all for the interpretation of u u test results. In the next lecture we will continue with
the interpretation of other triaxial tests, which is essentially CU and CD so that is all for now.
Thank you.
286
Lecture - 29
Interpretation Triaxial Test - CU
Welcome back all of you so, in the last lecture, we have been discussing about the interpretation
of triaxial test and we have finished UU test and UCS. In today's lecture, we will see the
interpretation of consolidated undrained CU test.
(Refer Slide Test: 00:45)
So, let us start the consolidated undrained CU test. Now in this as different from UU test we
have to understand that the consolidation at every confinement alters the initial effective stress
condition, what does this mean? We need to understand because while learning this we might
not have paid much attention to what is the implication of consolidation? Now, what happens
during consolidation we are conducting the test at least maybe 3 different confinements of σ3
values.
And at every σ3 for CU test, we allow drainage to happen and the consolidation to occur. Now,
this consolidation alters the initial state of the sample, which was not happening for UU test
and when the initial state alters definitely the strength of the soil will differ. Now, what is the
significance of this statement? When the strength of the soil differ is in terms of Mohr circle,
the diameter will keep changing, which was not the case for UU test that is the significance of
what is meant by consolidation at a given cell pressure.
287
Now, the second part of this test that is the shearing part is at undrained condition. Now, during
undrained condition, pore water pressure will develop and the same is measured in this
particular test. So, due to this, u measurements we can present the result in terms of both total
as well as effective stress. Now, just like we discussed about the stress condition in the soil, for
UU test we will see what is the stress condition in CU test.
So, first is saturation stage again for each of these tests, we will repeat these because for
completeness and for clear understanding. So we need to ensure that the B parameter is greater
than 0.95 to ensure that it is fully saturated. So the next stage is consolidation now consolidation
can be brought about or saturation can be brought about by the application of back pressure as
well and we may also not opt to take the back pressure.
So there are 2 conditions of consolidation where back pressure is used where back pressure is
not used. So first let us see a consolidation stage without back pressure. So, in the figure for
confinement application a particular sample is denoted with is σvc that is the vertical stress for
confinement, c denotes confinement and σhc is the horizontal stress for confinement that is σhc.
Now, there can be isotropic or anisotropic consolidation in general, conventional triaxial testing
we always opt for isotropic consolidation because this is more simple and easy to understand.
Anisotropic consolidation would also induce some sort of shearing because there is a difference
in stress. So, here we are talking about a condition where σvc = σhc which is an isotropic
consolidation.
Since it is a consolidation stage, drainage valve will be opened and we do not expect any pore
water pressure to develop. So the pore water pressure will be equal to 0 because of
consolidation stage. If there is some residual pore water pressure, we need to account for that
will be the initial pore pressure that is soon after it is mounted. So, in this consolidation stage,
since the drainage valve is opened, pore water pressure is equal to 0 and hence we have σvc =
σ’vc and σhc = σ’hc. So, that is consolidation stage without backpressure.
288
Now, let us see the consolidation stage with back pressure. So, whatever we are doing this
isotropic consolidation this may not be actually realistic in the field because in the field always
an isotropic consolidation happens because you have an overburden because of this overburden
will there will be lateral earth pressure lateral pressure is not always equal to vertical pressure.
Hence, what we are adopting in triaxial testing we need to understand that there is some
deviation from the reality and hence, but they still go ahead because this is very easy to
understand and follow and there is no shear which gets induced. So, total stress what did we
understand in backpressure? In backpressure both the cell pressure and the back pressure is
increased simultaneously. So, it is more like a static pore water pressure that is what is denoted
here.
So, σvc is raised by u BP and σhc is also raised by u BP and but here it is isotropic consolidation
and the pore water pressure within the soil is also increased by back pressure. So, u BP will be
the back pressure pore water pressure so, what has happened u BP is increased outside and
there is an increase in inside also, so, net effect is 0. So, effective stress is σ’vc and σ’hc. Now,
in the previous case also that is when back pressure is not applied, there also the σ’vc and σ’hc
is the same.
So, we are not altering the initial effective stress condition. So, once the consolidation is
completed, that means, there is no more drainage of water happening then the next stage is
shearing stage and this is done under undrained condition. So, here we are increasing by Δσ
289
that are the stress difference or whatever is the actual stress which is imparted to the sample
that is Δσ. Now, this results in an increase in or increase or decrease there is both positive and
negative of pore water pressure during shearing stage.
So, during shearing stage there is either positive or negative pore water pressure. Now, why
this positive or negative pore water pressure we can see that it depends on the tendency of the
sample to contract or dilate let us say that there is a volume change which is permitted for this
particular sample during volume change, what happens to this soil during shearing we have
seen this in our previous lecture that this sample will have a tendency to either compress, or it
to dilate.
So, that is the volume change in this case volume change is not permitted so, that translates to
pore water pressure. So, a dilating sample will try to exhibit negative pore water pressure or
decrease in pore water pressure. Whereas, a compressing sample at low sample which tends to
compress will achieve positive pore water pressure because drainage is not permitted it
translates to pore water pressure. So, hence, the effective stress will be σvc + Δ σ ± Δ u.
So, the sign depends upon the behavior during shearing similarly, for σhc so, this is the stress
condition of the sample during consolidation and during shearing stage. Here there is also
important aspect which we need to understand which we know but it is better that we confirm
this again and again. So, σcell = σvc = σhc (isotropic consolidation) = σ’1c= σ’3c provided pore
water pressure is 0 and σ’3c = σ3f that is only in terms of total stress.
But since the shearing stage is undrained there will be pore water pressure that gets generated.
So, definitely final cell pressure is going to be different that is final effective cell pressure that
is σ’3f, which is σ3f ± Δ u. So, that is why this is not same and the final deviatoric stress at
failure will be Δσdf = (σ1 - σ3)f.
290
Now, let us see the typical stress strain response of CU test. Now, we have already discussed
in our previous lecture about a typical stress strain response. Now, there is nothing different
from that, when you consider UU, CU, CD test, but it is important for every stage of the test
that we repeat this and reaffirm that this is what is the stress strain response? Why this is
important is this knowledge is very much important for your subsequent modules.
So, there is no harm if we hear this maybe 1 or 2 times more, so, that, things and the concepts
will be pretty clear. So, here it is deviatoric stress versus volumetric strain and a typical OC,
NC and LOC curves are given this we have seen previously here it is plotted in terms of
volumetric stream, one can do that in terms of actual strain as well and volumetric strain is
𝜖𝑣 = 𝜖𝑎 + 2 𝜖𝑟 . So, what is happening at typical OC has a peak and NC does not exhibit a
peak and here it is strain softening here it is strain hardening.
Now, in this case, instead of volume change, we have changes in pore water pressure and you
if it is under back pressure, then the port water pressure starts from the back pressure and for
NC there will be positive, so, this will be positive and that is NC and for OC after a small initial
adjustment, it will get to negative pore water pressure. So, that is what is going to happen it
may reach negative, but what it means is that it is reducing the pore water pressure is reducing
from back pressure.
So, depending upon the amount of OCR that means, how over consolidated the soil sample is
maybe it is heavily over consolidated. So, definitely it may reach to a negative pore water
291
pressure. So, this, the same figure is plotted in terms of principal stress ratio. This also we have
seen before as a very important aspect in terms of maximum stress obliquity, so, σ’1 / σ’3 versus
ϵv.
It gives the same pattern as that of the deviatoric stress pattern only thing is the peak it may not
be exactly at that particular point because of σ’1 / σ’3 please note it is in terms of σ’1 / σ’3 that
is effective stress ratio. Here in the case of deviatoric stress ratio σ’1 - σ’3= σ1 - σ3 it is same.
So, in the ratio maybe there will be some difference here and there.
So, that is why the trend is same, but it may not be exactly the same here it is starting from 1
why because it we are starting from an isotropic stress condition. So, OC exhibits softening
behavior NC exhibit strain hardening Δu increases for NC and decreases for OC principal
effective stress ratio exhibit similar trend like deviatoric stress.
Now, let us see the typical Mohr circle for CU test in the case of NC and OC. So, for the NC
case, the total stress circle that is at the end of the test like once the test is completed, we obtain
the failure stress then we try to plot the Mohr circle. So, the total stress circle is indicated and
a failure line for total stress is indicated that is ϕT. So, they have already measured pore water
pressure and measurement of pore water pressure is invariably necessary in the case of CU test.
So, once we know u f that is a pore water pressure at failure, one can plot the effective stress
circle so, this is the effective stress circle. Now, if we draw the failure envelope that is given
by ϕ’ now, what is the implication of the two tangent lines? Let us see, based on failure stress
292
obtained from triaxial tests, we have plotted this plus u, because you shift the effective stress
circle towards origin.
Because we are minusing from the total stress, same diameter for effective and total stresses,
because σ’1 - σ’3 = σ1 - σ3 MC failure envelope passes through origin now, this is a very
important information, which you have to keep in mind, let us say that you are dealing with
normally consolidated clays and if we have certain wrong notion that clays are always cohesive
material, then we tend to question the fact that why there is no cohesion and there is only
friction which may be sometimes difficult to understand.
We need to understand that all soils are granular in nature and it can mobilize friction. Now,
what happens to a normally consolidated material? A normally consolidated material it can
undergo fast volume change and here in fact, it is in terms of pore water pressure because
volume change is not restricted. So, you tend to have only frictional component for normally
consolidated clays and we do not tend to have cohesion for this particular soil and hence, the
failure envelope of CU for NC passes through origin.
So, that we need to keep in mind and ϕT is less than ϕ’. Now, let us try to understand whether
this is really important or not whether ϕT that is the frictional angle corresponding to undrained
condition is it relevant or not. In fact, it is not relevant the reason is Mohr Coulomb failure
envelope is always defined in terms of effective stress. So, only this particular failure envelope
is important and we generally do not use ϕT because the undrained strength is different.
Now, we will keep repeating this you will hear the statements at least 4, 5 times throughout
this module and we need to clearly understand this particular fact. So, for OC, the total stress
circle is given as follows and ϕT is drawn and you can see that it does not pass through origin.
So, here since the soil has the tendency to dilate the pore water pressure will be less so, it will
tend to be negative and hence, we will have effective stress circle moving towards the right and
the failure envelope drawn as ϕ’.
So, u decreases and may become negative shifts the effective stress circle towards right same
diameter same explanation. MC failure envelope does not pass through origin for OC which
means that the cohesive component may exist now, this may be due to cementation this may
293
be due to apparent cohesion, it is not very clear since the particles are already stiff, because it
is over consolidated and the stress which is applied is less than the over consolidation.
The pre consolidation stress the material tends to act like a stiff material and the particles are
sufficiently close. So, this would exhibit in terms of some sort of cohesive intercept that you
get. So, for OC cohesion may exist, and ϕ’ would decrease. Now, ϕ’ decrease means in terms
of NC and OC do not think that OC will have a lower strength it is not like that. So, when you
plot the envelope for both, then we will try to we will be in a better position to understand. So
for OC question may exist ϕ’ would decrease that is ϕ’ is less than ϕT, again ϕT does not have
much meaning.
Mohr circle for CU test for a wide range of normal stress. Now just we will try to understand
what is the kind of Mohr circle when we consider a wide range of consolidations stress or
normal stress? So, τ versus σ: Let us first draw the pre consolidation pressure σ’p is the pre
consolidation pressure, which divides the behavior into OC and NC. So, let us say the total
stress circle for NC, which gives ϕT and this passes through origin effective stress circle.
And the envelope the failure envelope, which is denoted by ϕ’, this also passes through
origin. Now for what is 𝛼𝑓 ? 𝛼𝑓 according to Mohr failure hypothesis is the inclination of
failure plane which is nothing but 45 + ϕ’ / 2 and that is as per the Mohr failure hypothesis.
Now, total stress envelope for OC is T and if you draw it, you can see that there is a kink or
there is a discontinuity at this point P.
294
Now, this point P is roughly 2 times σ’p for typical clays and this information is obtained from
Holtz and Kovacs. The original paper is different. If you are interested to find out more, you
can always refer to this book and cross refer the literature. The effective stress circle for OC is
drawn and the failure envelope is drawn in both these cases there will be sort of cohesive
intercept.
Now failure again, this part is I am telling this again repeated and I am intentionally repeating
these statements. So, that it is pretty clear as to what we tend to understand. So, and this
understanding again it is important for accumulating your further understanding. So, failure is
fully governed by effective stress. MC failure envelope is valid only in terms of effective stress
that is long term, short term strength is only in terms of undrained cohesion.
So ϕT does not have much relevance with respect to Mohr Coulomb failure envelope. Now,
what is the relevance we will not talk about ϕT much the reason is for failure defining the failure
bound we tend to use only effective stress. And hence Mohr Coulomb failure envelope in terms
of effective stress only is important.
Now, what is the relevance of CU test it is used for stability problem where consolidation
already occurred and further developed an undrained condition. For example, there is a typical
case of an embankment construction. So, this let us say this is clay and consolidation already
occurred when you construct it this particular one, so is the kind of staged construction, the
entire height of the embankment is not constructed in one go first this part is constructed and
it was left for some time before the second part got constructed.
295
So, this particular clay would have undergone some amount of consolidation maybe it might
have undergone a complete consolidation depending upon the clay properties, then comes the
construction of 2 and this construction of 2 induces again a undrained condition provided clay
is saturated. Now, this is a typical case of CU condition some other conditions are rapid
drawdown of embankments, slopes of reservoir. Now when there is full of water and then there
is a rapid drawdown that is happening that induces more effective stress.
And this additional effective stress is going to cause an undrained condition within the core of
the dam. So I am not going to the details of it. I just need to understand that these conditions
do exist and it is close to CU conditions. And this is very true with clays where undrained
conditions are common. And other cases case where such a case can occur is when you are
constructing an embankment over a natural slope. So in case of certain developments, if it is
done, then this can undergo a kind of CU response.
Let us just try to see quickly what is the significance of what we have just understood? NC clay
is subject to CU test. The soil sample is consolidated under isotropic stress of 150 kPa. The
principal stress difference at failure is 100 kPa. What is meant by principal stress difference
that is deviatoric stress at failure pore water pressure at failure is 88 kPa. What is the shear
strength parameter for this soil?
Nothing is told about back pressure so, we will not get into that. But we will assume that the
pore water pressure is the final pore water pressure at failure and whatever it is it is all inclusive.
296
Now, based on effective stress principal, shear strength is governed by effective stress what is
given σ1 - σ3F is 100 kPa, uf is 88 kPa, σ3C is 150 kPa. So, this much is given and these are in
terms of total stress.
So, σ1f that is from this particular equation one can find out what is σ1f that is equal to 250 kPa
you are just adding σ3C to it and σ’1f is minus pore water pressure that is 162 kPa, σ’3f = 62 kPa
and for NC cohesion is 0. Now, think of a situation that we do not understand this particular
point like for NC there is no cohesive intercept then it becomes difficult to solve this problem.
So, now, we know that NC does not possess cohesive intercept.
So, that is equal to 0 if that is equal to 0 then one can always write the Mohr Coulomb failure
𝜙′
envelop σ’1f = σ’3f tan2 45 + . Substituting the effective stress condition into this equation,
2
we can get ϕ’ = 270 and that is plotted as show in the figure and the line touching the circle is
the bound now, only 1 Mohr circle is there if you have more or 3 Mohr circles as we do for the
axial testing.
And average line which is tangential to 3 of the Mohr circles gives the failure bound or Mohr
coulomb failure bound or failure envelope, we can also obtain the value of ϕ that is ϕT if you
substitute the total stresses in the same equation, which is not relevant, but if you substitute you
will get 140 and that is what has been plotted in the figure. Now, that is the differences u f for
design consideration 270 is useful and this ϕT = 140 may not be useful, I am telling in terms of
the failure criterion and the failures on the strength parameter.
297
So, CU test is a relatively quick test, which can be used to determine Mohr Coulomb failure
envelope shear strength parameters. So, what we are doing here is we know that for clays
drainage is poor and it takes weeks or maybe even months for completing a consolidated drain
test, if we are conducting it with a condition that pore water pressure does not generate. So,
practically it is not possible for conducting a drained test in clays whose permeability is less.
So, in that condition, we have CU test where it is still faster, why because the shearing stage is
undrained, but, there is a problem associated with it, which is like in CU test, the rate of
shearing is faster, I mean the second stage the rate of shearing is faster and which is different
from a normal field case or a CD case. Now, we also need to understand that the rate of shearing
considerably influences the strength. Hence, what we get from CU that is from CU we do the
test get the result in terms of effective stress.
This effective strength parameter closely simulates but we need to understand that it is not
exactly what is expected in the field. The shear strength parameters obtained from CU based
on u measurements may not represent the actual long term conditions in the field is exactly
what I just told. Because in the field it takes time in what manner it is getting loaded and in
what rate it is getting loaded. And in what rate the pore water pressure is dissipated.
And it gets to drained condition these things are quite, quite different from what we simulate
in the lab. And that is the reason why we need to understand like, we will use CU parameters
effective strength parameters. But we need to understand that there will be some sort of
298
deviation what we are trying to explicitly state is the uncertainties involved in these tests one
should not be. So, confident that we have got this particular value and this is the final it is not
so, a lot of inherent uncertainties associated with each of these parametric values.
Now for engineering purposes, we need a design value based on which we can design various
structures which are founded or maybe the design the foundation or various other applications.
So, we need to understand that these are representative values and not really true values and
because of these hidden factors involved and this has something to do with the loading rate.
And we have already explained this in actual condition for drain condition to exist the loading
rate will be very low in NC, u will be positive and in OC it will decrease.
And tend to negative value the Mohr Coulomb failure envelop passes through origin for NC
this is important and the Mohr Coulomb failure envelope exhibit cohesive intercept for OC soil
you please note that I have used the term cohesive intercept. So, this does not mean that again
it is a true cementation there is a kind of value which you obtain because of the OC condition
this is associated with dilation.
Mohr coulomb failure envelope is valid only in terms of effective stress which is long term
short term strength is only in terms of Su probably you will hear the statement once more when
we discuss the next interpretation which is ϕT. Because I find that or I understand that this is
very important to clearly get into your mind that undrained condition the strength is mostly in
terms of Su, ϕT that is the total strength parameter which you obtain from the total stress does
not have much relevance with respect to Mohr Coulomb failure envelope.
So that is all for the interpretation of consolidated undrained test. We will see about the
interpretation of consolidated drain test in the next lecture. I again, request the participants to
be patient, those who know these particular concepts very well and those who have been τght
these particular concepts very well. For them, it may look like it is getting repeated. But for
me, these particular aspects are extremely important for all the participants to clearly
understand this, so that the understanding of the subsequent modules will be easy. So that is all
for now. Thank you.
299
Lecture - 30
Interpretation Triaxial Test - CD
Welcome all of you in the last lecture we have seen the interpretation of CU test. Now, the last
one is consolidated drained CD test. So, we will see the interpretation of CD test in today's
lecture.
So, it is all about consolidated drained test CD test just like CU test consolidation at every
confinement alters the initial effective stress condition at least we have to have 3 samples for
obtaining Mohr Coulomb failure envelope and shear strength parameters. So, when we do we
will have at least 3 consolidation pressure or 3 confinement σ31, σ32, σ33 in increasing order.
So, at every consolidation pressure, the initial state of the soil gets altered, it undergoes more
densification and hence, the strength will also be different. So, that is what it means, and why
it is happening? It is happening because the initial effective stress conditions are getting altered.
Now, we need to keep in mind that the soil samples are initially consolidated in the range close
to field condition or design requirement.
So, when we apply σ3 for any specific field application, so, we have some clue about the
anticipated overburden pressure. So, the confinement in the field will be definitely σ’h which
300
is coefficient of earth pressure multiplied by effective overburden pressure. So, we know the
range what is expected in the field. So, in the lab also the confinement or the cell pressure that
we choose will be in line with this.
This we have to keep in mind when we are doing to triaxial testing for very specific projects
where it is extremely important that shear strength parameters should be close to reality. Now,
consolidation can be isotropic or anisotropic. That is what is meant by isotropic consolidation
that is what we normally do. Anisotropic means, we can also consolidate at some k0 value for
maintaining more realistic situation because isotropic consolidation as we discussed in the last
lecture is not actually realistic we choose it for simplicity.
Now, if you choose to have anisotropic consolidation that can also be done, but we need to
remember that during anisotropy, anisotropic consolidation, we are bound to induce some
amount of shear stress into the sample. But since the consolidation pressure or confining
pressure is not that high, that is not going to create too much of a problem and it is more like a
densification. So, it is not going to create much issue but during anisotropic consolidation, we
are bound to have shear stress developed within the sample.
No excess pore water pressure is expected throughout the test and that is how CD test becomes
a very slow test. When you load a saturated soil sample, there will be immediately an increase
in pore water pressure. Now, we have to do the loading in such a manner that during shearing
pore water pressure does not develop and this is a very hard task when it comes to low
permeable material. It is practically difficult to avoid pour water pressure.
So the loading rate will be excessively slow and that is how it becomes a very slow test. Similar
to UU and CU test where we have discussed about the stress condition in the soil sample we
will see or we will repeat it for CD test as well. Of course, the first stage is saturation stage it
has to be fully saturated which is ensured by the B parameter greater than 0.95 so first let us
see the consolidation stage without backpressure we have seen exactly same for CU test.
But it is important for completeness that we repeat this here as well total stress σvc and σhc and
for isotropy consolidation σvc = σhc pore water pressure equals 0 because there is no back
pressure. So, effective stress will be σvc = σ’vc, σhc = σ’hc there is no pore water pressure.
301
Consolidation stage with backpressure so, σvc = σ’vc + uBP similarly, σhc = σ’hc + uBP here it is
same isotropic consolidation we are raising the pore water pressure also by uBP. So, that there
is no alteration in the effective stress condition it remains same as σ’vc and σ’hc as we have in
the previous stage. Now, here when you apply back pressure, the back pressure line is kept
open for drainage as well and the drainage happens against the back pressure and that is okay.
We should not think that since there is back pressure that is going to oppose the drainage
coming out of the sample, this is mainly because, uBP is not acting like an excess pore water
pressure it is as I told in the previous lecture backpressure is more like a static pore water
pressure. So, any pressure over and above will drain will create a hydraulic gradient which
helps the flow from sample outwards so, that is not going to be a problem.
So, back pressure is kept open during the shearing stage as well so, uBP remains in the sample.
So, next comes the shearing stage so, here the deviatoric stress or the axial stress is applied Δ
u = 0 that is not considering backpressure, it is only for the shearing stage or the loading stage
there is no development of pore water pressure because that is the condition which is important
for this particular test.
Hence, σvc+ Δσ = Δσ’1 and σ’3 so, it is the same. So, there is no pore water pressure so,
everything remains the same. And here we have σcell = σvc = σhc, which is equal to σ‘1c that is
only for confinement c stands for confinement which is equal to σ’3c = σ3f = σ’3f please note
here it is equal why?
302
Because shearing stage there is no pore water pressure in the CU test this was not equal to σ3f
σ’3f why because in the final sharing stage they will be poor to pressure development. So, the
final effective confinement pressure will not be same as the initial the pore water pressure will
be there and the final deviatoric stress is 𝛥𝜎𝑑𝑓 = (𝜎’1 − 𝜎’3 )𝑓 at failure.
Now typical stress strain response of CD test, it is more or less same as that of what we have
seen for CU test there is nothing changed changing and that is true also because all these stress
strain responses are more or less general in nature. There are 2 classes as we have seen in the
previous lectures. If you consider granular materials, it is loose and dense and when it comes
to clay type of soils we have based on stress history, one is normally consolidated the other one
is over consolidated.
Now, loose sand and normally consolidated clays have same class in terms of stress strain
response, whereas, a dense sand and over consolidated soil falls in the other class which has a
specific stress strain response. Now, this is not going to change so, that is why it will look same,
you have the over consolidated, normally consolidated and lightly over consolidated please
remember the peak of lightly over consolidated is not ensured always.
It depends upon how much over consolidation it has undergone rather it depends upon OCR
over consolidation ratio, if the over consolidation ratio is close to 1 it may not exhibit peak, but
it is still ligtly over consolidated. So, this I will not explain much, it is all the same and it is
plotted with respect to ϵv one can always plot it with respect to ϵa as well. Now,
303
here, please note here it is ΔV, ΔV is the volumetric change. Now, in the CU test, the only
difference is instead of ΔV there it is Δu.
So, for over consolidated sample, there is an initial compression followed by dilation and for
NC there is progressive compression. For the same confining pressure, OC exhibits strength
higher than NC, let us say these are all conducted at same confining pressure you can see that
OC is exhibiting a higher strength OC also has a higher modulus if you see the slope this is
progressively reducing so OC has a stiffer material, because it is already subjected to that much
pressure in the past.
So it has more stiffness σdf occurs at lower strain in OC than NC. So, these are some very
minute important points which we have to understand the peaking of the OC happens at a lower
strain the peak happens because of dilation and the stiffness of OC material is more. So, all
these minute points are important and we have to keep in mind OC dilates and exhibit strain
softening I will not again get into those details, this has been repeated a lot of times.
So, you might be already knowing this NC compress and exhibit strain hardening and the
principles stress ratios just similar to what we have done for CU test, it also exhibit the same
kind of trend as we have got for deviatoric stress.
Now, typical Mohr circle for CD test for NC. So, remember it is CD test for NC yes, we have
seen and for NC, we need to keep in mind that the failure envelope it passes through origin. So
that is very important aspect which we have to keep in mind based on failure stress obtained
304
from triaxial test MC failure envelope passes through origin there may be deviation from
linearity, now, this appears to be pretty linear, we can also have kind of deviation from the
linearity as well.
Now, this may happen if the consolidation stress range is large, this aspect also we have
discussed in our previous lecture, as the normal stress goes on increasing there will be a kind
of restriction in the movement when there is restriction in the movement the friction mobilized
will be coming less and hence the friction angles start reducing. But, at the same time, since
the particles are coming closer to each other, it can exercise more cementitious effect and this
may not be a real cementitious effect.
But when the particles comes closer, there is a kind of apparent bonding that gets developed
because of which the cohesive intercept appears or it becomes more. So, that is what will
happen at a higher range or when we are considering a wide range of confining stress, it can
also deviate from linearity if the initial state is different. Now, when we are conducting the
shear strength test specific for a material at a particular place, we need to make sure that we
have identical samples, now what is meant by identical?
The initial state of these samples should be same and when initial state is same, it means that
the initial effective stresses are same. Now, if initial effective stresses are same, it is going to
have a comparable strength, but then the strength will be different depending upon the amount
of confinement that we give later so that is how it works. So the initial condition of the samples
should be similar or identical.
Now, if there is deviation from that, like we have a different water content or different density,
or even there is change in stress history, then that is going to create or that is going to give us
Mohr circle of unexpected diameter that means, it is not a proportional increase in the diameter
rather there will be changes now, this exhibits non linearity so that is what it means. Now, if
marginal c’ is observed, we know that the soil sample is normally consolidated we conduct the
test.
And we find that there is a marginal c’ that is coming we can easily ignore it provided we know
that the soil sample is NC we can neglect this marginal c’ and force the failure envelope through
origin. So, this c’ or the intercept can be considered 0 so, that is up to our judgement. So, there
305
are cases where you even though we know for NC it passes through origin it may so, happen
that it will not pass through origin and the very marginal c’ is observed so, neglect that.
Now, let us come to a typical Mohr circle for CD test for OC is this we have seen in the CU
test as well 𝜏 versus σ Mohr circle and the failure envelope gives ϕ’ and it also give a kind of
c’ the intercept Mohr Coulomb failure envelope does not pass through origin and how much
the intercept is that will depend upon the OCR how much the soil sample is over consolidated
if you remember the plot, we have seen this before.
So, this is for NC or critical state friction angle and we have already seen this response and we
have drawn like this. So, this comes for OC now, these points where it will fall this will depend
upon the OCR of the sample. So, if the OCR is high over consolidation ratio is high that means
that it is heavily over consolidated if it is heavily over consolidated the dilation angle will be
more than what is dilation angle this is the angle over and above the critical state angle.
Now, this is the critical state angle ϕ’c. So, if the over consolidation is more, let us say this is
the σn. Now, corresponding to this, if the over consolidation is more than what happens the
point shifts upwards. So, this dilation angle increases. So, OCR also determines the kind of c
that we are going to get for OC, cohesion may exist and ϕ’ would decrease as compared to NC
the ϕ’ is going to decrease.
Because it is the particles are pretty close to each other and hence the mobilization of friction
is limited by its movement but since it is close by it develops a kind of cementitious
306
characteristics, I will not say it is real cementation but a bonding is developed. So, that results
in a higher c’ then NC and lower ϕ’, now, low ϕ’ for OC does not mean that this strength will
be lesser than NC it will be still more but as you can see here, this is the OC part from here to
here.
You can see for every point the strength of OC is greater than NC, higher stiffness, higher
strength for OC, but the strength will depend upon the amount of cracks and fissures, when it
is over consolidated it is also likely to have more cracks and fissures depending upon the kind
of tensile stress that gets developed within the sample. Now, this will enable to have lesser
strength.
So, when we say higher strength, it is not necessarily always higher strength, the presence of
cracks and fissures also determine what kind of strength it will have the strength will depend
upon OCR and the confinement. Now, another aspect we discussed earlier is the kind of
confinement that we give. Now, when the confinement is more it suppresses the dilation so,
that also becomes very important.
So, that is why it depends upon OCR as well as the kind of confinement and confinement
determines the kind of dilation because confinement can suppress dilation for determining
strength parameter samples of identical OCR is needed and subjected to different σ’c. So,
sometimes when we explain based on this figure, there will be a sort of confusion on how the
OC parameters can be determined. Now, this remains true and but in this particular figure, there
are 2 things that is happening here.
For example, this particular point can be as a result of a particular OCR and which corresponds
to this particular confining stress or the normal stress. So, if I redraw this now, let us say that
there is an OCR corresponding to which a given confinement based on which I get a point here.
Now, for the same confinement, I have a different OCR, then or a higher OCR then the point
will go up because it exhibits more dilation.
So, that is what is going to happen so, one aspect is OCR. Now, if you want to determine the
strength of a given soil sample, it is ideal to have identical OCR. So, 3 samples of same OCR
so, then what is changing is the confinement now, for let us say this is a particular OCR now,
the same OCR at a higher confinement will give another point. So, this is what we draw and
307
we get c’ and ϕ’, so that is what it happens. So, when you read this particular graph and this
particular graph, we need to understand clearly how the shear strength parameters are
established for an OC sample.
For NC, ϕ’ from CU and CD may be comparable only if the rate of shearing is comparable,
this is very important. Now, why are we doing CU and CD we have already seen that CD is a
slow test and for most of the time for clay kind of soils we adopt CU for obtaining what the
effective stress parameters. Now, this effective stress parameters have to simulate or resemble
the long term condition that exists in the field which is in fact same as CD.
Now, a comparison of CU parameter the parameter which we obtained from CU, and the
parameter which we obtained from CD, if this, need to be comparable, then it is essential that
the rate of shearing also has to be comparable. Again this is not going to get fulfilled all the
time. Now, if the rate of shearing is going to be that slow CU test also will become slow. So,
we need to basically understand there will be some deviation from reality, that is the whole
purpose of telling here.
For OC and dense sand shear strength parameters may be higher in CD. Due to the actual
dilation the sample undergoes, what this means is that in the CD test, there is actual dilation
which is happening not that tendency to dilate and the particles have to do the work against the
loading the dilation is a phenomenon which is against the loading against the compression. So,
more work is done in this particular case.
308
And hence, that resembles in higher strength parameters in the case of CD. Volume change in
CD and u in CU it is also dependent on the sequence and manner in which the stress changes.
What it means is that when the shearing happens the volume change that happens because CD
is always associated with volume change and u is there in CU these changes is basically
dependent upon the kind of the manner or the sequence in which the stress also changes.
So, this also determines what kind of response we are going to get so, that we will understand
with the help of some examples. So, first example is loose sand subjected to compression. Now,
we can induce this compression in 2 different ways. The first way is case 1 that is σ1 increasing
σ3 constant this is the typical conventional triaxial testing, which you can see it here.
So, here the constant σ3, whereas, σ1 goes on increasing, case 2 is σ1 constant and σ3 decreasing
as you can see here σ1 is kept constant and σ3 is released. So, what is happening if you have a
cylinder which is acted upon by σ1 and σ3, where it may be equal in the beginning that is what
this particular point is the actual stress is kept constant and the σ3 is released.
So, when it is released, the tendency of the sample is to compress so, both action case 1 and
case 2 results in compression. So, both the cases results in compression now what is the kind
of volume change response that we obtain it is shown here sample can accommodate higher
volume change before failure in the case of case 1 for this particular case, the failure happens
a bit later the strain at which the failure happens will be delayed as compared to case 2 where
release of σ3 results in quick failure.
So, a higher volume change can be accommodated for case 1 as compared to case 2. Now,
please remember case 2 is more like a release and this is almost similar to an active condition
that is existing in the field. So, when there is a release the soil tend to expand or it goes away
from each other. So, it is a kind of same response that we simulate using case 2. So, we can see
that how the manner in which or the sequence in which the stress changes in these 2 cases will
result in a different kind of response.
309
Now, we will see another example, example 2 the earlier one was for granular soil sand, and
here we will take an example of NC soil for CD and CU under compression again, case is same
increasing σ1 σ3 constant, case 2 is σ1 constant σ3 decreasing, so, it is the same type of response.
Now, only important aspect that we need to keep in mind is in the case of comparison, both
CU and CD the shearing rate has to be same. Now, what will be the implication of this on the
overall test results so, for case 1 let us draw the case of CU first.
So, CU Mohr circle envelope ϕCU1 that is undrained case corresponding effective stress circle
is CU effective which is given by this now, this is a drained response this we have already seen
while discussing CU. Now, for the same sample we have conducted CD test, so, CD test is this
you can see that the strength is higher for the same σ3. Now, this will give us the drained
envelope and the failure angle that is ϕ’CD1 so, this is about case 1. Now, case 2 is σ is constant
and σ3 decreasing.
So, let us see how we can draw that is tau versus sigma. So, for case 2 first let us see CU and
we get the CU and this is the undrained failure and I will not say failure envelope undrained
trend or linear trend which is given by ϕCU2 the corresponding effective stress circle now, here
you have to understand in this case it has moved in the leftward direction because there was
positive pore water pressure.
Now, in this particular case, σ3 is getting released and hence, pore water pressure tends to
reduce and it tends to become negative, because it is a release that is happening for the soil
sample and that is the reason why it is shifting rightwards. And here σ1 is constant so, the
310
consolidated drained test envelope also should start from here. So that is what you can see that
the diameter of the CD is quite less.
Now that means is that the strength corresponding to CD is pretty less than the undrained case
for a constant σ1 case, why this is so we will explain in shortly. So the drained envelope or the
failure envelope ϕ’CD2 Mohr Coulomb failure envelope can be plotted like this that will give us
ϕ’CD2. So we have ϕ CU1 for case 1 ϕCU2 for case 2 ϕ’CD1 for case 1, ϕ’CD2 for case 2 now, let us
discuss this.
So, case 1, this is a typical compression test all of us know only this particular test. Now, for
the same σ3, CU Mohr circle is less than CD Mohr circle, this is quite expected case 2. Now,
case 2 is similar to active condition, we have already told this, this is case 2 for CU, as σ3 is
released, u will decrease hence strength would increase. So, that is why it is shifting towards
right for CD.
Now what is happening to CD as σ3 is released, what is happening is because of this release, it

has a tendency to draw water inside there is a kind of negative suction kind of thing developed
because of this release. Now this draws water in the case of drained enveloped or drained
testing, when water gets in, we know that the strength is going to be small or it is going to fall.
So that is what is happening and hence the strength would decrease and hence a smaller Mohr
circle corresponding to the same σ1. So, CD Mohr circle is less than CU Mohr circle.
311
Now, comparison of case 1 and case 2 the same envelope, which we are showing case 1. So
we have case 1 and case 2 so it is the same figure that we have drawn. Now we will try to
compare some important aspects of these 2 figures. The first comparison is ϕ’CD2 less than
ϕ’CD1 that is ϕ’CD2 the drained characteristics is less than that what we have from the normal
compression tests that we do in the lab.
So, this is a kind of release test whereas this is a compressive test that is having both are
compressive tests, but this is the general triaxial testing that we do now, note that this particular
point, what is the implication of this, we already know that case 2 resembles closely the active
condition. Now think of an active condition that exists in the field and we want to know what
is the shear strength parameter corresponding to that a sample is brought to the lab this
particular aspect is not clear to us.
And hence we do a normal compression test the one which is σ3 constant and σ1 increasing, we
end up with ϕ’CD1 this particular value or this particular value, but the actual condition of the
field is close to active condition and hence the available strength of that particular soil will be
ϕ’CD2. So, what we are doing is we are supplying a higher value of ϕ’CD1 compared to ϕ’CD2.
So, conventional triaxial compression gives ϕ’CD1 which is generally conducted and used. Now
for an active condition in the field, the available strength will be ϕ’CD2. So, if the design value
considered is ϕ’CD1 then a higher strength is considered then what is available in the field what
is available is ϕ’CD2. So, these are some important aspects which we need to keep in mind now
312
is this going to result in some type of failure may not be because we always give adequate
factor of safety to account for all such uncertainties.
So whether failure occur or not, is not the question, what is important is we need to perform a
triaxial testing, which is very close to the field situation. So that is the whole purpose of this
particular discussion. Now, at the same time, we also need to keep in mind triaxial simple
compression triaxial testing, it looks very simple, but it is really complicated. To get a very
good result from triaxial testing, there is a lot of pain involved and only those who continuously
work on triaxial testing will be able to produce a good set of results.
Otherwise, there will be a lot of over sightedness which is possible in this very simple test.
Because we have lot of controls in the case of triaxial testing, because of which it makes it
much complex than what it appears to be. So there are a lot of things like this and what actually
is happening in the field and for all those conditions we simply perform a triaxial compression,
because that is what is conventionally possible.
All other kinds of tests needs some specific and skilled manpower for performing this and this
is where the shear strength parameters that we obtained from the lab it deviates from the actual
results and for the same reason, geotechnical engineering or the strength based problems, it
becomes highly uncertain and it gives ample justification for considering these on a
probabilistic scale to account for these uncertainties just to add to why and what is the
implication of these kind of over sightedness or uncertainties.
313
Now, case 2 let us make a comparison in this particular case, we can see that this particular
angle is greater than ϕ’CD2. Now, there is not much use of this angle first of all it gives a higher
design value. And when we design a particular structure, we always need to be careful in the
sense we using a realistic value. Now, always people tend to be on conservative side if one is
not confident about a particular situation.
So, here ϕ CU2 is greater than ϕ’CD2. So anticipated undrained strength is higher than the drained
strength and this consideration may not be appropriate the first reason is Mohr Coulomb failure
envelope is relevant to only effective stress so, only this holds. So, ϕ’CD2 is relevant and
appropriate and for any undrained condition we have seen and we have repeated ample number
of times, it is better that we go ahead with Tresca criteria and the undrained shear strength Su.
Now stress history and CD strength consider a wide range of consolidation stress. Let us show
it in the form of the consolidation curve E versus log σ’d, this we have already discussed before.
Now, point E is on the reloading line which means that from this particular point C it is
unloaded to D and then it is reloaded to C now E is on the reloading path that means this is C
to D.
So, first C to D happens and then from D it comes back to C and E is on that unloading curve
points A, B, C and F are normally consolidated D, E are over consolidated it is very relevant
and we have seen this in previous lecture. So, now, let us translate these results to 𝜏 versus 𝜎.
So, what we are trying to do is we are trying to understand how stress history impacts this CD
314
strength we already know this, but again there it is important that we understand this particular
point.
So, here first let us mark the pre consolidation pressure. So, this separates NC and OC behavior
so less and more. So, more will be NC for the same confining stress strength of OC is greater
than NC that means we are considering point D and A. So, now D and A it is the same stress
condition now, but then D has a higher strength than A so, that is one important observation E
have higher strength than D now E is a point the soil has gone to D and then it is coming back
to E. So, what is happening is it is getting densified from D to E.
So, E have higher strength than D because it has lower initial w and lower e because when this
sample comes from D to E it is E is reducing. So, for E has a sample with lower e and hence
the strength of E will be more and here the state of OCR is also different for both D and E.
Now this becomes the curve failure envelope which we have already seen previously see curved
failure envelope for OC we have already seen this portion DEC is called pre consolidation hub.
So, this particular curve DEC, so C will come here reloading from D that is from D and going
past C that is from here it is reloaded to C and then when it is loaded further, it moves along F
so now F is the normally consolidated point and it will have higher strength, but the point will
be so, we are dealing with B and then C. So, from D to C it reaches here because C is the pre
consolidation pressure on further loading, it goes up to F.
So, now F is the normally consolidated line, but, this particular F has already been subjected to
this over consolidation in the past once the loading goes past C, now, once it goes past C it
erases all its previous memory. So, what has been done in CD link that particular thing is lost
now. Now, once it reaches see it again comes back to a normally consolidated state. So, that is
once the loading goes past C the previous stress history gets erased. So, now, it is left for a new
pre consolidation pressure maybe at F so, this becomes a new pre consolidation pressure if it
is unloaded from F.
315
Now, what are the relevance of CD test the way we have discussed for CU test more or less it
remains same. In fact, those conditions that we discussed in CU test, is more appropriate for
CD test in terms of its long term behavior so relevant for long term stability of earthen
embankments, excavations and slopes in soft and stiff clays. So, this is an embankment
constructed slowly were drained shear strength of clay will prevail.
Now, that is very important when there is a saturated clay and the construction is happening
very slowly then the drainage also would happen along with there will not be much of pore
water pressure development and hence, this will simulate closely the CD condition or it can be
excavation in clays or it can be an earthen dam with steady seepage. So, the clay core will
exhibit a CD behavior. So, the drain shear strength is persisting for clay core.
So, CD conditions are critical for heavily over consolidated soil for OCR greater than 4. And
these details are clearly given in Holtz and Kovacs book. It is a very fundamental book where
all these details which I am discussing is clearly written. So, CD conditions are critical why for
OC condition it is critical, we are considering a CD test drainage is permitted. Now, it is a
heavily over consolidated sample during shearing what is the tendency?
The tendency of OC sample is to dilate. Now, as it tends to dilate, it will suck water into it,
when it suck water into it water creates strength to reduce so it induces lowering of strength.
So, a drained condition in OC is going to be more detrimental because of this particular reason.
And this happens mostly in CD test drained tests where water can easily move in or move out.
316
So, undrained condition, dilating OC draw water into it causing strength reduction so, that is
what I just explained.
So, let us now just try to summarize this lecture. CD test is a slow test and useful only for freely
draining soil. So, we have to keep in mind that even though we have understood CD test, it is
practically not useful for clays the MC failure envelope passes through origin for NC soil, we
have repeated this several number of times and this should be there in your mind. The MC
failure envelope exhibit cohesive intercept for OC soil, mostly associated with dilation or
cementation if any.
As the consolidation over consolidation increases, that means OCR is more c’ increases and ϕ’
decreases as compared to NC the shear strength parameters of OC soils are influenced by both
normal stress and OCR. This also you need to understand very carefully like OCR also induces
changes in shear strength parameters, and the kind of confinement that we give controls the
dilation.
And hence the shear strength parameters, for stability analysis, we have to use Mohr Coulomb
effective strength parameters determined for the range of effective stress affecting normal
stress anticipated in the field. So, when we are planning the test, we have to make sure that the
confinement that we give closely simulates the overburden and the lateral stress condition that
is existing in the field.
317
Practically it is not possible to conduct CD test on clays because it is a time intensive analysis
or testing. CU test is a practical method for obtaining Mohr Coulomb failure envelope
parameters for clays provided the rate of shearing is not very high. Now, I just summarize this,
a lot of these details on shear strength of clays come from the work of Charles C. Ladd,
Professor in MIT, USA.
So, you can see that in 1970s, there are quite a number of reports and publications from Ladd
and many of these details that have come into the textbook related to consolidated undrained
UU test; it is basically coming from his contribution. So I just want to acknowledge this
particular aspect also in this lecture. So that is all related to interpretation of triaxial testing so
we are concluded this here.
And in the next lecture, we will summarize and we will see certain aspects, which we have not
covered in any of these previous lectures, some very minor important points relevant to shear
strength and transactional testing. We will have it in the next lecture. And with that, module 2
will be completed that is all for now. Thank you.
318
Lecture - 31
Some Additional Aspects of Shear Strength
Welcome back to all of you. In the last lecture, we have completed the interpretation of triaxial
tests, the last one being the consolidated drain test. In today’s lecture, we will see some
additional aspects which has not been covered till now related to shear strength of soils. And
with this lecture, we will be winding up this particular module. So, we will get on to some left
out aspects which we will discuss just now.
(Refer Slide Time:00:58)
So, we will see the first aspect is the comparison of stress strain curve when you have all the 3
tests done on a particular soil. All the other conditions remaining same, how will these results
compare, I mean to say UU, CU and CD. This comparison is not conclusive but a conceptual
comparison we can say. So, the first one is UU test, the stress strain response and the pore water
pressure generation as u nought is the initial pore water pressure before shearing.
That is may be at the consolidation stage. And then during shearing the pore water pressure
goes up. CU, it is also the same. The stress strain response more or less is identical and in the
event of back pressure application, the pore water pressure corresponding to consolidation
stages u back pressure. And when it comes to consolidated drain instead of pore water pressure
it is volume change. Now, the quick comparison it is only in terms of stress strain response.
319
You can see that the final pore the failure stress corresponding to all the 3 is progressively
increasing as it comes from UU to CU, CU to CD. Don’t take this comparison as conclusive.
Because there are various other factors that can alter these results. So, that is why I told all
other factors remaining same, a more or less comparison will be in this form in terms of its
strength. So, that is all about the comparison of the 3 types of tests that you perform in triaxial
testing.
Now, the next one is the criteria for determining the shear strength. Now, why should we have
this question at the first place? We conduct triaxial tests, any of these 3 tests. Plot the Mohr
circle. Now, where is the criteria coming into picture. So, when we have conducted the test we
would have either taken the peak stress at failure for plotting the Mohr circle. It will not be
done such that in one of the tests we will take peak the other one will take ultimate or critical.
It will not happen like that. So, there will be some criteria based on which the failure state of
the soil is defined. And we are we have already seen that the peak stress is only possible in the
case of dilating soils. So, always it may not be possible. In the case of normally consolidated it
progressively densifies. So, there is no particular peak or particular way of defining the failure
state.
So, what should be done in those cases? Having known the different type of tests and its
interpretation now the job is how to define the failure envelope and for that what criteria need
to be taken. So, that is why we will discuss, what are the common criteria which are adopted
320
for defining the failure envelope or shear strength of the soil? Now, this deviator stress versus
axial strain this can be volumetric strain as well.
Now what are the different criteria? It is very it is quite explicit that in such a case we can
always take peak as the criteria provided we can identify the peak. So, the first one is peak
deviator stress. So, 3 samples we conduct in the triaxial testing. In all the 3 samples the identical
samples we will consider the peak deviator stress. Now, this is possible only if the soil that is
tested is either dense state or it is an over consolidated state.
So, this is the first criterion for determining shear strength. The second one is maximum
principal stress ratio or more specifically maximum effective principal stress ratio. And we
have seen this and we have discussed this before also maximum stress obliquity. So, that is
what we have already seen. The same thing it becomes a very useful criteria for defining the
shear strength of the shear envelope. And that is very specific to may be un-drained soil.
So, in the figure you can see that the plot between σ1’/σ3’ versus 𝜖𝑎 at the isotropic
consolidation it is 1 because σ1’= σ3’. So, it starts at 1. So, this is the second criterion which is
possible. Now, second this corresponds to this peak and we can see that the peak of deviator
stress. It is not matching with the peak of σ1’/σ3’.
The next one is limiting strain that is the third point we will discuss each of these separately.
So, we will first see what are the available criterion. Then, comes the critical state. So, this also
to some extent we have seen while discussing the stress strain response. And the fifth one is
residual state which is at a very low stress condition. So, we will see each of these one by one.
So, peak deviator stress does not need much of explanation because we have seen this all
throughout the discussion of this particular module. There is a common criterion for dilating
soils, σ1’- σ3’ = σ1 -σ3f. Pore pressure at failure is uf. It is not a fundamental characteristic of
soil. Now I think this becomes again a testing or revision of whatever we have explained before.
Why it is told that it is not a fundamental characteristic of soil. It is because it is dependent on

the initial stress history and the confinement. Both is going to affect or determine how much
dilation it will undergo. So, it depends basically on how much the soil will dilate. So, hence it
321
cannot be considered as a fundamental property because it keeps changing. That is why most
of the time for the design there is a tendency to neglect the cohesive component.
The reason is you might have obtained cohesive component while doing the test in the lab but
what is the guarantee that the same cohesion will be available in the field. I am not talking
about the cohesion that is related to the actual cementation. That will be there. But if it is not
because of actual cementation and it is exclusively due to the kind of over consolidation it has
undergone then probably the dilation may not happen in the field while loading it is not clear.
So, there is a general tendency that we tend to neglect cohesion for design. So, what I mean to
say is peak characteristics and the shear strength associated with peak cannot be considered as
a fundamental property. To be very specific from whatever discussion we had earlier the
dilation angle which is over and above the critical angle how much it would be it is condition
specific.
Now, the maximum principal stress ratio, a very useful way of representing stress strain
behavior and obtaining the failure envelope for especially for un-drained case. The ratio of
principle effective stress is plotted versus strain. You have seen this in the previous slide.
Maximum value of σ1’/σ3’ need to be same as need not be same as that of (σ1’- σ3’)f.
What it means is that the peak value that you obtain from deviatoric stress which is this one
may not be same as in the case of σ1’/σ3’ and this is very true for un-drained case. Why? So,
this is what it is. Here the peak happens to be at this particular point that is corresponding to
322
point 2. And it is different from that of point 1. Now, this is preferred to σ1’- σ3’ in un-drained
test.
If you have noted deviatoric stress σ1- σ3 = σ1’-σ3’. So, it is not going to make much difference
even if you measure in terms of pore water pressure the un-drained testing. But σ1’ and σ3’ will
be exclusively different. So, that is the reason why σ1’ /σ3’ will be more reliable in the case of
un-drained testing.
It is also useful when the deviatoric stress does not exhibit a well-defined peak. So, next one is
limiting strain. So, what happens is in some soils large strain is needed to mobilize the failure
condition or rather it is very difficult to identify the failure condition. For example, in the case
of normally consolidated soil, it goes on increasing. So, it goes on densifying. So, where to
stop the test during triaxial testing? So, we need to have some criteria.
And these are also very well explained in the code of practice as well. So, there is a limiting
strain based on which we can terminate the testing. So, that is what is meant by limiting string.
So, that is the point 3. For such soil, the limiting strain is defined by point 3. And this can be
any specific value or the value which is coated in the code. For example, in IS code also there
is limiting strain specified. In ASTM also, it is specified. Both are different.
One is close to 10 %, the other one is 20 %. So, it is it or it can be any sort of limiting strain
which the engineer finds it appropriate for that particular condition. Whatever be the testing is
stopped corresponding to a limiting value of strain. Point 3 is then used for defining the failure
envelope. Now, if it happens to be slightly lower then what we are ensuring is whatever stress
that is developed within the soil will be close to the elastic limit.
Depending upon what elastics what limiting strain that we are fixing that will depend that will
determine what sort of failure envelope that we are taking. Because now all these stresses will
be plotted with respect to this particular point 3. What does that mean? If this stress what we
are considering is fairly less then it means that we are shifting our failure envelope downwards.
That means we are limiting we are becoming more and more conservative.
We are not allowing the soil to fail beyond a certain point. So, that means that all our design
will also get influenced by this. Yes, there is a trade-off between the conservative result and
323
the economical part of it. But this particular aspect when you consider limiting strain is if it is
too conservative with the limiting strain then the failure envelope that we get also will be fairly
conservative. That is what it is written.
It is relatively conservative. And it can it will be more conservative depending upon the limiting
starin.
Critical state is the fourth criterion and this we have seen quite a number of times as we
discussed in our previous lecture. For OC, the stress falls to 4 after peak. That is strain softening
behavior. For NC, it reaches 4 directly. And point 4 is known as the critical state. Now, we will
be discussing at length this concept of critical state in the fourth module. But for our quick
understanding, at critical state, shearing takes place at constant volume or void ratio.
So, it is the state which the soil achieves during shearing such that the volume does not change
further but deformation happens. Deformation happens at constant volume. So, that state is
known as critical state. It is more like a concept. And hence it is also more like a fundamental
property for a particular soil. This we have seen a bit of it in our earlier discussion. We have
drawn the failure envelope and the peak is over and above this critical.
So, for any soil, this particular inclination that is 5 critical angle of internal friction
corresponding to critical state becomes more or less a kind of fundamental characteristics. It is
an idealized behavior of remolded clays. And it is assumed to apply for undisturbed clays as
324
well. In short it applies to most of the kind of soils but it was mostly idealized for remolded
clays. And it is assumed that it behaves well even for undisturbed clays as well.
Now, the fifth one which we have not discussed till now is the residual state. You can see that
the residual state that is the fifth one is fairly different from all other 1, 2, 3 and 4. It is mostly
applicable for cohesive soils only. The reason is you can see here residual state corresponds to
very large amount of strain. That means the soil has already undergone a large amount of strain
and it is almost at the verge of failure.
So, the amount of shear strength which the soil mobilizes at a very high strain is what is known
as residual strength. And how much of it, it mobilizes? It is a very tricky and a complex
situation. But what it means is that it is a very low shear strength which is mobilized in a soil
corresponding to a large strain or large shear displacement. That is what is meant by residual
state.
If the soil is sheared beyond critical state the shear resistance continues to decrease. So, that is
what it is shown here. It reaches a constant shear resistance. So, at 5, it more or less becomes
a stable value at large strain beyond which it is more or less going to be either fail at that
particular point. So, it almost reaches a constant value. Point 5 represents residual shear
strength of the soil.
This value is considerably less than the peak value. Now, residual condition is formed in a
narrow shear zone due to reorientation of clay particles. So, it mostly it is associated with clays
325
and as example of such residual state is in the case of slope stability where there is failure
planes which are formed or slip lines which are formed. So, the residual condition develops
when large shearing has occurred as in the case of landslides or slip failure in slopes.
In the case of landslides, there is a continuous movement before the actual failure happens. So,
it has already undergone sufficiently large strain but it has not failed. So, whatever is the shear
strength that the soil possess during that particular time. That is not time that particular strain.
That is what is meant by residual strength. Now, determining residual shear strength is totally
different from what we do in the conventional triaxial testing.
It is very difficult to withstand the failure of a soil in conventional triaxial testing. So, there are
some special instrument which are used for determining residual shear strength like a ring shear
test where a large strain can be induced into the soil without failure. So, those type of testing
will help us to obtain the residual shear strength and which corresponds to large movements
like that in landslides and in slope failure.
So, now we have more or less finished like the whatever shear strength aspects which we need
to discuss. Now, after knowing all these things, what to apply where is an important question.
Again this discussion is not conclusive because the amount or the kind of possibilities that we
have in the field is innumerous. It is very difficult to pinpoint each of these possibilities in a
course.
326
So, a broad understanding of what to use, where, I will discuss in brief. Sliding along a pre-
existing slip surface. So, it is very clear like already existing failure surface or slip surface and
the soil is moving. In that case, what shear strength to use? So, based on our discussion today
we can easily say that it will be residual shear strength. That is what residual shear strength
𝜏𝑓𝑟 = 𝜎’𝑓 𝑡𝑎𝑛 𝜙’𝑟 . r corresponds to residual strength
See please note here, r stands for residual. Now, how do you get 𝜙’𝑟 ? We get 𝜙’𝑟 by considering
the deviatoric stress corresponding to residual state from the previous stress strain diagram.
That deviatoric stress value is used for plotting the Mohr circle and obtaining the ϕ value. Then
that ϕ value corresponds to residual strength. Newly formed slip surface that is it is just about
began to slip or the failure plane is just forming.
So, in that particular situation, which value we need to adopt? Now, this will depend on what
type of stress history or the initial state the soil has. So, for a normally consolidated state and
for short term strength we will use the un-drained shear strength 𝑆𝑢 . Please note we are not
using 𝜙𝑢 . For any un-drained case, for all practical purpose, we use 𝑆𝑢 , un-drained shear
strength for NC, for short term strength. And for long term, it will be 𝜏𝑓𝑟 = 𝜎’𝑓 𝑡𝑎𝑛 𝜙’𝑐 .
Please note here, it is critical state value because for normally consolidated, we do not have the
kind of peak. For OC, if it is small strain then for un-drained it remains the same whether it is
large or small 𝜏𝑢 = 𝑆𝑢 . And it is 𝜏𝑓 = 𝑐’ + 𝜎’𝑓 𝑡𝑎𝑛 𝜙’𝑝𝑒𝑎𝑘 . Please note here it is a small strain
problem. We know that the peak condition is always with respect to smaller strain.
And at larger strain, its strain softens and reaches the critical state. So, this is what we have
already seen. So, that information is very vital in appreciating this particular point. For long
term if you have a small strain problem it is conveniently we can use the concept of peak shear
strength, a ϕ’peak. And that is what is written here. And in the case of long strain, S u remains
the same. It is not sensitive to this.
So, in the case of large strain, we can see that it is convenient to use ϕ’c because by the time it
reaches a large strain it would have crossed the peak and strain softened to the critical state.
And for coarse grained soil is a freely draining material it is always we are using drained shear
strength parameters that is ϕ’, ϕ’peak or ϕ’ critical. So, it depends upon the denseness of the soil.
327
So, that is all about which shear strength to be used. Now, let us summarize. Like we have now
gone through most of the important aspects of shear strength. What are the different factors?
Now, when you start any lecture normally these factors are discussed in the beginning like
what are the different factors affecting the shear strength. But here I have chosen this to be
discussed in the end. Because many of these factors if discussed in the beginning it is very
difficult to appreciate.
So, here what are the different factors? There are 2 sets of factors. First of all before discussing
the factor let us make it clear that shear strength is not a unique property of soil it is condition
specific. And this we have discussed at length throughout this module. And it is very clear now.
Failure of soil can occur as a whole or along a well-defined narrow zone called failure plane.
Now, the factors are classified as natural factors and induced factors.
So, let us first see the natural factors, mineralogy of grains because the clay behavior itself is
determined by mineralogy. And it is very easy to understand also. For example, an expansive
clay behaves different from a non-expensive kaolinite kind of clays. Particle size distribution,
shape and arrangement. So, shape of the particle very easy to understand. Particle size
distribution how distributed it is and that distribution will determine how it is arranged when it
is compacted.
So, all these are going to determine what is the shear strength. Now, this is a property which is
very relevant for soil type of materials because you cannot say that shear strength of soil is this.
328
Because it is also dependent on how you are packing it a loose soil or a dense soil. So, for the
same soil you have different strength depending upon particle size and its arrangement.
Initial state in terms of void ratio and water content very important because that determines
again how compacted the soil is and compaction, it induces strength into the soil. Stress history,
we have seen this very clearly. So, if this I have told in the beginning it is very difficult for you
to understand. Now, since we have gone through the entire aspects of interpretation in terms of
normally consolidated, over consolidated, it is easy to appreciate how stress history would
affect the strength.
Existing in situ stresses, so, when there is a kind of undisturbed sample what kind of stresses it
was already subjected to that also it is a part of a bit of stress history itself. But, what is the in
situ stress condition that available in the field? That also determine how much strength it can
mobilize.
Induced factors. So, what the whatever we have discussed is by its own we are not doing any
additional aspect for that will affect the shear strength. So, induced factors includes stress
changes that happens during sampling. We have again seen this with an example what causes
during sampling. Initial state imposed before testing, this also it is governed by us what type
of initial state do we tend to give.
May be it is an isotropic, anisotropic consolidation. So, all these possibilities are there. So, this
depends on the quality of sampling, handling and how we are preparing the sample for the test.
329
So, this is exclusively dependent upon how we do the test. Stresses applied prior to testing, it
is more or less similar to that of initial state and whether it is isotropic, anisotropic, it can be
controlled for remolded samples.
Method of test, what type of test and sequence we have seen all these factors indirectly
throughout the lecture. And we have seen how that is going to influence the final strength.
Method of test, rate of loading, drained or un-drained test, these need does not need any sort of
discussion at this stage but we are just summarizing it. Can vary considerably and determine
the pore water pressure in the case of un-drained test.
What is the rate of development of pore water pressure? The criterion adopted for defining
shear strength. So, that is what we have just discussed. So, that also will determine what type
of failure strength and failure envelope that we are referring to. Shear strength varies depending
on the criterion adopted.
So, that is all related to what are the factors affecting the shear strength. Next is limitations of
triaxial test. In triaxial test, we do not consider the effect of intermediate principal stress quite
obvious. Rotation of stress is not considered which is a common phenomenon. If you consider
the slope and the slip line you can see that the principal stress rotation happens. Now, this is
not considered in triaxial testing rather it is not possible.
Not applicable for plane strain condition, we have already seen that plane strain is a very
common condition while dealing with problems related to soils but triaxial testing conditions
330
are different from that of plane strain. So, but we still tend to apply ϕ or c that we determine
from triaxial to any condition. But we need to keep in mind that it is different and the kind of
strength development also will change.
A very good example is the kind of pore water pressure equation that we obtained. For plane
strain, it is different from that of axisymmetric triaxial condition. Isotropic condition may not
simulate the actual field consolidation. Now, this is with respect to the conventional test that
we conveniently do. We do always isotropic testing. Now, there are ways of doing anisotropic
testing as well but not very common.
We need to keep in mind that this may not actually simulate the field condition. Effect of end
restraints, this we have also discussed in the previous one of the lectures where the pore water
pressure development may not be uniform within the sample because of this cap end restraints.
Now, for that we need to be careful while loading. So, that we have already seen. So, the effect
of end restraints remains in the case of triaxial testing.
In certain cases, the strength determined from triaxial test will be higher than the available field
strength. A very good example is the isotropic consolidation. So, there it will be different.
Strength that we get finally will be different.
With that we finish the discussion and the additional aspects. We will just see a very small brief
problems just to make sure that what is the influence of pore water pressure on the shear
strength. So, the first problem is results of triaxial CU test is conducted in NC soil. ΦCU = 13.3
331
degrees and ϕ’CU = 30 degrees. Deviatoric stress at failure is given 60 kPa. Cell pressure is
given as 100 kPa.
We need to find out what is the final pore water pressure. Now, if you don’t understand that
for an NC soil it passes through origin then we will struggle at this point thinking what will be
the cohesive component. So, that is very clear like for an NC soil, we do not have cohesion.
So, this is the kind of problem which has been given. We are asked to find out what is the final
pore water pressure. So, it is a very simple problem.
We just need to substitute in the Mohr Coulomb failure envelope. This is not given. 𝜎1 ’ , 𝜎3 ’ is
not given but 𝜎1 and 𝜎3 can be determined. 𝜎3 = 100 kilopascal. So, 𝜎1 - 𝜎3 = 60 kPa. 𝜎1 can be
determined which is 160 kPa. Now, substituting that into the equation, only unknown is 𝑢𝑓 .
𝜙′
(𝜎1 − 𝑢𝑓 ) = (𝜎3 − 𝑢𝑓 ) tan2 (45 + )
2
So, ϕ’ is also given which is 30 degrees. So, if you substitute that you get 𝑢𝑓 =70 kPa.
Another problem that is normally consolidated clay has 𝜙’ = 260 . Triaxial test was carried
out as follows. We are discussing 3 different cases. The first case is specimen is isotropically
consolidated at 200 kPa. Shearing is conducted as un-drained. Final pore water pressure is 50
kPa. Finally, deviatoric stress at failure, it is a straightforward question.
332
𝜎1′ = 385 𝑘𝑃𝑎. We have 200 kPa and the final pore water pressure is given. So, we can always
determine what is the kind of the deviatoric stress at failure.
𝜙′
𝜎1′ = 𝜎3 ′ tan2 (45 + )
2
So, 𝜎1 ’ = 385 𝑘𝑃𝑎 and 𝜎𝑑𝑓 = 385 − (200 − 50) = 235 𝑘𝑃𝑎.
So, ϕ’ is given. Substitute it, get the answer for 𝜎1 ’ then you can directly get what is 𝜎𝑑𝑓 .
Now, in the same case a, let us say that shearing is drained with zero back pressure. So, what
will be the deviatoric stress at failure? So, again it is 𝜎3 = 𝜎3 ’ = 200 kPa because now pore
water pressure is not there and there is no back pressure as well. So, 𝜎1 ’ = 𝜎3 ’. Same equation
we will get 𝜎1 ’ = 512𝑘𝑃𝑎. And 𝜎𝑑𝑓 = 312 𝑘𝑃𝑎.
In the case 3, shearing is done but there is a back pressure of 80 kPa. So, what will be the
deviatoric stress? Same procedure, only thing is here 𝜎3 ’ changes. You need to find out what is
𝜎3 ’. And 𝜎1 ’ = 307 𝑘𝑃𝑎 and 𝜎𝑑𝑓 = 187 𝑘𝑃𝑎.
333
So, that is all with this particular lecture. Let us summarize what we have seen in today’s
lecture. Different criteria that are used for defining the shear strength of the soil is discussed.
Drained and un-drained shear strength parameter should be used depending on the respective
field condition. So, that also we have discussed briefly. The natural and induced factors
affecting soil shear strength is discussed. And the limitation of triaxial test is listed.
So, that is all with this particular module. In the next lecture, we will summarize what all we
have learned in Module 2. That is all for now. Thank you.
334
Lecture - 32
Summary of Module 2
Welcome back in the last few lectures, we were discussing about shear strength of soil. So,
now, we have almost finished this particular module. Before winding up, we will just see
what are things we have learned from this module?
So, let me make it very clear like this module is a revision of shear strength concepts of soil.
Now, why I call it as revision is for some participants of this course, these concepts may not
be very clear during the undergraduate. So, we are just trying to bridge that gap so that
certain concepts are clear. And these concepts, the understanding of these concepts are very
important to appreciate the stress path and the critical state modules which are going to
subsequently come up.
So, it is important that even though if you have learned it is important for me to explain those
shear strength concept which is relevant for the upcoming modules. So, basics of shear
strength we started off with stating what are the important basics of shear strength. The
Coulomb’s friction model where we started off with and we have seen it gets integrated with
Mohr circle resulting in the very popular Mohr Coulomb failure envelope.
335
And the integration was done essentially to represent what is the stress acting on the soil.
That is σ’1and σ’3. And MC failure envelope is independent of normal effective stress. I
wanted to add the statement here so that you do not confuse this. What it means is that the
failure envelope as a whole it is linear with the inclination a constant slope. So, whatever be
the normal stress this particular envelope is not going to change.
So, that is what it means. And always Mohr Coulomb is expressed in terms of effective
stresses only. Different methods of stress representation and slope of failure line has been
determined. The relationship between maximum shear stress plot and Mohr Coulomb failure
envelope, the plot that is s-t plot or t-s’ plot, σ’1 + σ’3 /2 versus σ’1 - σ’3 /2.
So, the relationship between that plot and the shear strength parameters c and phi has been
determined. And we have also told that this is quite handy, because one can always find out c
and φ without plotting Mohr circle. Shear strength of granular soil, we have discussed in
detail, how the loose and dense state of the soil influences the shear strength. Loose soil
compress and strain hardens during shearing.
Dense soil dilates and strain softens during shearing. So, these aspects we have seen in this
lecture. Peak shear strength is applicable for dilating soil.
Now, peak shear strength is influenced by 2 important aspects. One is the normal effective
stress and the other one is initial state of the soil. Now, when I say normal effective stress that
confinement restricts the dilation and when I say initial state it has something to do with how
336
over consolidated the soil is. Both this influences how much dilation and the dilation angle.
And this dilation angle influences the peak angle of friction.
And hence, it is not a fundamental characteristics of soil. φ'cs is a φ corresponding to critical

state the effective friction angle you will see shortly that it is it can be considered as a
fundamental soil parameter. Concept of intrinsic friction, Taylor’s model and critical void
ratio was discussed. Shear strength of course, if soil is influenced by stress history, essentially
normally consolidated soil compress and over consolidate soil dilates during shearing.
Skempton’s equation for pore water pressure estimation is derived both for the triaxial
condition and plane strain condition. Overall pore water pressure parameter and its
importance for field problems have been discussed. The most versatile among all that is the
triaxial shear strength test and the interpretation of the important test which are UU, CU and
CD results have been done.
The criterion for shear strength determination is clearly explained. There are 5 criteria for
determining the shear strength. Factors affecting the shear strength has been summarized.
And the limitation of triaxial test has been explained. So, this is all what we have learned in
Module 2. So, with this background, now if you do not remember some of the concepts of
Module 1, it is still.
But it is invariably necessary that you remember all these concepts when we move on to the
next module that is Module 3 on stress paths. So, we will see in the next lecture. That is all
for now. Thank you.
337
Lecture - 33
Stress path and representation
Welcome back, all of you. With last lecture, we have completed Module 2 which was basically
on shear strength. We discussed various aspects which are important in terms of strength of the
soil. And today we will be starting a new module and that is on stress path. So, we will see
what is the importance of stress path vis-à-vis the shear strength of the soil and how this
information becomes handy.
So, we will move on to what is known as stress path. So, stress path in short it is SP. I am
calling it as SP is a 2-dimensional graphical plot representing theoretical or experimental
relationship between 2 stress parameters. If this is the discussion if this is the definition of stress
path I think we have already seen a stress very relevant stress path in terms of Mohr Coulomb
failure envelope and in terms of τ and σ’ that is shear stress versus normal stress because these
are also 2 stress parameters and the plot indicates the relationship between these two τ and σ.
That is shear stress and normal stress but relevant to the failure condition. So, that is what we
have seen in terms of Mohr Coulomb. So, it is a similar kind of thing that it relates the variation
of 2 different stress parameters.
338
So, that is what we understand by stress path and it is a 2-dimensional graphical representation.
It is the locus of stress points during loading and unloading. So, whatever is happening to the
soil during loading, what is the variations in the stresses? During loading as well as during
unloading can be captured and when it is plotted it becomes a stress path. It is a convenient
way to visualize how stress changes during loading and how it approaches failure state.
So, this is the difference what we will find generally when we discuss about stress path and the
Mohr Coulomb failure space that we have discussed before. We never talked about what is the
manner in which these stress changes and then how it approaches failure. We have simply
defined the bound or the failure state within which the soil mass would operate. So, that is how
we have conceived before.
But in the case of stress path, it clearly shows how the soil changes in terms of its stresses
during loading and how it reaches its failure state. This is a very important information for non-
linear materials like soils. So, it is a common practice to use stress invariants for plotting stress
path. Now, you should understand the relevance of why we have discussed about stress
invariants before.
So, it is very important to understand what is the why we one should go for stress invariants.
Now, that those aspects we have already discussed. So, I am just go just highlighting that like
it is very important that certain things are plotted in a particular manner. For example, stress
path it is quite handy when you use stress parameters like stress invariants than the normal
stresses.
Normal stress means not σn in general terms like if you use shear stress and normal stress rather
than that it is always convenient to use stress invariants. Now, stress path can be total stress
path and effective stress path. If it is drained or un-drained response in the case of drained
response, total stress response and the effective stress response will be same. But if it is un-
drained response, we will have both because there will be pore water pressure. Now, what is
the implication of both? All these things, we will see in the subsequent lectures. Importance of
stress path, for an elastic material, the progression of stress during loading is very well defined
and it is known if the elastic properties are known. How do you generally represent this the
mechanical behavior of elastic materials?
339
We have the stress strain response what is happening. The moment you know what is the initial
state and the elastic properties more or less everything is defined for an elastic material. So,
there it may not be that relevant. The concept of stress path may not be that relevant. For elastic
behavior, there is no importance on the manner in which it is loaded. Because once the slope
is fixed that is the modulus is fixed then everything is known.
We are not bothered how a particular stress rate changes from beginning of the loading to the
end of elastic limit. So, this is valid only between initial and final state. Now, final state here
for an elastic material is not in terms of failure it is up to its yield limit. So, it does not matter
how these stress changes. Ultimately, it has to be within the elastic limit or the yield point or
the yield limit.
Now, for elasto-plastic materials like soil, it is essential to know the manner in which stress
changes from initial to final state. This is very important because depending upon the initial
state we have already got a clue from our previous lecture. Especially with respect to Module
2 like depending upon the initial state there can be different possibilities for the same soil and
that can be further dictated by the kind of drainage it possess and also the kind of stress
history’s.
Initial state means it can be a kind of how it is packed or what disturbance it has occurred. But,
on top of it, it is also dictated by stress history as well. So, here in the case of soil, it is not
enough like if you know, what is the initial state and, what is the final state. Final state here
340
refers to the failure state. But it is equally important how the stress states in the soil changes
during its course from initial to its failure state.
So, it is important to simulate stress sequences in the lab which is relevant to the field condition.
The conventional triaxial compression is not realistic always. So, if you are really serious about
the kind of performance of the soil in the field. Especially for very important structures where
a lot of load is going to get imparted then we also need to simulate the actual field condition.
For example, if it is building load, we know that it is going to be compressive most of the time.
But if it is an excavation that need to be made then we know that there will be a kind of release
that is happening. If it is for a retaining wall then the things may be different. If you are looking
at a slope then the stress conditions may be different. Now if we want to actually understand
the performance of the soil under different sequences.
Sequences means all these problems what I just mentioned. It all induces a different type of
stress response within the soil. If you want to actually understand that with respect to its failure
then we need to simulate different kind of conditions in the lab. But in general practice for
finding out the design parameters it is a common practice that we adopt only triaxial
compression test and conventional triaxial compression test.
So, that may not be realistic all the time. And it is also important even in a conventional triaxial
compression test. It is also important like what is the initial state of the soil and how the state
changes during loading. So, stress path is essentially a function of soil type, stress history,
drainage and loading sequence. All of these factors we have discussed in detail when we
discussed in our Module 2. So, all these things have been touched upon.
So, I am not getting into the details. All those factors which affected shear strength is going to
affect the evolution of stress path as well. Now if we want to adopt stress path for our further
understanding we also need to know how these are represented. So, this is one way of
representation. That is actual stress as a function of radial stress and there is a failure line which
is represented by kf.
So, if σa and σr is a more general representation now I would always prefer and recommend
that we always adopt axial versus radial than in terms of principal stresses. Because, depending
341
upon conditions, axial and radial major minor principal stresses would change. And this we are
going to see in this particular module. So, it is better in general represent σa’ on y axis and σr’
on x axis.
So, that is what we will be following in terms of stress path representation for this module. So,
this is another way of representing σa and σr’.
We have also understood in detail q versus p’ plot. And that essentially came from Cambridge
plot and what is q and what is p’ is all well explained. Now the terms stress invariants comes
into picture where q and p’ can be treated as stress invariants. And p’ just to revise because we
will be using extensively these symbols in this particular module. It is just to summarize.
(𝜎𝑎′ + 2𝜎𝑟′ ) (𝜎1′ + 2𝜎3′ )

𝑝′ = =
3 3
𝑞 = 𝜎𝑎′ − 𝜎𝑟′ = 𝜎1′ − 𝜎3′
It can also be represented in terms of t and s’ which is essentially MIT plot or maximum shear
stress points. So, here t versus s’ we have also dealt these aspects of stress representation in our
earlier lectures and we have already seen that.
So, it is just to revise in this particular module because we will be using it for stress path
𝜎𝑎 ’+𝜎𝑟 ’ 𝜎𝑎 ’−𝜎𝑟 ’
plotting. So, s’ = . And t = . So, t is plotted on y axis, s’ is plotted on x axis.
2 2
𝜎𝑎′ − 𝜎𝑟′ (𝜎1′ − 𝜎3′ )

𝑡= =
2 2
342
Similarly, q is plotted on y axis and pp’ is plotted on x axis. That is what we generally use.
Normally, we talk mostly about compression.
But now in this particular module to cover the entire aspect for holistic discussion we will be
discussing the cases of extension as well. Extension not necessarily mean the soil is sample is
just held on both ends and it is extended. It is not that. We can based on the sequence in which
we apply stresses we have already seen in the previous module that a kind of compression or
extension can be simulated based on how we change the compressive stresses.
We only talk about compressive stresses σ1 in the axial and σ3 radial for a typical triaxial
sample. Now, controlling these two , we can even simulate extension conditions. So, I am
talking about that particular extension and not the realistic extension of soil. We know that soil
cannot tolerate such tensile stresses. Now plotting stress path for given stress variation. So, we
will slowly get into the plotting of stress path for different conditions.
Now in this particular module, we will be dealing only with this like how to plot stress path for
different cases. Now this manner in which how you plot stress path is important for the fourth
module. We need to know the inside out of how stress path emerges and how we will define
the stress path for different cases.
So, it is basically to represent the given stress variations. First let us talk about there is an
isotropic increase in effective stress and which is represented by OA’. Now we will be
essentially dealing with 3 stress path plots. One is σa, σa’ versus σr, σr’. Second one is s or t
343
versus s s’. I hope you understand why I am not telling t t’ and s s’ because deviatoric
component does not get influenced by pore water pressure.
So, t = t’. So, t versus s s’. Similarly, for q versus p p’. So, these are the 3 different plots which
will be adopting whole throughout this module. So, first let us consider σa, σa’, σr, σr’. So, this
σa, σa’ is floated on y axis and σr, σr’ is plotted on x axis. So, first is isotropic increase. Now
σa= σr or σa’= σr’ means it is a 45 degrees line.
And that 45 degrees line represents the isotropic condition or hydrostatic condition, isotropic
line. So, this is another important representation in a given stress path. So, first our condition
is isotropic increase in effective stress. So, it is OA’. So, we start from O and hence the first
aspect is OA’. It is along the isotropic line. So, you may be wondering what is so great in doing
this.
In fact, there is nothing great. What we are doing is we are simply following what are the given
increments of stresses. That is all. But it may look very simple. But as you get into the details
of it even the simple plotting can sometimes create a bit of confusion. Now with that confusion
it will be difficult to learn critical state soil mechanics chapter. So, that is why even the very
simple things are described in this particular module.
So, the first one is it is this particular exercise what we are discussing is extremely simple.
There is nothing great but then there are some simple aspects which becomes clear from this
discussion. So, OA’ is the isotropic increase. So, we know that this is the isotropic line. So, it
goes along OA’. Now forget about the magnitude. We are only understanding it notionally.
That means it is only conceptually we are trying to understand.
Forget about all the magnitude now. The second one is increase in σa’ and σr’ is constant. So,
that is represented by to be represented by A’ B’. So, the starting point is here. So, here it is at
constant σr’ its σa’ is increasing. Now third condition is increase in σr. So, σr is increased and
σa is constant. Now please remember here. I have represented it in terms of total stress.
Because, total stress is what actually happens onto the soil, the external stress which gets
imposed onto the soil. So, what is done is σa is now kept constant, σr is increased. So, now there
344
is an additional condition which is stated there during the increase in σr, the valve is closed or
the drainage is not permitted or it is an un-drained loading. So, we need to highlight.
How the stress path moves during the un-drained loading which is to be represented by B’ C.
Now this is only for your conceptual understanding may not be actually realistic. Now when at
this particular point B’ it is a saturated condition at a particular B’ if we raise the σ r component
without allowing drainage to happen then obviously there will be pore water pressure
development and it is a saturated condition. So, what happens?
So, then this is represented by B’ C. Now, what is happening here? We have not increased σa.
We have increased only σr. But remember it is an un-drained condition and hence it is a
hydrostatic pressure condition. So, when it is an un-drained condition when stresses are
increased there is an all-round increase in pore water pressure. And that means it is a hydrostatic
condition. σa also gets increased σr also gets increased within the soil mass or it is in both the
direction the rise in pressure will be same. So, pore water pressure is all round. So, it will have
in increment in both the direction provided it is un-drained. So, definitely the stress path BC
will be parallel to isotropic line because it is equal increment in both the directions. So, Δu
=Δσr. So, equal Δu will be available for σa as well.
This happens only under un-drained condition. So, the fourth condition is increase in σr’. Now
if we are allowing drainage to happen. So, drainage condition when excess pore water pressure
is dissipated. So, it is increase in σr’ and constant σa’. Now how this point is going to get shifted.
This is going to get shifted on the line C’. So, it does in fact it need not be at C’.
It is again for your understanding the if it is not specifically mentioned that it is going to be at
C’. It only states that σr’ increases and σa’ remains constant. So, it has to be horizontal. So, it
can be anywhere in between. Now we are just for convenience we are just showing with respect
to C’ because from here it just falls down. So, B’ C’ represents the fourth condition.
Now this is very simple because it is in terms of σa, σr, σa’, σr’ whatever happening we can plot
it. Now let us plot the same exercise on st plot and qp plot. Now there we need to determine
how the stress path moves here it is straightforward.
345
So, next is the plot on MIT. That is t, s, s’ plot. So, we know that
𝜎𝑎 + 𝜎𝑟
𝑠=
2
′ ′
𝜎𝑎 + 𝜎𝑟 (𝜎𝑎 − 𝑢) + (𝜎𝑟 − 𝑢)
𝑠′ = = = (s − u)
2 2
So, t = t’. Both are same.
𝜎𝑎′ − 𝜎𝑟′ 𝜎𝑎 − 𝜎𝑟
′
𝑡=𝑡 = =
2 2
′
𝛥𝑡
=0
𝛥𝑠 ′
So, this is deviatoric component and hence not affected by pore water pressure. It gets cancelled
off from both ends. So, the same representation OA’ that is on an isotropic line. So, this is t on
y axis, s, s’ on x axis. So, now here it is already written it is isotropic line. Why? Why this x
axis is isotropic line. You can refer to here.
That means if you take this isotropic condition means these two are same. So, σa’- σa’=0. So, t
is always 0 for isotropic conditions. So, that is why this particular x axis is an isotropic line in
t s’ plot. So, now OA’ we cannot plot simply directly for in this case it is possible but it is not
direct. So, we will find out how we have to do that.
𝛥𝜎𝑎′ + 𝛥𝜎𝑎′ 2𝛥𝜎𝑎′
𝛥𝑠 ′ = = = 𝛥𝜎𝑎′
2 2
So, Δs’ that is in this direction it will be Δσa’. And Δt’ = Δσa’- Δσa’ = 0. So, it has to be on this
particular line.
346
And hence
′
𝛥𝜎𝑎′ − 𝛥𝜎𝑟′
𝛥𝑡 = =0
2
So, it will give OA’ that is on the isotropic line and this is isotropic condition.
3)B’C
𝛥𝜎𝑎 − 𝛥𝜎𝑟 𝛥𝑢 − 𝛥𝑢
𝛥𝑡 = = =0
2 2
𝛥𝑡 ′
=1
𝛥𝑠 ′
We will be dealing more with isotropic loading in the subsequent lecture. Here it is only for
our notional understanding. So, here it is A’B’. What is A’B’? A’B’ represents σa’ is increasing
𝛥𝜎𝑎′ +0 𝛥𝜎𝑎′ 𝛥𝜎𝑎′ −0 𝛥𝜎𝑎′
whereas σr’ is constant. So, that is Δ𝑠′ = = and then 𝛥𝑡 ′ = =
2 2 2 2
𝛥𝑡 ′
So, that gives the slope of = 1. So, how this point is going to move? Now probably you
𝛥𝑠′
will understand why we need to pay a lot of attention when we are plotting stress path. It is the
information to us is only in terms of σa’, σr’ or in terms of major minor principal stresses. Now
for us to plot in terms of stress invariants, we also need to make this calculation and we also
need to know, what is the slope at which the stress path would move? So, entire exercise of
𝛥𝑡 ′
stress path plotting is all about finding out the slope in which it moves. So, here = 1. So,
𝛥𝑠′
we know that A’ B’ should be moving at an inclination of with a slope 1. So, that is what is
shown there A’ B’. Second part it is with a slope of 1 is to 1.
347
Now B’C is that momentary un-drained condition at point B’. So, we will see that how it
𝛥𝜎𝑎 +𝛥𝜎𝑟 𝛥𝑢+𝛥𝑢
happens. 𝛥𝑠 = . Now the pore water pressure is equal. So, 𝛥𝑠 = = 𝛥𝑢. So,
2 2
increment in along this direction is Δ u. That is what it means. But Δ t since both are same it is
=0. So, that means the slope is 0 but Δ s has an increment of Δ u.
How do we plot it? We plot it like this. B’C with an increment in Δu. And it is horizontal
because Δ t is 0.
Now let us see the last one. That is σa’ is kept constant and σr’ is increasing.
0 + 𝛥𝜎𝑟′ 𝛥𝜎𝑟′
𝑠= =
2 2
𝛥𝜎𝑟′
So, here Δσr’ is increasing. That is 𝑠 = .
2
0 − 𝛥𝜎𝑟′ −𝛥𝜎𝑟′
𝛥𝑡 ′ = =
2 2
So,
𝛥𝑡 ′
= −1
𝛥𝑠 ′
Δt’/Δs’ is -1. That means it has to drop in the stress path.
As you can see is B’ C’ is =1 is to 1 slope but with a negative slope. So, this is how we plot for
st plot.
348
Now we will move on to the third plotting. That is Cambridge plot which is in terms of q, p,
p’. So, once we have done with st plot, it is quite easy to understand even q, p, p’ plot. More or
less is the same procedure where
𝜎𝑎 + 2𝜎𝑟
𝑝=
2
′ ′
𝜎𝑎 + 2𝜎𝑟 (𝜎𝑎 − 𝑢) + 2(𝜎𝑟 − 𝑢)
𝑝= = = (𝑝 − 𝑢)
3 3
𝑞 = 𝑞 ′ = 𝜎𝑎′ − 𝜎𝑟′ = 𝜎𝑎 − 𝜎𝑟
Because this plus this becomes p minus u minus 2u becomes 3u by 3 that is u. And q is =q’.
This we have already seen. So, first is OA’. That is isotropic stress path and here also in q, p,
p’, similar to t; 𝑞 = 𝜎𝑎 − 𝜎𝑟 . So, if they are same that is σa’ =σr’ then it will be 0. The slope
will be 0. So, this becomes the isotropic line. That is x axis is the isotropic line.
So, Δp’ for OA’ now both are it is an isotropic increase. So, σa’ and σa’ this becomes same.
𝛥𝜎𝑎′ + 2𝛥𝜎𝑎′ 3𝛥𝜎𝑎′
𝛥𝑝 = = = 𝛥𝜎𝑎′
3 3
And Δ q is =
𝛥𝑞 = 𝛥𝜎𝑎′ − 𝛥𝜎𝑟′ = 0
𝛥𝑞
=0
𝛥𝑝′
So, definitely it is slope with 0. So, that OA’ is on the x axis.
349
3) B’C
Next is A’ B’. So, A’ B’ is σa’ is increasing σr’ is 0 which is given

𝛥𝜎𝑎′ + 0 𝛥𝜎𝑎′
𝛥𝑝 = =
3 3
𝛥𝑞 = 𝛥𝜎𝑎′ − 0 = 𝛥𝜎𝑎′
𝛥𝑞
=3
𝛥𝑝′
So, it has a slope of 3. So, that is plotted in s 2. That is A’ B’ is with a slope of 3. Now we need
to understand for such a particular increase the slope is 3. Now there can be different ways by
which σa’, σr’ are changing. Here it is a simple case where σr’ is kept constant. So, that is why
it is giving a slope of 3. Now B’C now that intermittent un-drained condition. How it will look
like.
𝛥𝜎𝑎 + 2𝛥𝜎𝑟 𝛥𝑢 + 2𝛥𝑢
𝛥𝑝 = = = 𝛥𝑢
3 3
There is an increment of Δu.
𝛥𝑞 = 𝛥𝜎𝑎 − 𝛥𝜎𝑟 = 𝛥𝑢 − 𝛥𝑢 = 0
𝛥𝑞
=0
𝛥𝑝
So, slope is 0 for B’C where the increment is Δ u in an un-drained condition.
350
Now the last one is B’ C’. So how the slope changes when you have σa’ constant and σr’ is
increasing.
′
0 + 2𝛥𝜎𝑟′ 2𝛥𝜎𝑟′
𝛥𝑝 = =
3 3
𝛥𝑞 = 0 − 𝛥𝜎𝑟′ = 𝛥𝜎𝑟′
𝛥𝑞 −3
=
𝛥𝑝′ 2
So, you see here, slope should be known and the slope is specific to the kind of stress increment
which we have imposed.
So, this may change again if the stress variations are same. So, we should understand how
important it is the sequence and the manner in which the stress changes. You can plot this. Now
this is negative slope of 3/ 2 which is plotted here. So, this is a quick example of how a stress
path variations look like. Now what we will be doing in the subsequent lectures is to understand
how these stress path need to be plotted for common conditions.
That we normally find in soil mechanics. So, that is what we will be do. So, we will summarize
for today’s lecture.
351
As stress path is the graphical representation of any 2 stress parameters. It is a convenient
method to trace the progression of stresses during loading and how it approaches towards
failure. Now this particular sentence will be very clear as we move forward and understand
stress path much better. So, stress path can be effective stress path or total stress path. Now the
relevance of these 2 will also be known in the subsequent lectures.
Stress path is a function of soil type, stress history, drainage and loading sequence. Different
stress path parameters can be used for plotting stress path. It is a common practice to use stress
invariants for plotting the stress path. Now plotting of stress path in terms of σa, σa’, σr, σr’,
then t versus s, s’, q versus p, p’ is demonstrated for simple stress variations. So, that is all for
today’s lecture.
In the next lecture, we will add more meaning to this stress path plot. That is all for now, thank
you.
352
Lecture - 34
Failure line in stress path
Welcome back, all of you. In the last lecture, we have started a new module that is on stress
path and we have seen what is the importance of stress path? Now, we will move on to another
particular aspect of stress path. That is what how one can make the stress path plot meaningful.
So, let us start with today’s lecture and we will start with an example. So, before moving on to
the actual aspect of failure line in stress path, we will just try to see a simple example of a
drained and un-drained triaxial compression test, which are given. And, we have to plot this
stress path in σa, σa’, σr, σr’ plot, MIT plot and Cambridge plot. So, we will see how this stress
path will look like.
So, this is the data for the drained test, drained triaxial compression test where σ1 is given 300,
400, 500, 565, 590 and σ3 is constant, which is 300 and back pressure is given as 100. I hope
you remember what is meant by back pressure. And the un-drained test results are given. Again,
there is an increase in σ1 and σ3 is constant. And, at every instant of loading, we have the pore
water pressure as well. So, this is how it looks like. So, this is the preliminary data which is
given.
353
So, let us first try with σa, σr, σa’, σr’ plot with a drained envelope. So, this is the drained data
which is given. Now, is there anything to be changed in this? Yeah, first of all, we need to
understand in the case of compression σ axial is equal to σ1. So, we have written σa and σr now
separately. σa is the same as σ1. And σr is nothing but σ3.
And for every σa, it is 300. Then, we have σa’ and σr’. So, there will be total and effective stress
path that hope you remember. For every loading, there can be total stress path and effective
stress path. Those, which is plotted as per total stresses gives total stress path and the other one
gives effective stress path. So, σa’, σr’ is given. How did we get?
Because there is a back pressure of 100 kilopascal. So, we need to minus it from all these minus
100 will give σa’. And σr minus 100 that is 200 gives σr’. So, how the plot will look like. So,
the plot will look like this. So, σa σa’ on y axis and σr σr’ on x axis. So, here the total stress is
plotted. We can see from the figure that σa is increasing σr is constant. So, that is what is given.
And σr’ is constant at 200 and σa’ keeps increasing. So, this is increasing in this direction for
both effective stress and this total stress as you can see here and this is effective stress. So, now
what we can make out from this plot. Yes, we have plotted the variation in stresses as it
progresses towards failure. But, from this particular plot, what is that vital information that we
are getting.
In fact, we are missing something in this particular plot. So, we will sum up in the end what is
that we are missing.
354
So, now for the same triaxial compression for the un-drained test results, so, we have un-
drained test results and again, we need to find out what is σa, σr that remains same, same data
and 300 is the σr. But σa’ will change corresponding pore water pressure has to be detected. So,
that is how it looks like 300 minus 100, 200 like 398 minus 232, 166. So, σr’ here you can see
it is no longer constant.
So, in initial case, it was constant. Now, even though σr is constant, σr’ keeps changing so
depending upon the pore water pressure. So, you can see that 300 minus 100, 200 whereas 300
minus 232, 68. So, how this particular stress path will look like. Now, the stress path has
changed corresponding to drained and un-drained. Now, things have changed. But you can note
that the pattern of total stress path is more or less same.
But total stress path is same then the effective stress how it changes in un-drained test is totally
different from what we have seen in the previous slide. Like both were vertically rising but in
this particular case you can see that it shifts towards left. Still, we are not able to make out
much from this particular plot. Let us go ahead because let us see. We will be as I told in the
beginning we will be considering 3 stress path plots.
Basically, σa σa’ then st plot and qp plot. So, we will see for all the 3, how it looks like. Then,
we will come back and summarize.
355
Next is t, s, s’ plot with the same data. So, drained data remains same but whatever drain data
we have seen in the previous slide, so, this one. So, this is the drain data. Now, from this drain
data, we will have to find out what is the data for t, s, s’ plot and we have the relationship for t
and s, s’. So, that is given here. So, s and t is given for drained and s’ will be this minus the
back pressure is there. So, that will get changed.
Whereas in the case of t, there is no change as we have seen earlier that t is not affected by pore
water pressure. So, then how the stress path will look like. Again, now the pattern of stress path
has changed in ts plot. So, this is total stress and this is effective stress. But the pattern of both
effective and total stress remains same in the case of drained stress path.
356
So, corresponding un-drained, so, we have pore water pressure. So, accordingly s will change
and we know that s’ is nothing but s - u. So, that relationship we have seen. So, then 300 325,
this is s, t is 0, this up to 49. That remains same even for effective stress whereas total stress s
effective stress s’ will change. So, this is the data for s t and s’ t. So, how it will look like.
Again, the patterns have changed.
So, in this total stress path, the pattern is more or less same whereas in the case of effective
stress it moves leftwards. So, this how it looks like for t, s, s’ plot.
Now, for q, p, p’, it is more or less same as that of st plot. Again, we need to find out p and q.
I hope you remember, what is p and q? If not, please refer back. And corresponding p’ is
obtained and q is obtained. Note that q remains same. q is equal to q’. So, in this particular case
for the drained envelope, it is very much similar to st plot and it the pattern of both total stress
and effective stress paths are same.
357
Now, for the corresponding un-drained case for p q and p’ q. So, that will give you again the
total stress path maintains the same kind of pattern as that of drained but in the un-drained
effective stress path moves towards left. So, we have now presented a very simple example
how the total stress total stress path and the effective stress path looks like in different stress
path plots.
So, now what we have understood from these stress path plots. Why we are plotting this and
why we are studying this. We want to understand when the soil will fail or and from the
beginning how it traverses how this stress changes up to its failure state. So, definitely that
information is missing from whatever we have discussed till now. Now, there has to be a bound
which gets attached to each of these stress path plots.
Without which this information is missing. So, that is the importance of today’s lecture, failure
lines in stress path plots.
358
So, we will now get on to that what is meant by failure line. So, it is need of failure line in
stress path plot. What is the relevance of stress path plots? So, to make you understand this, we
have explained the given examples so in all the 3 plots we have seen. So, what is the relevance
of stress path plots? Locus of stress point, stress point means variation of any 2 stress
parameters. So, that gives you a particular point. So, how this point changes?
So, locus of stress path point is not relevant unless the bounds are defined. We should know
like there is a space given to us that is st plot or qp plot or whatever in that domain what are
the possibilities for this stresses to change for a given soil. This is very important. So, that is
why it is told that we need to know the bounds within which this stress path points can change.
So, that bound is defined by failure lines in each of the stress path plots.
Failure can occur both in compression and extension. Normally, it is a general tendency that
we discuss only about compression and we do not really talk about extension. But, we should
also understand that extension is not like you are holding soil and pulling it apart. That is not
the only case of extension. Extension can be created even by compressive stresses the manner
in which you are applying. I think some examples we have seen in Module 2.
So, here when we are discussing for stress path, we will discuss both. That is one is the mirror
image of the other. So, we will discuss both. So, failure can occur both in compression and
extension. Failure lines correspond to both compression and extension because of this.
359
So, first let us see σa, σa’, σr, σr’. Now, I have already told you before that let us identify all
these stresses in terms of axial stress and radial stress. The reason you will understand in this
particular slide. So, here is a typical case of compression where the triaxial compression what
we generally do in the lab. So, this σ1 is the major principal stress, σ3 is the minor principal
stress, the confining stress.
Now, here let us see that both are compression but here it is a case of extension. Why? Here
you can see that σ1 is major and σ3 is minor. So, please understand that when there is a larger
stress which is acting readily it has a tendency to squeeze the soil apart. So, squeezing is just
like an extension. So, same compression but in a different manner of stress application can
result in extension.
So, that is what it is not the direct extension that we are talking about. So, these are 2 cases.
Now, you can see that the axial stress in this. And the axial stress in this is changing. So,
depending upon whether it is compression or extension σa, σa’ will change, σr, σr’ will change.
So, according to Mohr Coulomb failure envelope now that is the only failure criterion that we
know.
That is Coulomb’s envelope and the extended version of the modified version which is the
Mohr Coulomb failure envelope. These are the failure criteria we know. How the soil will fail
based on this. So, according to Mohr Coulomb failure criterion, we know that σ 1’ is equal to
360
𝜙′ 𝜙′
𝜎1′ = 𝜎3′ tan2 (45 ′
+ ) + 2𝑐 tan (45 + )
2 2
Now, for convenience, let us assume c’ to be 0. So, that it becomes a bit simplified.
Otherwise also, we know that c’ is a parameter that can happen only if there is cementation or
if there is dilation. So, for the time being, let us consider c’ equal to 0. So, we can write
𝜙′
𝜎1′ = 𝜎3′ tan2 (45 + )
2
So,
′ ′
𝜎1′ 𝜙 1 + sin 𝜙
2 (45 + )=
′ = tan 2 1 − 𝑠𝑖𝑛𝜙 ′
𝜎3
′ ′
𝜎3′ 𝜙 1 − sin 𝜙
2 (45 + ) =
= cot
𝜎1′ 2 1 + 𝑠𝑖𝑛𝜙 ′
So, in a way, we have now defined the failure lines in the σa, σa’ plot.
For compression
′
𝜎𝑎′ 𝜎1′ 𝜙
2 (45 + ) 𝐹𝑎𝑖𝑙𝑢𝑟𝑒 𝑙𝑖𝑛𝑒 𝑘𝑓𝑐
= = tan
𝜎𝑟′ 𝜎3′ 2
So, failure line is denoted as kfc where c stands for compression. So, here σa, σa’, σr, σr’, this is
the 45 degrees line and then the kf compression line that is the bounds of compression failure
is given by this line. So, this is nothing but the bound in compression failure. So, this line
governs that any soil mass which is starting from any initial point in this domain has to fail
here. So, there is no stress state which is possible for a given soil above this line. So, that is
361
what it means. So, we have now defined one bound. Similarly, for the other bound that is for
the extension case.
For extension
′
𝜎𝑎′ 𝜎3′ 𝜙
2 (45 + ) 𝐹𝑎𝑖𝑙𝑢𝑟𝑒 𝑙𝑖𝑛𝑒 𝑘𝑓𝑐
= = cot
𝜎𝑟′ 𝜎1′ 2
So, in the figure the red line is the failure line for extension. So, kf extension, this is denoted
𝜙′
by kf extension. And this is cot 2 (45 + 2
) is the bound in extension failure.
And for example, 𝐹𝑜𝑟 𝜙 ′ = 300 , then you substitute this you will get kf of compression will
be equal to 3. To show that the relative position of kf compression and kf extension with
respect to isotropic line. Now for isotropic line you have k is equal to 1. Now, here kf of
compression is equal to 3. So, the slope will increase and kf of extension is 0.33. So, slope
will be less than 1.
So, the relative position of kf compression and kf extension is known. Now, any stress state that
we have discussed in the previous slides if you put we know that where it should fail. Since the
friction angle for that particular example is not given. We can find out and then we can see
where it falls like. So, we will not get into that. Just want to make sure that any initial point we
know where the soil would fail. Now in what manner it will fail whether it is a straight upwards
or moving towards left it depends upon the type of the test. So, this is about σa, σa’, σr, σr’
failure line.
362
−𝜎3′ 𝑠𝑖𝑛𝜙 ′
=
1 − 𝑠𝑖𝑛𝜙 ′
So, next we will come to t, s, s’ plot. Now, we have already discussed this in our previous
lecture. If you remember the stress representation in Module 2, we have touched upon MIT
plot. And we have derived the relationship between MIT plot and Mohr Coulomb failure
envelope parameters. So, I will not take much time in this particular lecture. For compression,
if you remember, we have obtained 𝑘𝑓𝑐 = 𝑡𝑎𝑛𝛼 = sinϕ′
Where alpha is the inclination of origin of the line that joins the origin to the maximum shear
stress points because here c’ is 0. If you join all the maximum shear stress point that line
extended that will give a’ is the intercept and 𝛼 is the angle. So, if cohesion is present, so a will
also be present. If cohesion is not present, a will not be present. But the inclination of the failure
line is 𝛼 and 𝑡𝑎𝑛𝛼 = sinϕ′
For compression, we have already derived. So, this is how it looks like t, s, s’ and we know
that this is the isotropic line kf compression. Here 𝛼 is the angle. So, here the inclination is
equal to sinϕ′ or the slope is equal to sine phi’. Now, what about for extension? For extension,
we have not derived. So, we will see that what will be the slope for extension. For that
𝜎𝑎′ − 𝜎𝑟′
𝑡=
2
363
Now, for the extension, σa’= σ3’. Now, this is some important points or the minor points which
we have to keep in mind. We need to know what stress is acting in axially or radially. So, in
this case σ3’ is σa’. So,
𝜎3′ − 𝜎1′
𝑡=
2
If you take σ3’ outside, we have
1 + sin 𝜙 ′
𝜎3′ (1 − 1 − 𝑠𝑖𝑛𝜙 ′ ) −2𝑠𝑖𝑛𝜙 ′ 𝜎3′
𝑡= =
2 2(1 − 𝑠𝑖𝑛𝜙 ′ )
Similarly, s’
𝜎𝑎′ + 𝜎𝑟′ 𝜎3′ + 𝜎1′
𝑠′ = =
2 2
Same procedure, only thing is here the sign changes.
1 + sin 𝜙 ′
𝜎3′ (1 + 1 − 𝑠𝑖𝑛𝜙 ′ )
=
2
𝜎3′
=
𝑡
𝑘𝑓𝑒 = = −𝑠𝑖𝑛𝜙 ′
𝑠′
So, that will give −𝑠𝑖𝑛𝜙 ′ as the slope. So, we have defined the failure line in both t, s, s’ plot
for both compression as well as extension.
364
So, next is q, p, p’ plot for compression. Again, this particular slope, we have already defined
in Module 2 during stress representation. So, we have
𝜎𝑎′ + 2𝜎𝑟′ 𝜎1′ + 2𝜎3′
𝑝′ = =
3 3
𝜎3′ 1 + sin 𝜙 ′
= [ + 2]
3 1 − 𝑠𝑖𝑛𝜙 ′
𝜎3′ 3 − sin 𝜙 ′
= [ ]
3 1 − 𝑠𝑖𝑛𝜙 ′
𝑞 = 𝜎𝑎′ − 𝜎𝑟′ = 𝜎1′ − 𝜎3′
1 + sin 𝜙 ′
= 𝜎3′ [ − 1]
2𝜎3′ 𝑠𝑖𝑛𝜙′
=
𝑞 6𝑠𝑖𝑛𝜙′
𝑘𝑓𝑒 = ′
=
𝑝 3 − 𝑠𝑖𝑛𝜙 ′
365
Now, for extension,
𝜎𝑎′ + 2𝜎𝑟′ 𝜎3′ + 2𝜎1′
𝑝′ = =
3 3
𝜎3′ 1 + sin 𝜙 ′
= [1 + 2 ( )]
3 1 − 𝑠𝑖𝑛𝜙 ′
𝜎3′ 3 + sin 𝜙 ′
= [ ]
3 1 − 𝑠𝑖𝑛𝜙 ′
𝑞 = 𝜎𝑎′ − 𝜎𝑟′ = 𝜎3′ − 𝜎1′
1 + sin 𝜙 ′
= 𝜎3′ [1 − ]
−2𝜎3′ 𝑠𝑖𝑛𝜙′
=
𝑞 −6𝑠𝑖𝑛𝜙′
𝑘𝑓𝑒 = ′
=
𝑝 3 + 𝑠𝑖𝑛𝜙 ′
So, what we have done? for all these stress path plots now the bounds are defined. Now, in
whatever manner we need to study the stress path, we can do that.
366
So, summary of today’s lecture, this stress path is meaningful only if the bounds are defined.
The bounds are defined in terms of failure lines. And in this lecture, the failure lines are
determined for σa, σa’, σr, σr’ plot, t, s, s’, q, p, p’ for both compression and extension. So, that
is all for today’s lecture. In the next lecture, we will see, what are the stress paths for common
cases that we encounter?
The common type of stress variations, we will try to see the stress path. So, that is all for now,
thank you.
367
Lecture - 35
Stress path - some common cases
Welcome all of you. In the last 2 lectures, we have seen, what is a stress path? And, we have
defined the failure line. And, for all the 3 cases of stress path plotting, we have defined the
failure line and its slope. And, we have told that bounds are very important in understanding
the stress path variation. So, from today’s lecture onwards we will see some common cases for
stress path plotting first.
Then that will be followed by specifically for the triaxial testing in the lab and some examples
relevant to field conditions. So, that will be the progression of this module further. So, in
today’s lecture, we will see the stress path for some common situations. Obviously, you will
be having some knowledge now based on our previous lectures like how things would move
or how the stress path would move for some specific conditions.
We need to keep in mind that whenever we are doing the stress path what we have to understand
is we need to determine how the 2 different stress parameters changes with respect to each
other. And for that we may need some sort of computations. For example, in the case of σ a σr
variation, it is straight forward. Because that is the external stress that we are imposing onto
the soil mass.
So, it is very clear how σa and how σr if the variations are given it is very easy for us to plot.
But if it comes to st plot or qp plot the issue is now this becomes a function of this σ a and σr.
So, we do not know exactly how this stress path would move and plotting in terms of st plot
and qp plot becomes very important because they are more like stress invariants. And there are
several failure criterion which are based on this.
So, it is necessary for us to actually determine the variation even though the variation is known
to us we may have to derive. For example, the slope in which the stress path move. Because
unless we know the slope we will not be able to know when it will it is going to reach the
368
failure. So, all these aspects we will see in our few lectures from now. So, with that we will
start our first case.
That is stress path for isotropic consolidation. Now, without explaining much in this lecture
itself you will be in a position now to know how this stress path will move in a isotropic
consolidation case. Because we have started off with some examples in the beginning. But for
completeness let us again go through these because these forms some of the very important and
very unavoidable situations in geotechnical testing and in the field situations.
So, consolidation is done to simulate to the required stress history and stress state before
shearing. All of us know if it is a triaxial testing we try to simulate some condition initially
with respect to that is what is available in the field. But in many cases when the testing is done
in the lab this particular aspect of how much the initial condition relates to field is generally
not checked to that extent. But that is invariably necessary.
We should have an overall understanding of what was the overburden pressure the soil was
subjected to in the past. I mean at least the overburden pressure because whether it has been
subjected to more stresses in the past. Sometimes, we may not be knowing. The reason is if a
hill is cut and the engineer in charge he does not know whether there was a hill initially then
he could not he will not be in a position to anticipate the possibility of over consolidation.
But at least looking at the subsurface investigation plot we can always understand what could
be the possible overburden pressure which the soil is subjected to from a particular sampling
369
point. And any stress range that we simulate in the lab the initial state which is less than this
would definitely give an over consolidated behavior of the soil. So, it is very important that we
should know what is the overburden and at least we should know how much more stresses the
soil will be subjected to in future. So, based on this we make a judicial selection of initial state
simulation in the lab. So, that is what we have to do. Now, consolidation helps us to attain this
particular state of soil. For example, the whatever stress history we want to simulate, maybe
we can load it and then unload it to create a certain amount of over consolidation.
Or, maybe we can consolidate and then start the test on that without any unloading. So, what
sort of stress history we want to simulate or what sort of initial stress condition we want to
simulate depending upon that we need to decide upon the consolidation. It is possible to do
isotropic or anisotropic consolidation. Even though both are possible, majority of the cases we
prefer isotropic consolidation because it is fairly simple and it can be executed well without
much error. Whereas it is also possible in a specific triaxial testing system that we can also
induce anisotropic consolidation and remember we have told this in the previous lecture. When
we have anisotropic consolidation, we should also expect some amount of shear stress also
developed within the soil.
Now, this is a typical case of initial isotropic consolidation and which is represented by point
A, A’. We will have all the stress path plots. First, we will start with σa, σa’, σr, σr’ Now, we
can draw this isotropic line. So, this is this part of stress path plotting is fairly simple because
we know that for isotropic consolidation all the points would lie on the isotropic line.
And this we have seen initially also in while we started this lecture. So, the first point the initial
isotropic consolidation point is A, A’. Now, this is a saturated soil sample initially isotropic
consolidated. Now, we will increment this stress by Δσ. So, the from the initial point A, A’, we
apply an incremental stress of Δσ to it. So, what happens? Failure line is added to stress path
plot.
So, that is first and foremost thing that we have to do because we know that without that the
stress path plot is not complete. So, we have added. This we have already discussed in the
previous lecture. Helps to identify what happens with respect to stress increment that we have
already seen. At time t = 0, so, now we will consider both situations. An instantaneous un-
drained condition that will get generated which corresponds to t time t = 0.
370
That is at a very small time immediately after the application of Δσ there will be a total stress
condition. So, we will just try to understand that even though it may not be relevant. But we
will try to understand how it will look like and then at time t = t when this pore water pressure
aspect is not there it will be a drained or an effective stress path. So, when time t = 0 Δσ is
taken by the pore water pressure.
And Δσ’ that is gain in effective stress 0 at time t = 0. That means nothing has happened to
effective stress but there is an increment of Δσ in the total stress condition. So, that is what
again it is on the isotropic line because it is a hydrostatic stress increment. So, the point B will
be here. And, where will be point B’? Since, there is no change it will be at its initial point
itself.
So, B’ will be here. And this, please remember the point B’ and B may not be relevant. It is
relevant only with respect to time t = 0, condition that is an instantaneous loading condition.
So, that means this will increase by Δσ = u, Δσ = u. That is why it is on the isotropic line. Now,
time t equal to t that means a lot of time has gone.
So, whatever we have to expect, we have to expect in terms of effective stress only. So, in that
case all the stress increment is now in terms of effective stress, pore water pressure is 0. And
this will be indicated by point C, C’ because there is no total stress condition here. If there is
total stress condition at time t equal to t then the only change is point C and C’ will not coincide
so, here in this case point C and point C’ are the same. So, if you actually look at the given load
increment of Δσ there is only a case of A’ and C’. But we have expanded it and shown both in
terms of effective stress path and total stress path. Now, if you see the total stress path will be
A B and the effective stress path will be A’ C’. Now, imagine that there is sort of pore water
pressure also happening then because of pore water pressure the points C and C’ will not be
same. So, that is the only difference in the case of effective stress path and total stress path. So,
isotopic stress increment is along 45 degrees line as it is visible. Now, imagine so this is another
important point which we have to keep in mind. We know this but in the form of stress path
we need to understand this.
Initially itself we have discussed that for failure in soil to happen and when I say failure it is
shear failure we need to have shear stress developed. If we want shear stress to develop then
371
the soil should be subjected to unequal stresses. This means that for isotropic stress there is no
failure possibility. Now that is also visible here now if from A you keep on increasing Δσ what
will happen is it will simply move along the isotropic line.
Now, for failure to take place, the stress path should deviate and move towards either of these
failure line either in compression or in extension. Now, this is not a possibility in the case of
isotropic consolidation case. And hence isotropic stress does not induce shear stress and shear
failure. This we have to keep in mind. So, that is all about isotropic consolidation. We will
quickly see this on t, s, s’ and q, p, p’ plot as well.
So, in t, s, s’, we have given k f compression and k f extension. For isotropic stress, Mohr circle
has 0 diameter. There is no Mohr circle for isotropic case because σ1’ and σ3’ is same. Initial
isotropic consolidation is A, A’. Isotropic stress increment of Δσ same case instantaneous
increment in Δis u. So, that is what is given B, B’. And then the total stress path is AB.
Time t equal to t, ESP follows B’ C’ as is given. The same thing marked on q, p, p’, A, A’, B,
B’, C, C’. So, that is all with respect to isotropic consolidation, a very simple case to
understand.
372
So, now let us move on to the next set of consolidation where we have one dimensional
consolidation. Now, the stresses are not same in the radial as well as in the axial direction. So,
that is the most important difference as we do one dimensional consolidation. So, in the case
of one-dimensional consolidation a special case you can call it as isotropic consolidation. This
is a more general case.
So, 1D consolidation, the ratio of radial to axial stress is given by coefficient of lateral earth
pressure at rest. Now, we presume that there is no lateral strain that is happening. And that is
exactly what is simulated in an oedometer. It has got an infinite confinement because of these
metal container or metal cylinder in which these metal ring to be very specific metal ring in
which this soil is placed.
So, we do not allow any lateral strain to happen. And, where do you find such a situation? If
the loading is happening over a large area then we presume that the amount of confinement is
good enough and there is no lateral strain that is happening in the field and that closely relates
to one dimensional consolidation. So, that is how we take it. Now, for one dimensional
consolidation as it is already told.
The ratio of radial to axial stress is defined by coefficient of lateral earth pressure at rest because
there is no strain that is happening. Had it been a passive active then there is a moment. So,
then it will deviate. So, presently for this particular lecture and for subsequent analysis we will
have the use of coefficient of earth pressure at rest which is k0. And k0 = σr’/σa’.
373
Please note it always has to be in terms of effective stresses. I think we have already discussed
this during while explaining the sampling of soil. For NC state of the soil, k0 is specifically
equal to k0NC. Now, for both NC and OC, the k0 parameter would change. Now, according to
Jacky 1944 which is a very popular equation 𝑘0𝑁𝐶 = 1 − 𝑠𝑖𝑛 𝜙′
You will find different forms of this equation in the literature. For me, in this particular lecture
that is not important. What is important is how the radial and the axial stress varies. Now, for
that I have taken a very simple case of Jacky’s equation which is 𝑘0𝑁𝐶 = 1 − 𝑠𝑖𝑛 𝜙′ . And
this is applicable for a certain range of 𝜙′. I am not going into those details. Even you will find
different forms of this equation in the textbooks.
But we will just consider this particular one. So, we need some sort of value for k0NC. And this
is very prominent equation that people use. So, now, let us say that the sample is initially
normally consolidated. Now, let us get on to the stress path. So, what change will it have? So,
now this is an isotropic consolidation line. Initially our point was located here. Now, what will
be the change in A, A’? Should we plot it here?
No, because it is governed by this ratio k0NC. So, let us put the failure line. And now this is k
NC which is the line corresponding to normally consolidation, so that k NC. So, where should
A, A’ fall? Definitely it is normally consolidated initial point. So, it has to fall on k NC. So, A,
A’ is on k NC line. Now, slope of kNC is σa’ upon σr’ which is the reciprocal of this reciprocal
of k0. So, it will be
𝜎𝑎′ 1
𝑘𝑁𝐶 = =
𝜎𝑟 ′ 1 − 𝑠𝑖𝑛𝜙 ′
So, this will be the slope of kNC line. Stress increment, let us say that it is Δσa is incremented
by Δσ. So, what happens at time t = 0 again the concept of instantaneous pore water pressure
happens and hence there will be point B and B’. Now, remember this is hydrostatic stress
condition when it comes to pore water pressure at time t = 0 this you have to keep in mind
because it is not relevant but we are discussing it. So, it will help us to understand. When there
is an immediate or for an un-drained condition when you are in incrementing the stress if you
have an un-drained condition how this stress path would move. So, this is very important. So,
that is why time t = 0 is discussed. Now, what would be the possibility of the point A, B.
374
Since, it is hydrostatic, definitely the stress path should be parallel to isotropic line because that
is the hydrostatic line. Hence, this from AB will be at 45 degrees with whatever is the stress
increment. Now, here both = Δσ = Δu. Now, at time t equal to t that is when there is no issue
of pore water pressure effective and total stress path is same, what will happen?
The point will lie on the normally consolidated line itself and that points will be C, C’. So, what
happens is there is no point which is joining from here to here because that is not what we are
simulating. It is not a stress path which is going from A to B and B to C. AB is only a concept
which has to be described here. Actual stress path is A, A’, C’. So, that is what it means and
then at time t equal to 0 TSP is AB and time t equal to t, ESP follows B’ C’ or A’ C’.
So, I am just showing an enlarged version of kNC line A, A’, AB and C, C’ is falling on this.
Just want to make clear this particular point here AC or A’ C’ this vertical distance will be Δσa’
= Δσ. Because we have incremented Δσ but what will be the horizontal distance I mean this.
So, that will be k naught NC times Δσa’.
So, that is governed by the slope of k NC. So, here Δσr’ = kONC Δσa’. So, that we have to keep
in mind. So, what ultimately we learned from this exercise is that in stress path we need to
determine the slope with which this stress path will progress and how that slope compares with
the slope of failure line. So, then we will be able to understand in what manner and in what the
distance of this stress path the failure is going to take place.
375
The same thing is repeated for the t, s, s’ plot. So, I will not spend much time here. We will
just quickly glance through it, same the A, A’ B. That is 𝛥𝜎𝑎 = 𝛥𝜎. And the final stress path
is A’ C’. Now, here we need to understand that this relationship is known and it was fairly
simple for us to determine in σa, σr plot. But here the slope will be slightly different in the
case of kONC line. So,
𝐴𝑡 𝑡𝑖𝑚𝑒 𝑡 = 0, 𝛥𝜎 = 𝛥𝑢(𝐵)𝛥𝜎 ′ = 0(𝐵 ′ )
𝐴𝑡 𝑡𝑖𝑚𝑒 𝑡 = 𝑡, 𝛥𝜎𝑎′ = 𝛥𝜎, 𝛥𝑢 = 0(𝐶, 𝐶 ′ )
𝛥𝜎𝑟′ = 𝑘𝑂𝑁𝐶 𝛥𝜎𝑎′
𝛥𝜎(1 − 𝑘𝑂𝑁𝐶 )
𝛥𝑡 =
2
𝛥𝜎(1 + 𝑘𝑂𝑁𝐶 )
𝛥𝑠 ′ =
2
𝛥𝑡 (1 − 𝑘𝑂𝑁𝐶 )
=
𝛥𝑠 ′ (1 + 𝑘𝑂𝑁𝐶 )
The slope is given by Δt/Δs’. So, determining the slope is important in stress path. So, we
need to understand what exactly is happening when I say isotropic consolidation one
dimensional consolidation and so on. Substituting 𝑘𝑂𝑁𝐶 = 1 − sin 𝜙 ′ , the slope can also be
written as
𝛥𝑡 𝑠𝑖𝑛𝜙′
′
=
𝛥𝑠 2 − 𝑠𝑖𝑛𝜙′
376
The same is repeated for q, p, p’. The points are not important because now we are converse
in with that. We will just see, what is the slope of kNC?
𝑆𝑡𝑟𝑒𝑠𝑠 𝑖𝑛𝑐𝑟𝑒𝑚𝑒𝑛𝑡 𝛥𝜎𝑎 = 𝛥𝜎
𝐴𝑡 𝑡𝑖𝑚𝑒 𝑡 = 0, 𝛥𝜎 = 𝛥𝑢(𝐵)𝛥𝜎 ′ = 0(𝐵 ′ )
𝐴𝑡 𝑡𝑖𝑚𝑒 𝑡 = 𝑡, 𝛥𝜎𝑎′ = 𝛥𝜎, 𝛥𝑢 = 0(𝐶, 𝐶 ′ )
𝛥𝜎𝑟′ = 𝑘𝑂𝑁𝐶 𝛥𝜎𝑎′
𝛥𝑞 = 𝛥𝜎(1 − 𝑘𝑂𝑁𝐶 )
𝛥𝜎(1 + 𝑘𝑂𝑁𝐶 )
𝛥𝑝′ =
3
𝛥𝑞 3(1 − 𝑘𝑂𝑁𝐶 )
=
𝛥𝑝′ (1 + 2𝑘𝑂𝑁𝐶 )
𝑆𝑢𝑏𝑠𝑡𝑖𝑡𝑢𝑡𝑖𝑛𝑔 𝑘𝑂𝑁𝐶 = 1 − sin 𝜙 ′
𝛥𝑞 3𝑠𝑖𝑛𝜙′
=
𝛥𝑝′ 3 − 2𝑠𝑖𝑛𝜙′
So, all these things are repetitive. So, you can just go through it and you can understand it.
377
we have completed isotropic consolidation one dimensional consolidation. Now, we will see
the next case which is stress path for unloading. What is the relevance of unloading? The
relevance of unloading that is that it creates an over consolidated state. I hope if you recall the
oedometer test, we do the loading and then we will do staged unloading. What is the relevance
of staged unloading?
Because we are also interested in getting the slope of swelling reloading line and that state
represents over consolidated state. We have already discussed this in our previous lecture. So,
how or what is the relevance of stress path with respect to unloading or over consolidated state.
Now, let us again start with σaa’. The normally consolidated state is represented by C, C’. That
is where we ended up with. So, all these things are there.
So, now C, C’ is the normally consolidated point. Now, the soil is unloaded by decreasing σ a.
So, what we are doing we are releasing the axial stress σa is released. Now, what will happen
if we release σa? So, we are starting with C, C’ the point. When it is released because it has
started from A to C’, does it mean that when σ a is released that it will trace back kNC line.
It will not and this information we already know because we know this from oedometer test as
well. So, when we load soil to a particular point and then unload it, it is not going to trace back
rather it takes a different route and that is what we known as the swelling line is the same aspect
which is happening here as well. The unloading stress path is in the direction of C’ D’. So, the
unloading will be here.
378
Now, this unloading will be restricted by the presence of kfe. So, it does not mean that you can
unload the soil the way you want, there will be formation of tension cracks and the soil is not
going to be existent. And that is defined by the bound is defined by kfe. Now, unloading what
have what does unloading do to the soil? Unloading induces over consolidated state into the
soil. Now k naught equal to kONC is not valid.
And that is very relevant from this plot also. Only those points on kNC line, it is valid. The
moment it deviates it is not valid. So, kO = kONC is not valid for the unloading line. Now, the
information that we have is that over consolidated state of the soil exhibits elastic behavior.
Why? Because when you load to a certain point, we know what is the pre-consolidation
pressure?
Now, pre-consolidation pressure, we have seen it is also stated as yield stress. Now, once it is
unloaded from it then again reloading till the point it reaches pre-consolidation it is bound to
have better stiffness and hence it will exhibit more like an elastic material elastic behavior. It
will exhibit elastic behavior. So, you can safely assume that the over consolidated soil behaves
elastically.
So, over consolidated soil, it is normally reported that 𝑘𝑜 = 𝑘𝑂𝑁𝐶 (𝑂𝐶𝑅)𝑚 . Again, there are
several relationships which are reported in the literature. I am just selecting this which is
generally found in literature. And the value of m is in the range of 0.4 to 0.5 or m is taken equal
to 𝑜𝑟 𝑚 = 𝑠𝑖𝑛𝜙 ′ 𝑜𝑟 𝑚 = 1.25𝑠𝑖𝑛𝜙 ′ . So, different researchers have proposed different type of
relationship for m and these are the ones which are generally considered. And hence kO for OC
soil will depend upon OCR.
379
𝛥𝜎𝑟′ µ′
=
𝛥𝜎𝑎′ 1 − µ′
So, we have understood that we need to understand in what manner this C’ D’ is going to move.
It can have multiple possibilities but where exactly it is going to move that will depend upon
the elastic behavior of the soil. So, that particular assumption is very important to understand
the slope of C’ D’. So, elastic isotropic behavior is assumed during unloading. The logic is
because it has become an over consolidated state.
For one dimensional, lateral strain is 0. This we have already discussed. So, 𝛥𝜖𝑟′ = 0. And we
also have the relationship
− ′ (𝛥𝜎𝑟′ + 𝛥𝜎𝑎′ ) = 0
𝐸′ 𝐸
𝛥𝜎𝑟 µ′
′
′
= ′ (𝛥𝜎𝑟′ + 𝛥𝜎𝑎′ )
𝐸 𝐸
So, this gives the relationship of
𝛥𝜎𝑟′ (1 − µ′ ) = µ′ 𝛥𝜎𝑎′ .
So, this will move with respect to this particular ratio defined by the Poisson’s ratio of the
soil.
380
So, reloading the heavily over consolidated from point D’. Now, if you happen to reload from
this particular point D’, now let us say that based on the location of point D’ the soil is heavily
over consolidated. HOC means heavily over consolidated. Now, if you are reloading it from
this point D’ then the effective stress path for reloading it quickly approaches the kNC line. That
is the ESP approaches k NC quickly to the point E’.
So, this is what it means. Now, in oedometer test, we have idealized it to be a straight line. So,
loading, unloading and reloading it is traces the same path. But we know that in reality that is
not going to happen it is not the same path. So, that difference is shown clearly in this stress
path. So, when you reload it, it is going to approach the kNC line very quickly. And that reaches
at point E’. It is not joining at C, C’. Please remember that.
So, point E’ is still over consolidated. So, that means when it reaches E’ that point is still under
over consolidation even though it has approached the normally consolidated line. But after E’
when it is subsequently loaded it will move along the NC line itself. So, that is the only
difference. So, instead of joining C, C’, it joins at E, E’. It approaches the NC but then we have
to remember that the point E’ is still OC. Why?
It has not reached the point C, C’. It is still less than the stresses at C, C’. Now, from point E’
onwards kO = kONC.
381
The same thing is plotted for t, s, s’ plot. So, that we know this relationship now,
=
𝛥𝜎𝑎′ 1 − µ′
We have to use this for finding out what will be the slope in t, s, s’ plot. So, this Δt = Δσ a’.
This now by this time you are already converse in with this formulation. So, I will not spend
much time here.
𝛥𝜎 ′
𝛥𝜎𝑎′ (1 − 𝛥𝜎𝑟′ )
𝑎
𝛥𝑡 =
2
Substituting for Δσr’ this ratio here we have
µ′
𝛥𝜎𝑎′ (1 − 1 − µ′ )
𝛥𝑡 =
2
𝛥𝜎𝑎′ (1− 2µ′ )
𝛥𝑡 =
2(1 − µ′ )
𝛥𝜎 ′
𝛥𝜎𝑎′ (1 + 𝛥𝜎𝑟′ )
𝑎
𝛥𝑠′ =
2
µ′
𝛥𝜎𝑎′ (1 + 1 − µ′ )
𝛥𝑠′ =
2
382
𝛥𝜎𝑎′
𝛥𝑠′ =
2(1 − µ)
𝛥𝑡
= 1 − 2µ′
𝛥𝑠 ′
So 𝛥𝑡/ 𝛥𝑠 ′ is the required slope. So, what turns out? Again, we are determining the slope in st
plot. For any stress path plotting, this is what we have to do. Depending upon the variation, we
need to determine the slope. So, this how it looks like is the same as that of σa σr’ plot.
When we repeat this for q, p, p’, we will have

=
𝛥𝜎𝑎′ 1 − µ′
µ′
𝛥𝑞 = 𝛥𝜎𝑎′ (1 − )
1 − µ′
383
𝛥𝜎𝑎′ (1 − 2µ′ )
𝛥𝑞 =
1 − µ′
2µ′
𝛥𝜎𝑎′ (1 + 1 − µ′ )
𝛥𝑝′ =
3
𝛥𝜎𝑎′ (1
+ µ′ )
𝛥𝑝′ =
3(1 − µ′ )
So, this will give this slope
𝛥𝑞 (1 − 2µ′ )
=
𝛥𝑝′ 3(1 + µ′ )
And the slope of the unloading line will look like C’ D’ and D E’. So, that is how it looks like.
So, in all the cases we need to determine the slope of the stress path and that is exactly what
we have to do. So, these are some of the common cases like isotropic one dimensional
consolidation and unloading which is over consolidation. There are a few more cases that we
need to discuss and that we will see in the next lecture.
For the time being, let us stop here. And we will see some more cases in the next one and then
that will be followed by the triaxial testing. So, that is all for now, thank you.
384
Lecture - 36
Stress path-some common cases – 2
Okay, welcome back all of you. In the last lecture, we were discussing about stress paths and
some common cases. We have seen isotropic consolidation, one dimensional consolidation, the
unloading of one-dimensional consolidation. That means the kind of OC state that gets
generated. And when an OC soil getting again reloaded, what happens? So, these stress paths
we were discussing in the last lecture.
Now, this lecture is a continuation of that, we will discuss some more examples related to stress
path. And this will help us understand how we need to plot or how we need to represent this
stress path for different cases.
𝛥𝜎𝑎 = 𝛥𝜎𝑟
𝛥𝑠 ′ = 𝛥𝜎𝑎′
So, we will just see the first one. Again some examples of stress path, let us start the initial
condition. It is an isotropic consolidation, where 𝜎𝑎 = 𝜎𝑟 . And then we will discuss some
specific cases of loading or unloading. So, it can be either loading unloading or loading
unloading in the radial direction. So, different stress paths, we will just consider t, s, s’ stress
path plot.
385
All the procedure remains same. But, for convenience, we are just referring to only one stress
path now. So, initial condition is given by Po which is the isotropic consolidation condition.
Now, when I say this it is very easy now for all of you to identify where the point would be on
the isotropic line. So, let this be the initial point Po. Now, the second one is P1 which represents
isotropic loading.
Please take this as a revision, because we have been discussing this quite a number of times.
So, you will be knowing now without my help like in what direction it should move. It should
definitely move along the isotropic line. So, P1 is fairly easy and why it is so, because
𝛥𝑡 = 0 and 𝛥𝑠 ′ = 𝛥𝜎𝑎′ . You will get this without any difficulty. So, P1, how will we represent
P1?
P1 is represented by the arrow along the isotropic line. So, you keep on increasing the isotropic
stress condition, the arrow keeps on progressing.
So, that is about 2 cases. The third case is P2 which is given by there is no increase in radial
stress Δσr = 0. And Δσa that is axial stress increases. Now, if this is the case on σa σr plot it is
very easy to plot. But when it comes to st plot and qp plot which are mandatory because they
are invariants. So, identifying this stress path needs some sort of mathematical understanding.
So, let us see what this translate to in t, s, s’ plot. You need to first understand what will be the
increment in 𝛥𝑡.
386
𝛥𝜎𝑎′ − 0 𝛥𝜎𝑎′
𝛥𝑡 = =
2 2
So, here Δσa increases. 𝛥 σa’ minus there is no increment in a radial stress so it is 0 by 2. That
𝛥𝜎𝑎′
gives plus Δ . Now, you need to keep certain important things in mind. One is whether it is
2
increasing or decreasing whether the radial stress or axial stress increases or decreases.
Now, if it is increasing then it is positive. And if it is decreasing it is negative. So, that is one
part. Now, whether this increase, decrease or the relative increase, decrease of both axial and
radial what it will result in terms of t and s. Whether t will be positive, s will be positive with
whether there can be possibility of both being positive and negative. All these aspects need to
be taken into account while understanding the stress path.
So, prima facie it looks a bit simple. But, there are several small small nitigrities that can hinder
your understanding about stress path. So, we need to be careful about these small aspects as
well. So, here Δt is known. Now, what is the next task? Δs’
𝛥𝜎𝑎′ + 0 𝛥𝜎𝑎′
′
𝛥𝑠 = =
2 2
𝛥𝜎𝑎′
Same thing you get . So,
2
𝛥𝑡
=1
𝛥𝑠 ′
So, we know that it should be with a slope of 1. That means at a 45 degree.
So, how do we plot P2? So, P2 will be at an inclination of 45 degrees from the initial point. Next
is P3 which is given as Δσa is = 0. There is no change in axial stress. And there is a decrease in
radial stress. That means Δσr will be - Δσr (because this is decreasing). Now, when you
substitute this in the same equation of Δt and Δs’, what we will get.
(0 − (−𝛥𝜎𝑟′ )) 𝛥𝜎𝑟′
𝛥𝑡 = =
2 2
(0 + (−𝛥𝜎𝑟′ )) −𝛥𝜎𝑟′
𝛥𝑠 ′ = =
2 2
𝛥𝑡
= −1
𝛥𝑠 ′
387
𝛥𝑡
So, that will give = −1. So, now, how P3 will translate. So, P3 is also with a slope of 1 or
𝛥𝑠′
at an inclination of 45 degrees.
But here we need to understand that this inclination is negative. So, that is why it is towards
this direction. Now, another aspect is whether Δt is positive Δt increment is positive. Again an
important thing which we need to understand is the actual value of t and s’. That is t s’ and the
increment now increment is Δt Δs’. Now, we are talking about increment because we are just
talking about the slope.
But, what can be the actual value? That also matters. That will be t ±Δt or s, s’±Δs’. So, what
will be the net? Only the final computation we will be able to understand. So, here Δs’ is going
towards in the negative direction whereas, Δt is moving towards in the positive direction. So,
if you can see here from this particular point s’ negative direction is in this direction.
So, that is why it is moving in this direction whereas Δt is in the positive direction that is in the
upward direction. It can be like this as well. From Po, it can come down. But then accordingly
then Δs’ will be positive, Δt will be negative. So, all these aspects we need to keep in mind.
What is the direction of increment or decrement, whether it is increment or decrement in the
respective stress parameter.
Now, increment decrement in Δσa, how it translates to these stress parameters like t and s, s’.
Both aspects, we need to see. So, here it is clear it is - 1. So, slope has to be negative. That we
know, but in what direction. Well it can in this whether it is downwards or upwards. So, that
will information will get from here. So, here Δs’ is towards negative s direction and Δt is
towards positive t direction.
388
So, in the next example, it is P4 where 𝛥𝜎𝑟 = −𝛥𝜎𝑎 . So, here again substituting this
𝛥𝜎𝑎′ − (−𝛥𝜎𝑎′ )
𝛥𝑡 = = 𝛥𝜎𝑎′
2
𝛥𝜎𝑎′ + (−𝛥𝜎𝑎′ )
𝛥𝑠 ′ = =0
2
𝛥𝑡
=∞
𝛥𝑠 ′
So, Δt upon Δs’ is = infinity because there is no change in Δσa’. So, how this will move, there
is no change in Δs’, but an increment in Δt is happening. So, it will be the P4 failure or this
stress path is in this direction P4. P4 means it is vertically moving upwards. This means that
there is no change in Δs, but only change in Δt. So, that is what is meant by Δt by Δs’.
That is infinity. That is keep on increasing. But this increase, please remember all these
increases are restricted by the failure line on compression and extension part. That also we need
to keep in mind. Those failure part, we will come a bit later. Now, the next one is P5 where we
have Δσr is increasing that is plus Δσr and Δσa decreases that is - Δσa. Now, how this will
translate to Δt.
(−𝛥𝜎𝑎′ − 𝛥𝜎𝑟′ ) −𝛥𝜎𝑎′ − 𝛥𝜎𝑟′
𝛥𝑡 = = (−𝑣𝑒)
2 2
′
(−𝛥𝜎𝑎′ + 𝛥𝜎𝑟′ ) −𝛥𝜎𝑎′ + 𝛥𝜎𝑟′
𝛥𝑠 = = (+𝑣𝑒 𝑜𝑟 − 𝑣𝑒)
2 2
𝛥𝑡 −𝛥𝜎𝑎′ − 𝛥𝜎𝑟′
=
𝛥𝑠 ′ −𝛥𝜎𝑎′ + 𝛥𝜎𝑟′
389
Δs’ can be positive or negative depending upon the relative magnitudes of σa’ and σr’. So,
accordingly this can be positive or negative. Now, this is always negative. This can be positive
or negative. So, Δt/ Δs’, the expression is known and how this is going to affect our stress path.
So, P5, how do we plot? One important aspect is Δt is always in the negative direction. So,
from this particular point, t negative is in this direction. So, that is moving towards this, the
arrow is moving towards this. Why?
Because t is negative and that is in this downward direction whereas in this particular case, Δs’
is considered to be positive. So, this has to move towards the positive s direction. So, that is
how it is written. Now, what will be the actual inclination of P5 that will depend upon what is
the type of magnitude of these stresses? So, that will determine in what variation what is the
actual magnitude of the increment that will determine the slope?
1
So, the next one is P6 where 𝛥𝜎𝑟 = 2 𝛥𝜎𝑎 . And this is another aspect. There is another way in
which the stress increments are changing

(𝛥𝜎𝑎′ − (0.5𝛥𝜎𝑎′ ) 𝛥𝜎𝑎′
𝛥𝑡 = =
2 4
𝛥𝜎𝑎 + (0.5𝛥𝜎𝑎 ) 3𝛥𝜎𝑎′
′ ′
𝛥𝑠 ′ = =
2 4
𝛥𝑡 1
′
= 𝑔𝑖𝑣𝑒𝑠 18.40
𝛥𝑠 3
390
So, accordingly P6 can be plotted. Both are in the positive direction. That is upwards and
towards this, so here P6 with an inclination of 18.4 degrees.
So, these are some of the examples wherein the starting point was isotropic consolidation that
is Po Now, let us see a few more cases wherein the starting point is not isotropic consolidation.
So, here we have anisotropic consolidation where 𝜎𝑎 ≠ 𝜎𝑟 , again loading unloading. The
same t, s, s’ plot is considered. So, the first initial condition is Po which is given by
𝜎𝑎 ≠ 𝜎𝑟 (𝜎𝑎 > 𝜎𝑟 ) , a typical triaxial compression case. If that is the case, then the point will
be in the upper half, where Po is the starting point, so this Po.
Now, why it remains in the upper half? Again this will become even more clearer when we
discuss the next slide on k value how it is changing. So, here let us say that this is in this
particular point Po. Now, P1 is Δσa decreases. That is (−𝛥𝜎𝑎 )𝛥𝜎𝑟 = 0. So, how this will
look like.
(−𝛥𝜎𝑎′ − 0) −𝛥𝜎𝑎′
𝛥𝑡 = =
2 2
′
(−𝛥𝜎𝑎 + 0) −𝛥𝜎𝑎′
′
𝛥𝑠 = =
2 2
𝛥𝑡
=1
𝛥𝑠 ′
391
𝛥𝑡
So, how P1 will look like. Now, please note here, = 1 and it is positive. So, our general
𝛥𝑠′
tendency is to draw it in the upward direction. For example, I may draw like this, because it is
a positive slope.
So, here, we also need to see, it is negative Δt is pulling it down because it is negative Δσa’/2.
Δs’, that is also in the negative direction. So, that qualifies this P1 stress path. So, t is also
coming down, s is also moving in the negative direction. So, you need to keep in mind both
this variation and the final slope as well.
The next stress path is Δσa increases that is positive Δσa and Δσr decreases - Δσr. So, if you
substitute this
𝛥𝜎𝑎′ − (−𝛥𝜎𝑟′ ) 𝛥𝜎𝑎′ + 𝛥𝜎𝑟′
𝛥𝑡 = = (+𝑣𝑒)
2 2
𝛥𝜎𝑎′ + (−𝛥𝜎𝑟′ ) 𝛥𝜎𝑎′ − 𝛥𝜎𝑟′
𝛥𝑠′ = = (+𝑣𝑒 𝑜𝑟 − 𝑣𝑒)
2 2
Now depending upon the relative magnitude of these two it can be positive or negative. So we
have
𝛥𝑡 𝛥𝜎𝑎′ + 𝛥𝜎𝑟′
=
𝛥𝑠 ′ 𝛥𝜎𝑎′ − 𝛥𝜎𝑟′
So, how P2 will look like in the stress path plot. So, here we are considering P2 that is t it is in
the upward direction because it is always positive. And here Δs’ is considered negative. So,
that is why it is going towards the negative direction. So, both ways it is possible. It is possible
392
in the positive direction as well. It is positive in it is possible in the negative direction as well.
And, what will be the actual slope of P2? That will depend upon the relative magnitude.
Next is P3 where Δσa decreases and Δσr increases.

(−𝛥𝜎𝑎′ − 𝛥𝜎𝑟′ )
𝛥𝑡 = (−𝑣𝑒)
2
′
(−𝛥𝜎𝑎′ + 𝛥𝜎𝑟′ )
𝛥𝑠 = (+𝑣𝑒 𝑜𝑟 − 𝑣𝑒)
2
The slope expression is
𝛥𝑡 −𝛥𝜎𝑎′ − 𝛥𝜎𝑟′
=
𝛥𝑠 ′ −𝛥𝜎𝑎′ + 𝛥𝜎𝑟′
So, how P3 will look like. So, P3 will look like it moves in the downward direction because t
tends to negative value. So, that is why it is coming down and here it can be positive or negative.
So, here we have considered Δs’ to be in the positive direction. So, that is why it is moving
towards rightwards. So, and the actual slope will depend upon the relative magnitudes of the
stress increment.
𝛥𝑡 1−𝑘
′
=
𝛥𝑠 1+𝑘
𝑘 < 1, 𝜎𝑎′ > 𝜎𝑟′
𝑘 > 1, 𝜎𝑎′ < 𝜎𝑟′
So, that is about some examples to make us understand how we need to plot the stress path
direction. And probably depending upon what direction the stress path is, we will be able to
393
understand when this stress path is going to touch the failure line. If at all that kind of stress
increment decrement is going to fail, all these information becomes quite handy from the stress
path plot.
More of that we will follow because the actual application of when it is going to fail and how
it is going to fail is in the Module 4, when we discuss about critical state. So, here we understand
the philosophy of stress path plotting for Module 4. Now, as I told you in the previous slide
that what is the importance of k. Now, we have seen different types of k. The k for isotropic is
1. Then we have seen k for normally consolidated condition, k for failure condition. So, all
these things we have seen now. k for unloading case. All these things we have seen. Now, what
will what is the actual importance of k? So, Δt /Δs’, again I am referring to this. We can do this
with any other stress parameter plots. So, this is k = 1 which is isotropic line. Now, for k < 1,
what is k?
k is σr’ upon σa’. So, k < 1 means σa’ is > σr’. So, k > 1 that is a case of maybe triaxial
compression where σa’ > σr’. It comes in this particular plot. So, this is k = 1. All conditions
where k is < 1 that means σa’ > σr’ will come in this particular domain. So, it is in the
compression domain. So, that is k < 1.
And please remember, this is initial condition mostly because where it falls in the anisotropic
consolidation. We are referring to that. Now, for k > 1 that means σa’ < σr’ we obviously know
it has to be < the bottom domain. So, this is k > 1. Now, if we know any point on the k line, let
us say we know any point on the k line it is possible to determine σa’ and σr’ graphically. Why?
Because each of these points in t, s, s’ plot is maximum shear stress point. I hope you remember
that. So, let us say that we know the point P. We want to determine, what is the σa’ and σr’
corresponding to this particular point? Now, for us in the triaxial sample, σa’ and σr’
corresponds to major and minor principal stress conditions.
So, here if I draw a horizontal line like this, and then draw 2 lines which is at 45 degrees or
maybe 45 degrees with this particular line, either way here and at the meeting point considering
these 2 meeting points, if we draw a circle, then these points will represent σa’ and σr’. So, that
is how we determine. If you know a point you can easily draw a circle there.
394
Both ends will give you σ1’ σ3’ which is nothing but our axial and radial stress condition in a
triaxial sample.
Now, there is a specific case of stress path for sedimentation. We have seen this. We have
already discussed this in our previous lecture. I am just trying to map that understanding with
the actual condition that can happen in the field. Soils are deposited in low lying areas, lake or
sea. That is how the evolution of soil takes place. The river carries it deposits at some particular
place, then this deposition keeps on going.
Now, this is a virgin deposition. And hence, any sort of consolidation we know that is going to
be normally consolidated sample. So, there is a gradual buildup of overburden as the deposition
keeps on going. And because of this the overburden pressure increases as the deposition
increases. Sediments consolidate and decrease in volume because as the deposition goes on the
overburden keeps on increasing.
So, self-weight consolidation happens densification happens, the volume decreases. So, this
process keeps on going and that is how the sedimentary soils are produced. If the area of
deposition is large, it satisfies the one dimensional criteria. This also we have seen. As the area
is large then the one dimensional consolidation condition is satisfied. It also is a normal
consolidation. Why?
Because, these are freshly laid deposits of soil and hence the consolidation will be will adhere
to the normally consolidation. So, this stress path would be essentially OA. So, now we know
395
that this is normal consolidation. So, we have to have a normally consolidation line. So, this is
k NC line which we have already discussed in the previous lecture. So, how will be OA? So,
OA will be it is starting from some point.
And then it is increasing and this follows the normally consolidation line as the sedimentation
keeps on increasing overburden keeps on increasing the sedimentation keeps on happening. So,
that will reach to OA. Now, during sampling or removal now, we have reached the point A.
Now, what we do, if there is any subsurface investigation, we need to sample it.
So, during sampling, what we are doing essentially we are removing the overburden. Now, this
removal of already deposited sample can also be due to anthropogenic reasons, because of
clearing of that particular area for construction or it can be natural reasons because of again
erosion happening. So, any of these factors can ease out the deposit. So, during sampling or
removal of overburden, because of anthropogenic activities erosion, the overburden pressure
is removed.
So, what will happen? This results in a stress path which follows AB. So, from A, there is a
release which is happening because of which there is a downward trend. So, AB represents the
removal of the overburden. So, on further removal of overburden, soil would further become
OC. So, now, the moment it is unloaded this portion represents over consolidation. Now if
further removal happens then it may further go down.
And soil would become OC and the stress path moves even below the isotropic line. So, which
I am not showing we have just stopped here up to isotropic line.
396
Now, we will see something what is known as elastic stress path essentially which confirms to
the elastic behaviour. So, here again the plot is same t, s, s’. When a soil is sheared, effective
stress path, ESP may be predicted assuming elastic behaviour until the yield point is reached.
This is possible, very much possible. So, we tend to have the elastic behaviour up to the yield
point. And this aspect will be again discussing in the next module.
We know that what are the elastic strains where

𝛥𝜎1′ µ′
𝜖1′ = − (𝛥𝜎2′ + 𝛥𝜎3′ )
𝐸′ 𝐸
𝛥𝜎2′ µ′
𝜖2′ = ′ − (𝛥𝜎1′ + 𝛥𝜎3′ )
𝐸 𝐸
′
𝛥𝜎3 µ′
𝜖3 = ′ − (𝛥𝜎2′ + 𝛥𝜎1′ )
′
𝐸 𝐸
Let us say a typical case of un-drained triaxial compression test where because of triaxial
compression we know that
𝜖1′ = 𝜖𝑎′ 𝛥𝜎1′ = 𝛥𝜎𝑎′
𝜖2′ = 𝜖3′ = 𝜖𝑟′
Again because of axisymmetric condition 𝛥𝜎2′ = 𝛥𝜎3′ = 𝛥𝜎𝑟′
397
𝛥𝑝′ = 0
𝜖𝑎′ can be written as

𝛥𝜎𝑎′ 2µ′ 𝛥𝜎𝑟′
𝜖𝑎′
= −
𝐸 𝐸
′ ′
𝛥𝜎𝑟 µ
𝜖𝑟′ = − (𝛥𝜎𝑎′ + 𝛥𝜎𝑟′ )
𝐸 𝐸
Now, what will be volumetric strain? The volumetric strain is 𝜖𝑣′ = 𝜖𝑎′ + 2𝜖𝑟′ . This is for
triaxial sample. Now, we know for un-drained condition 𝜖𝑎′ = 0. So, the final expression is
1 − 2µ′
𝜖𝑣′ = (𝛥𝜎𝑎′ + 2𝛥𝜎𝑟′ )
𝐸
For un-drained condition, we know that there is no volumetric strain, so 𝜖𝑣′ = 0. Now, µ′ ≠ 0.5.
Please remember, in our earlier lecture, we have already discussed that µ corresponding to un-
drained condition is = 0.5 because it is an incompressible state. Now, please remember that mu
corresponds to µ un-drained. So, there it is = 0.5. You can conveniently take = 0.5 because a it
is a un-drained condition. Here, even though the condition is un-drained we are talking about
µ’.
That is the effective Poisson’s ratio which is not supposed to be 0.5. Now, in this expression,
we know 𝜖𝑣′ = 0, but that can happen only when µ’ is = 0.5. Under effective condition, µ′ ≠
0.5. So, then what is the possibility? The only possibility is (𝛥𝜎𝑎′ + 2𝛥𝜎𝑟′ ) = 0. So, this is a
particular condition. This means Δp’ is = 0.
398
What is p’? p’ is the mean stress. What is mean stress? (Δσa’+2σr’)/3 which is = 0. Now, if you
plot it in q p’ plot, this particular stress path will look like something like this. So, starting point
there is no change in Δp’. So, q/p’ will be infinity. So, it moves vertically upwards. How this
will be in t, s, s’ plot.
So, we will see in t, s, s’,

𝛥𝜎𝑎′ − 𝛥𝜎𝑟′
𝛥𝑡 =
2
𝛥𝜎𝑎 + 𝛥𝜎𝑟′
′
𝛥𝑠′ =
2
𝛥𝑡 𝛥𝜎𝑎 − 𝛥𝜎𝑟′
′
=
𝛥𝑠 ′ 𝛥𝜎𝑎′ + 𝛥𝜎𝑟′
So, Δt/Δs’ is known. Now, we know this condition, 𝛥𝜎𝑎′ + 2𝛥𝜎𝑟′ = 0. Substituting that in the
particular expression for Δσa’ putting 𝛥𝜎𝑎′ = −2𝛥𝜎𝑟′ We get the slope as
𝛥𝑡
=3
𝛥𝑠 ′
So, how does it look like in the t, s, s’ plot. So, if this is the starting point it goes in this direction.
So, one is in the upward direction the other one is towards the extension. So, this is the kind of
stress path we get.
399
Now, the soil specimen which follows elastic stress path will have different strength in
compression. That is tfc and extension tfe. So, the earliest stress path let us say this is the starting
point. And we know that now, this is 3 that we have already derived, then this is the strength
corresponding to compression which is given us tfc and for extension it is tfe. We can note here
that this tfc ≠ tfe when we follow the elastic stress path.
So, tan 𝛼 ′ let us try to understand what is the ratio tan 𝛼 ′ , it can you can write as tfc. This
particular triangle tfc divided by so’ + tfc/ 3. tan 𝛼 ′ can also be written as
𝑡𝑓𝑐
tan 𝛼 ′ = 𝑡𝑓𝑐
𝑠𝑜′ + 3
So, tfc can be written as

𝑡𝑓𝑐
𝑡𝑓𝑐 = 𝑠𝑜′ 𝑡𝑎𝑛𝛼 ′ + 𝑡𝑎𝑛𝛼 ′
3
tfe can be written in this manner.

𝑡𝑓𝑒
𝑡𝑓𝑒 = 𝑠𝑜′ 𝑡𝑎𝑛𝛼 ′ − 𝑡𝑎𝑛𝛼 ′
3
Substituting for we will get
𝑡𝑓𝑐 𝑡𝑓𝑒
𝑡𝑓𝑐 − 𝑡𝑎𝑛𝛼 ′ = 𝑡𝑓𝑒 + 𝑡𝑎𝑛𝛼 ′
3 3
So, rearranging, we can get
𝑡𝑓𝑐 3 + 𝑡𝑎𝑛𝛼 ′
=
𝑡𝑓𝑒 3 − 𝑡𝑎𝑛𝛼 ′
400
which is ±1. So, when we follow an un-drained triaxial compression following the elastic stress
path, the strength in compression and the strength in extension follows this particular ratio.
So, we have discussed quite a number of cases related to stress path. We will quickly see an
example from what we have already learned. So, the given example is the saturated clay sample
is consolidated in triaxial cell by maintaining 0 lateral strain. So, it is as good as simulating a
one dimensional compression. The axial stress increment is by 80 kPa up to a maximum of 480
kPa.
So, maximum stress is 480 kPa with an increment of 80. Soil is then allowed to swell under 0
radial strain. Again maintaining one dimensional consolidation with a final decrement to 40
kPa. So, this starting from 40 to 80 it comes then further on it is an increment of 80. Once after
reaching 480, it is unloaded in steps till it reaches the final value of 40 kPa. So, that is what it
means to a final decrement to 40 kPa.
𝜙 ′ = 24 degrees angle of internal friction of the sample is given. Assume m = 0.41. Where is
this m coming from? This is for determining the coefficient of lateral earth pressure at rest
corresponding to OC state. For kOC, that is why it is written m = 0.41 for OC. The Poisson’s
ratio effective Poisson’s ratio is given as 0.25. Now, we need to determine the following. What
is that? What is the final OCR?
Plot this stress path for the given condition on t, s, s’ plot. If the soil is reconsolidated under
increasing axial stress, I mean to say after reaching 480 you are allowing it to unload. It reaches
401
up to 40 kPa axial stress. Now, if you again reload it, if the soil is reconsolidated, reconsolidate
means reloaded under increasing axial stress, what would be the axial stress when stress ratio
of σr’ by σa’ again equal to koNC.
And you need to assume it to be elastic behaviour. Remember, we have loaded unloaded. And
while reloading we have seen that this particular reloading curve quickly approaches the kNC
line. So, just to prove that particular point just to see that how it happens, we have given an
example. And in that reloading curve, we need to assume that it behaves in an elastic manner.
It is just giving you a hint to use that appropriate equation.
So, let us see how we will solve this. So, the normal consolidated loading is given σa’ is
increased. And we are adjusting the σr’ such that it maintains a normally consolidated
condition. No lateral strain is permitted. This is a one dimensional consolidation case. Now,
you can see that the starting is not at 80. Why? Because the final decrement we have to reach
to 40.
So, from 40 to 80, further, it is an increment of 80 up to 480. So, we are again seeing that in t,
s, s’ plot. Now, we know that 𝑘𝑂𝑁𝐶 = 1 − 𝑠𝑖𝑛𝜙 ′ . 𝜙′ is given 24 degrees. Substituting it, we
will get ko NC is = 0.6. Now, all this increment corresponds to ko NC line and hence σr’. What
will be σr’? Because, we do not know the value of t and s, s’.
So, that is possible only when we have σa’ and σr’. So, σr’ is ko NC of σa’ because that is the
coefficient of lateral earth pressure at rest condition. So, knowing this kONC is known, σa’ is
known. So, here it is not there because that is not where the loading starts. It starts from 80. So,
σr’= ko.
So, we can determine the value of σr’ which confirms to one dimensional consolidation. And
these are the values which we can determine using this expression. Once we know these 2
values, we can determine t and s’ values. And these values can be plotted on the stress path.
𝛥𝑡 1–𝑘
And Δt/Δs’ the slope is known which is 𝛥𝑠 ’ = 1 + 𝑘𝑂𝑁𝐶 . And that is = 0.25.
𝑂𝑁𝐶
So, we need to draw a line in this plot t, s, s’ with a slope which is given by 0.25. That is what
we have just drawn. Slope is = 0.25. And these points which are marked in red, these are the
402
points which is 16, 64, 30, 128, up to 96, 384. So, this is one part of the exercise where we have
plotted the given NC stress path. Now, what is the next condition? Now, the next condition is
after reaching 480, σa’ now this is released one by one.
And it will follow an OC state. So, there will be unloading which is happening. Now, how to
determine that?
So, this is the initial condition what we have already determined and with this we have plotted.
Now, unloading and OC state. Now, 𝑘𝑜 𝑂𝐶 = 𝑘𝑂𝑁𝐶 (𝑂𝐶𝑅)𝑚 Now, this m is given 0.41 kONC is
known. So, knowing OCR we can determine ko OC. Now OCR, how to determine OCR? Here
we know that the maximum pressure the soil is subjected to is 480.
So, 480 can be considered as the preconsolidation pressure. So, here 480 is yield stress or
preconsolidation stress. Now, we know how to determine OCR. The preconsolidation pressure
divided by the pressure it is subjected to. Now, if during unloading next is 400. So, 480 up by
480 that is on the normally consolidated line. Now, when it is unloaded to 400 it is 480 divided
by 400 is the OCR. That is 1.2.
Similarly, 320 by 480 that is 1.5 and if last value is 480 divided by 40 which is 12. So, one of
the question is, what is the maximum OCR? So, the final OCR is = 12. So, this point of 40
corresponds to OCR of 12. Now, once OCR is known, we can substitute here and get the values
of ko OC and that keeps changing depending upon the level of OCR. Now, the same thing can
be written in terms of elastic conditions as well but here we are using ko OC. And here ko OC
403
is = substituting each of the OCR values in this equation we will get 0.6, 0.65. Now, it does
from down upwards. Because it is unloading, 0.6, 0.65, 71 up to 1.66 is the ko OC. Now, once
ko OC is obtained we can determine what is σr’ because σa’ is known. This is σa’.
Hence, we can determine what is σr’ which is 66 100 up to so, it is from 288 upwards up to 66.
Now, once σa’ σr’ is known, we can find out what is t and s’. So, the final value is from 96 384
to - 13 53. So, how does it look like? Plot this stress path. It is keep on changing. You can see
this because ko OC keeps on changing. So, this will give up to the last point is - 13 53. So, this
is the unloading path.
Now, we have one more aspect to be considered, which is the reloading and at what stress
condition of t and s’ it will reach the normally consolidated line. So, that is what we need to
find out next. So, reconsolidated under increasing axial stress and we need to assume elastic
behaviour. So, if that is the case then this equation is already available to us,
𝛥𝜎𝑟′ µ′ 0.25
= = = 0.33
𝛥𝜎𝑎′ 1 − µ′ 1 − 0.25
Now, this is the slope with which it is go back to the ko NC line. So, from here reloading means
it goes at some inclination and meets the ko NC line. That is what it happens. Now, the slope
is known 0.33. Now, we need to find out at what stress condition it joins back the ko NC line.
So, given condition is
𝛥𝜎𝑟 ’ + 𝛥𝜎𝑟 ’
= 𝑘𝑂𝑁𝐶 = 0.6 .
𝛥𝜎𝑎 ’ + 𝛥𝜎𝑎 ’
404
Now, from here if Δσr’ and Δσa’ is given it should reach the ko NC line. So, that is what it
means. So, this condition is already given. At what stress increment now, when you reload it
what stress increment of Δσa’ which is a which is again determine Δσr’ under what increment
this particular point will reach the kONC line. So, that is why it is equated to kONC. And its value
is known which is 0.6.
Now, σr’ the final point σr’ is 66. So,

66 + 0.33𝛥𝜎𝑎′
= 0.6
40 + 𝛥𝜎𝑎′
Now, solving this, we will get 𝛥𝜎𝑎′ = 156 𝑘𝑃𝑎.And hence 𝜎𝑎′ + 𝛥𝜎𝑎′ = 196 𝑘𝑃𝑎
Similarly, you can find out

𝜎𝑟′ + 𝛥𝜎𝑟′ = 117 𝑘𝑃𝑎
So, knowing this stress we can determine t and s, s’
𝑡 = 40 𝑘𝑃𝑎
𝑠 ′ = 157 𝑘𝑃𝑎
So, this is the point where the reloading line or the stress path is going to meet. So, that is given
by this particular point. So, this particular Blue cross it represents t = 40 kPa and s’ is = 157
kPa. So, the reloading happens to this particular point.
So, it is a clear demonstration of the loading kNC loading unloading OC and again reloading
which we have discussed in the previous lecture. So, this is all about some examples, let us
summarize the last 2 lectures whatever we have learned till now.
405
This stress path SP for isotropic and one dimensional consolidation has been discussed. Stress
path for one dimensional consolidation we have seen that it is governed by coefficient of lateral
earth pressure at rest ko corresponding to NC which is kONC. Stress path for unloading is
governed by ko corresponding to OC. That is ko OC. Stress path for unloading and reloading also
confirms to elastic behaviour µ’/(1 - µ’).
So, therefore, this = 1 − µ’ particular expression is also valid for unloading reloading OC.
𝛥𝜎𝑎′
So, you can always approximate by unloading reloading to be of the same path. That is also
possible. If it is not then the unloading will confirm to ko OC. And again reloading will confirm
µ’/(1 - µ’) as the example we have seen. Reloading the heavily over consolidated soil, the ESP
approaches kNC line quickly by following elastic behaviour.
We have seen this already and we have also demonstrated this in the last example. Different
examples of SP variations for isotropic and anisotropic initial consolidation has been discussed.
Stress path for gradual built up of overburden pressure due to sedimentation and its removal
has been presented. And finally, the elastic stress path for un-drained triaxial compression is
presented.
Now, these are some of the cases which will help us to draw this stress path better. Now, in the
next lecture, we will see specifically stress path for triaxial testing in the lab. So, that is all for
now, thank you.
406
Lecture - 37
Stress Path for Drained Triaxial Testing
Welcome back all of you. So, we are dealing with stress paths in these lectures. So, in the last
lecture, we have discussed about some common cases and to some extent now, you are
conversant with the plotting of stress path. Now, let us be a bit focussed on the use of stress
path. First let us see the laboratory stress paths mostly for triaxial testing. So, today’s lecture
will be about stress path for triaxial testing and very specifically for drained triaxial test.
So, today’s lecture is exclusively for drained triaxial test. So, let us start with the statement
that all of these most of us know, but for completeness we need to understand this.
Intermediate principal stress interaction test is either equal to σ1 or σ3. The effect of σ2 is not
taken into account explicitly. A wide range of stress paths are possible by independently
varying σa and σr.
This we know that in triaxial testing, we have better flexibility in modulating these stress
paths or for very specifically, when we have stress path triaxial system, it is very much
possible to vary the stresses to suit the requirements of whatever is the scenario existing in
the field. A stress path gives a better representation of un-drained test specifically, when we
have u measurement as compared to Mohr circle representation.
407
Otherwise, also if you want to understand clearly how the failure is taking place, at what
instant the failure would happen, and why it is happening at that particular instant, these
informations are not available, when we plot the stresses in terms of Mohr circle. So, stress
path is much more handy in understanding this and to some extent you have already
appreciated this point in the last few lectures.
And it is very specifically for un-drained test that also we need to keep in mind because it
clearly shows how the un-drained stress path it routes to its failure. Representation of failure
progression by stress points are convenient, specifically when we do it on q-p’ and t-s’ plot.
Hence, stress path method can be a convenient replacement over Mohr circle for triaxial test
representation.
But, having said that, we need to keep in mind that if our primary objective is to obtain shear
strength parameters, even plotting Mohr circle or t-s, s’plot also would do. If you want to
understand further if not failure mechanism, the progression of failure then stress part would
be much more handy. Another aspect before we start seriously into the stress path plotting we
need to keep in mind that the stress path has to be indicated by arrows because we have to
specify the direction in which the progression is happening. So, arrows are a must for stress
path.
So, let us start with different cases of drained stress path for triaxial test. So, the first case is
drained compression with constant radial stress and it is designated as D1. So, it is the
408
conditions are given already. It is constant radial stress means σr is not changing. So,
compression is brought out by σa. Now this is a very typical common triaxial compression
test.
In drained test, total and effective stresses are same and hence TSP is total stress path is same
as effective stress path. u is assumed to be zero. Now, there can be possibility of initial border
pressure within the sample, because of back pressure. Even if it is there, then it acts like a
static pore pressure as we have seen in the previous lectures, then the total stress path and
effective stress path will not be same.
And this we will be discussing towards the last part of today’s lecture. Initial stress conditions
are considered to be isotropic. So, that is the starting point which is designated as M. So, we
keep everything same as that of the triaxial testing conditions, simple triaxial testing. Now, if
we want to create anisotropic consolidation that is also possible, but we are going ahead with
simple initial isotropic stress condition.
So, now, we are plotting in terms of all the three stress paths, but later, you will see that we
will adopt any one of these three stress paths as and when we depending upon the
convenience So, it can be t-s, s’ or q-p, p’ plots. So, now, let us first start with σa σ’a, σr σ’r
plot. It is given the isotropic is the starting point and we should not have any difficulty in
identifying this, because it will be on the 45-degree line.
So, this is the starting point M and similarly, for t-s, s’ it is M on the isotropic line and q-p, p’
M. So, let us start now, how the stress path will move. It is constant radial stress. So,
compression is caused by increasing sigma a dash. So, it is clear now, these things have to be
kept in mind. So, it is increase in σ’a and hence Δσ’a = + Δσ’a. This aspect again we have
discussed in the previous lecture.
So, depending upon whether it is increase or decrease whether it is compressing or loading or

unloading accordingly the sign of the increase would change or increase or decrease would
change. So, here it is brought about by increase in σ’a. So, Δσ’a in constant radial stresses
means it is not changing. Δσ’r is equal to 0. Now, for σa σ’a, σr σ’r plot there is nothing great
because it is a direct replication of what this stress path would be and it is very easy to plot.
409
So, this will be the designated stress path D1 for in this particular plot. Now, how it will look
like in t-s, s’ plot? So, you just need to substitute this and that is where we have to be careful
about. While substituting do not forget to put the sign. So, here that is Δσ’a - 0 / 2 that is Δσ’a
/ 2. Here plus again Δσ’a / 2 so, it will be the slope will be Δt / Δs dash equal to 1.
So, D1 is one is to one. Now, this again you have to keep in mind like in t-s, s’plot, if it is
triaxial compression, the stress path and for specifically for drained stress path it will be at an
inclination of 1 is to 1 or 45 degrees. That aspect we need to keep in mind. Now, what will
happen in q-p, p’ plot? Δp’ is equal to Δσ’a /3, q is Δσ’a, So, Δq / Δp’ is equal to 3. So, this is
the next understanding in q-p, p’ plot, if it is drained triaxial compression then the slope will
be 3.
So, this is D1 and this represented by 3 so, all the three has been done. So, this is the first
case.
Now, the second case is drained compression with constant axial stress designated by D2.
Now, axial stress there is a triaxial sample, there is an axial stress, there is a radial stress how
do we account for or how do we bring about compression in the sample? It is told that σ’a
remains constant. Now, what is the possibility of compression? A released σ’r.
So, when you release it, there is a tendency to undergo compression under constant σ’a. So,
compression caused by release of σ’r. Now, when there is a release and here it is constant
axial stress so, it is zero and just because it is released it will be minus of Δσ’r. So, in this
410
plot, there is a starting point a’ remains constant and Δσ’r = - Δσ’r and so, it has to be in this
manner.
So, this is constant σ’a and this is decreasing and finally, it reaches the failure line. And that is
given as D2. What about t-s, s’ plot? Δt is equal to remember Δσ’a is zero, minus of minus so,
that will become plus , so plus Δσ’r / 2. Here it will be zero minus Δσ’r / 2 and hence the
slope will be Δt / Δs equal to minus 1. So, this will be minus 1 slope.
And remember here again your Δt is positive Δs is negative So, you should be able to
understand in which direction it will move, minus 1 can be in this direction as well. But why
it is specifically in this direction because s is going towards negative, t is going towards
positive as you can see here. So, that is why it is negative slope in this upward direction and it
meets the compression failure line.
What about q-p, p’ plot? So, Δp’ = -2Δσ’r / 3 and Δq = Δσ’r because minus of minus that will
be plus Δσ’r. Δq / Δp’= -3/2. So, it will be in this direction at a slope of 3 upon 2. So, this is
the second case.
Now, the third case is drained extension with constant radial stress. Now, if it has to create a
sort of extension based on the compressive stresses, so, how it will happen? It is told that
radial stress is constant. Now, if extension has to happen, then we need to release the axial
stress. That is why it is like kind of squeezing effect. You are squeezing something so, it
extend. So, that is the same philosophy that is adopted. So, here it is constant.
411
So, the only possible way of extending is release σ’a. So, extension caused by release of σ’a,
that is because of that, it will be minus Δσ’a , Δσ’r = 0. So, a is moving in the negative
direction. So, a is changing a is changing in the negative direction. So, that will be D3 and
that will be downwards. How it will be in t-s, s’ plot?
So, Δt = -Δσ’a /2, Δs’ = -Δσ’a /2. So, both Δt and Δs is changing towards negative direction,
but the slope is 1. Both are moving towards but then slope is 1. So this slope, what is the sign
of this slope and what is the sign of individual stress parameter also need to be considered.
So, here it will be in D3 direction. So, this is positive 1 slope but both are moving towards
negative.
So, this is negative s direction and negative t direction. And the same will be there for q-p, p’
plot. We have Δq upon Δp’ equal to 3, but in the downward direction.
Now, the fourth case is drained extension with constant axial stress. Now, axial stress is
remaining constant. Now, how do we bring about extension? If you want to bring about
extension you need to squeeze it. So, σ’a constant so, σr is increasing. So, extension caused by
increase in σ’r. So, Δσ’a = 0; Δσ’r = + Δσ’r.
So, this is in the positive direction that is represented by D4. What about t-s, s’, Δt = -Δσ’r /2.
Now, you can see that t should move towards negative direction, Δs = Δσ’r /2 and hence Δt/
412
Δs = -1. So, you will have from here it moves towards you can see that s is in the positive
direction, whereas, t is in the negative direction.
So, that is all about t-s, s’, what about q-p, p’? So, Δp’ = 2Δσ’r /3, that is Δq = -Δσ’r. Hence,
Δq/ Δp’ = -3/2, so slope is minus 3/ 2. It will move in the downward direction that is for D4.
So, these are the important possible 4 cases one kept constant at a time, so, both compression
and extension.
So, this is a summary of drained stress path for triaxial tests. Now, another important aspect
which we need to understand is that, depending upon the stress path, you can see that the
failure when it is going to happen, that is also dictated by the manner in which the stress path
changes. Let us say for the case of D1, you can see that so, much of allowance is possible. So,
the strength available is this. So, when it reaches here, it fails.
Whereas, for D3, you can see that the failure occurs quite fast, even though the starting point
is same and that is true for all other stress path representation. So, this stress path
representation also gives such insight as to same initial point, but depending upon the manner
in which the stress changes, the failure condition also changes. To some extent we have seen
this when we discussed about shear strength tests.
So, depending upon the manner in which the loading changes the it is it influences the failure
conditions as well. So, now, you are just summarising this. These are the four different stress
path directions. If I draw a line here, then because this is the starting point above portion, this
413
particular portion it represents compression, because k it is dictated by what k you are talking
about. Here it is σ’a is more than σ’r. So, it is compression.
This region is basically an extension from here towards this direction we talk about unloading
and towards this direction it is loading. So, all the cases have been summarised in this
particular plot and this x and y can be any of these stress parameters either t-s, s’ or q-p, p’.
So, failure strengths in drain tests are governed by stress paths. For an identical initial
isotropic stress state, we have tfD1 equals tfD4.
Now, we are talking in terms of t here, you can see that tfD1, tfD1 means the strength which is
given by the failure stress path D1. So, this is what is meant by tfD1. So, if I draw here tfD1,
that means, it is the strength of the soil corresponding to the stress path D1. D1 this stress
path strength is same as tfD4 it is more like a mirror image. Similarly, tfD2 is equal to tfD3.
Now, let us see the influence of initial state and cohesion on drained spot for triaxial test.
Now, here till now we have been considering c’ equal to zero. Now, let us see that what will
happen if cohesion is there? Is there significant difference from what we have learned or how
it changes? And the other aspect is the initial state. Initial state means isotropic consolidation
or whether it is a normally consolidated soil or any other k value which we tend to use.
So, let the initial stress state correspond to k0NC. Now first case, let us consider k0NC condition
with k0NC is less than 1. That means σ’a is greater than σ’r a typical triaxial test that we
normally do in the lab. So, we are talking now, in terms of t-s, s’ as I indicated before, we
414
will not do for all the three now, as in when the case comes, we will generally use any one
and typically we will be dealing with t-s, s’, q-p, p’ plot.
So, now, let us try to understand what will be the influence on stress path because of the
initial state and cohesion. Now, initial stress state is given by s’0 and t0. So, the initial point
corresponds to k0NC state. So, if you draw a line here k0NC, this initial stress state falls on that
and it is clear that this k value is less than 1. So, we have already seen in the previous lecture
like if it is k value less than 1 which corresponds to σ’a greater than σ’r, it will be in the upper
part or in this particular region.
So, this is our starting point P. Now, to add value to this stress path plot, let us first draw the
failure line. Now, you can see that it is no longer through the origin. There is an intercept
here and in t s s dash plot what is the intercept? Definitely it is not c’, it is a’. And the
relationship between a’ and c’ is already known to us. So, this is the inclination α’ or tan α’,
which is equal to sine φ’ which we have already seen.
So, this is α’ and this is the point Q. This is a’ which is the intercept in t-s, s’ plot. Now, if
this is a’ we can write this to be a’ cot α’. Now, for drained stress path in compression, you
can see that the one which we have already seen before let this point be S. Now what has
happened? Initially, this stress path started off from here, because it is normally consolidated.
So, the point has shifted upwards, but at the same time the failure line also got shifted
upwards which means that strength because of C’ has been incorporated. This is the other
spot so, we will be generally dealing with D1 and D2 both refers to a case of compression.
Now, this particular point P is at s’0, the initial point and this is t0. We have also seen that if
you know this particular point P one can also find out what are the stresses it is subjected to.
So, this is σ’a, this is σ’r, what we have to do? Draw a horizontal line draw 45 degrees to both
sides, wherever it is intercepting the S line so, that will give us σ’a, and σ’r. This is already
discussed. In that case this will be t0 and this will be t0. Now, what we are trying to do is we
are trying to understand or determine what will be the expression for the strength represented
by D1 and D2.
415
Now, what is the strength? It is tf. That is a failure point in T. Let us assume this point to be S
and T. So, you are dropping a vertical from this particular point. So, let it be ST. Now, we
know that this angle is 45, this angle is 45. So, this angle is 45 degrees. And this is α and
hence this is also α’. This is tfD1, that is the strength corresponding to the stress path D1 is
given by tfD1.
Since this is 45 degrees, take this triangle, this particular triangle this is 45 degrees. So, here
it is tfD1. So, definitely this has to be tfD1. Now, referring to this figure QT is equal to this is
tfD1 divided by QT is equal to tan α’. So, QT will be equal to tfD1 divided by tan α’. That is QT
equal to tfD1 cot α’. So, what we are trying to do we are trying to get an expression for tfD1.
Now, what is RT? RT is equal to tfD1. That is already clearly marked on the figure. We can
also write RT = QT – (Q0 + OR). So, that will give RT. So, we have the expression for RT,
tfD1 substituting that, we also have the expression for tfD1. QT that is tfD1 cot α’. QRO is
already marked here a’ cot α’and OR that is OR = s’0 - t0. So, that will be OR.
So, all these are known. Substituted, we will get tfD1 = tfD1 cot α’ – (a’ cot α’ + (s’0 - t0)). So,
rearranging similar terms then we will get tfD1 cot α’ – 1. What we are taking is you we are
bringing this to this and tfD1 on to this side. That will give tfD1 = (a’ cot α’ + (s’0 - t0))/ (cot α’ –
1). So, this is about tfD1.
416
So, we need to also find expression for tfD2. Let us see how. Let us drop a vertical at U. This
is tfD2. This is at an angle of 45 degrees we can write UV = tfD2 because this is 45 degrees. QU
= tfD2 cot α’. This QU is equal to this one. So, that is very clear. We can also write QU = Qo+
OT + TV -UV. That will give you QU. So, all these are known to us.
QU is tfD2 cot α’, Qo a’ cot α’ + OT. This is s’0 - t0 UV that is tfD2. Again rearranging we will
get tfD2 (cot α’ + 1) = a’ cot α’ + (s’0 - t0). So, we will get the expression for tfD2 similar to tfD1
= a’ cot α’ + (s’0 + t0))/ (cot α’ + 1). So, this is about the expression for both tfD1 and tfD2.
Now, let us see what happens if k is greater than 1. So, earlier case was k less than 1. That is
σ’a less than σ’r. So, where will the points start from? It has to be in the bottom portion. So,
the same for the previous case. So, now, you can see that point P is in the bottom half. Every
417
all other aspects remain same. Now, D1, it starts from here. Then this is tfD1 same and this is
D2. And this is the strength tfD2.
You can see that there are more allowance now, because the it is starting from this particular
point it takes a while reaching the compression failure line. So, what is the effect? It has got
shifted. Now, here this will be t0, t0 because why it is t0 , t0 here? P
Because this is at an inclination of 1 is to 1, or 45 degrees. Previous equations for tfD1 and tfD2
are valid for this case as well.
But the only difference is the starting point is below. So, the value of t0 will be negative.
Now, let us see that what will happen if there is some initial pore water pressure? We know
that this is a drained test we are discussing about and we do not expect any pore water
pressure development during shearing. But what if it happens that there is some pore water
pressure in the beginning? And that we have seen there are different possibilities why there
should be pore water pressure.
A very good example is the presence of backpressure because of saturation. We know that it
is a static pore water pressure, but then still that acts as pore water pressure and hence, the
stress path will be different even though we know that total stress path and effective stress
path are same in drain triaxial testing. You should not confuse why there is a contradiction
here, there is no contradiction. The pore water pressure referring to is only the initial pore
water pressure.
418
Drain compression for a normally consolidated sample with zero increase in pore water
pressure. Here you can see that you are discussing about initial not during sharing so that is
why It is zero increase in u means u is pore water pressure. But there is an initial back
pressure which exists in the sample during saturation. So, again t-s, s’ plot kfc has been
drawn, the initial isotropic consolidation M is stated here. So, there is a back pressure.
So, now, the initial point, total stress point and the effective stress points are different because
of this difference in back pressure. So, now, here it is M, here it is M’. So, here the effective
stress path is in this direction. Now, what is the slope of effective stress path here, we have
already understood that in a drained triaxial compression in t-s, s’ plot it is 1 is to 1. So, 45
degrees, so, this is 45 degrees or 1 is to 1.
Now, total stress path will be parallel to effective stress path. Now, there is something
interesting, which is coming out here, why did I stop this total stress path here? You have the
failure line it should have extended. Now, this could be an immediate response for some of
you. Please note that failure is always brought about by effective stress. So, the failure is
dictated by effective stress path.
So, this is the effective stress path. It fails at this point. So, total stress path beyond this point
is not meaningful, because the soil does not exist that you have to keep in mind. This is 1 is to
1, I have already told. In the same q same condition for q-p, p’ is the same thing. Only the
slope changes all other details remains the same. There is a parallel shift in effective stress
path with a difference of back pressure.
Shearing under different isotropic cell pressure results in different parallel ESPs. Again, it is
better that we understand this point right now, because what it means is that now, this is this
corresponds to one particular effective cell pressure. Now, when we share at different cell
pressure in triaxial drained condition, we will have parallel ESPs. This is what it is, t-s, s’ and
this is the failure line you can see that this corresponds to the first cell pressure.
This corresponds to the second cell pressure and this corresponds to the third cell pressure.
So, these are different cell pressures, we can see that how the failure is happening. So, it is
just a parallel shift in effective stress path.
419
Now, let us deal with drained compression of over consolidated soil with zero increase in
pore water pressure. Earlier it is normally consolidated. Now, it is over consolidated. Initial
back pressure exist. Now, remember, this failure line corresponds to critical state friction
angle. All those things we have already discussed. Now with problem with over consolidated
sample is that it exhibits peak and then fails at critical state or ultimate state.
So, that is what why we have to specifically discuss it for over consolidated soil. So back
pressure exists all other condition remains same. Now, you may ask here, we have already
learned that in over consolidated state pore water pressure will be negative. You are correct.
But that corresponds to shearing state. Here we are referring to initial state. You have not
started shearing. Negative pore water pressure happens because of dilation.
Dilation happens during shearing, so, do not have to confuse that. So, here other things
remaining same now, here the effective stress path is crossing the failure line. So, what is
happening? How can this be possible? There is no state of this soil that is possible beyond the
failure state. Now, here we have to be very clear about our understanding, in fact, the same
aspect is going to be very clear, when we discuss about yielding.
Here, what has happened is the soil would fail only after exhibiting peak in the case of over
consolidated. That means, the yielding has to happen first then only it will result in failure.
But, unfortunately this line corresponds to the critical state lines. So, soil has not yielded, so,
420
first it will go to the yield point and then it will come back to failure. So, here it has reached
the peak, because it is always greater than the critical state.
So, it reaches the peak and then it comes down. You can see this arrow coming down and
then it will fail at this point of critical state. So, it goes up and then come down. And that is
just because, when you understand this, viza viz..., the stress strain response of over
consolidated soil which we have already discussed, this will this aspect will become very
clear.
And your total stress path will be up to this particular point. Now, if you want to find out
what is the pore water pressure corresponding to peak, then the total stress path will go here.
So, this is the difference. In fact, total stress path does not come into picture when we talk
about failure. So, this again the difference is backpressure. Here it is q-p, p’ so, the slope is 3.
So, during during sharing of OC, it would reach peak state first the stress then falls to critical
state by strain softening.
So, that is all about triaxial testing drained case specificall. We have not talked about un-
drained now. So, what we have learned in this lecture, let us summarise. Triaxial drained SPs
discussed for well defined cases of compression, extension, loading and release combinations
is discussed in three stress path plots. Drained compression loading, that means, we are
talking about increasing of the load is at 1 is to 1 slope in t-s, s’ and 3 is to 1 slope in q-p, p’.
421
This aspect you need to keep in mind. That is for triaxial case, compression and it is loading
that is increase in stresses. So, it is 1, positive 1 and positive 3. Initial stress state that is
which is defined by the k value and cohesion influences this response. We have clearly seen
that. So, here we have already derived drained strength tfD1 and tfD2.
Presence of initial pore water pressure, how this could impact the stress part we have
discussed, even though in drained triaxial condition. We do not expect ESP and TSP but just
because of the presence of initial back pressure both gets shifted parallel to each other, but
with a difference of u maybe that is back pressure. Shearing under different isotropic cell
pressure results in different parallel ESPs.
For drained sharing of over consolidated sample, the stress state reaches peak and then
returns to critical state failure line. A failure condition is totally defined by ESP and not by
TSP. This is a very important point that we have to keep in mind. For drained sharing, ESP
and TSP are same in the absence of initial pore water pressure. So some of these concepts are
very relevant. And you may be able to understand it very well. But then these points we need
to keep in mind so that we understand it better.
So that is all for today’s lecture. We will see few more cases that is exclusively triaxial un-
drained test and its stress path in the next lecture. Thank you
422
Lecture - 38
Stress Path for Undrained Triaxial Testing
Welcome back to all of you. So, we were discussing about stress path and in the last lecture
we have seen stress path for drained triaxial testing. So, in today’s lecture, we will see the
stress path corresponding to un-drained condition or un-drained test. So, what is the essential
difference? We know that from the shear strength discussion the essential difference would
be in terms of pore water pressure generation. So, we will see how that is going to impact the
stress path.
So, just like what we have done for drained compression test, there are four cases that we
discussed. Here the first case is un-drained compression test with constant radial total stress.
Now, when we discussed about drained stress path, we did not have to stress too much on
effective or total, but now, it is important. So, you can see here that it is with constant radial
total stress. And that is designated by U1, earlier it was D1.
But the conditions were same. The first condition we dealt was constant radial stress, but here
we have specified total stress. So, it is very clear that the TSP and ESP will not be seen for
the case of untrained test. Just like in normal triaxial testing, we always have the initial
423
condition to be isotropic. So, let us start with isotropic condition M. And again we will
discuss what three stress paths.
One force in actual radial, the other one is what t-s, s’ and other one q-p, p’. So, this is the
starting point the isotropic condition, so, M is the starting point. Like that, we have t-s, s’ and
q-p, p’. So, this is the starting point. Now, for the first condition, we know the radial stress is
constant, so, Δσr = 0. And for creating compression, we know that the actual stress should
increase.
So, Δσa = + Δσa . Now, you are more or less conversant with these usage. So, when there is
release it is negative and when there is loading or increase in load it was positive. So, this is
the starting point. Now, for this particular country condition, we know that TSP remains
constant with radial and it increases. That is actual load increases. So, this is increasing in this
direction.
Now, this is very much similar to the drained triaxial stress path for the same condition. So,
the total stress path in both the cases for the drained and the untrained is not going to change.
But, you may be wondering, why did I stop this total stress path at this particular point? Why
it is so? And there is a discontinuity here. In fact, when it comes to effect and during testing,
you will see that total stress path will not represent what it actually has to represent about
failure.
So, it is all governed by effective stress path. So, whether it is complete or not whether the
total stress path is meeting or whether it is crossing the failure line, it hardly matters. Now, let
us say that for this particular condition, initial pore water pressure u0 exists. What is the
implication of this statement? It means that the total stress path and the effective stress path
will not start at the same point. So, there will be a difference of u0.
So, let us find. So, there is a difference of u0 and that remains isotropic. So, the point M will
get shifted to M’ for effective stress paths. So, that is why It is called M’, for effective stress
path. Now is it straightforward to plot effective stress path with the given information? But
ESP will be separated from TSP by pore water pressure at any time. So, it is an un-drained
test, for every instant the pore water pressure changes during shearing.
424
So, we need to understand what we need to get what is the value of pore water pressure with
loading. How do we do that? Either we measure it, get the results and you plot the effective
stress parameters the variation or u is estimated by Skempton’s pore pressure equation. Now,
you will again appreciate why we have studied Skempton’s pore water pressure equation in
detail.
So, in the absence of measurement, one can also have some idea about pore water pressure by
estimating it corresponding to the change in stresses. So, Δu = B [Δσ3 + A(Δσ1 - Δσ3)]. So,
here replacing σ3, σ1 by σr and σa, we have Δu = B [Δ σr + A(Δ σa - Δ σr)]. It is a saturated
case. So, B is equal to 1.
Δu = Δ σr + A(Δ σa - Δ σr) . Now, this is a typical case of Δ σr = 0 and Δ σa = + Δ σa.

Substitute this into pore water pressure equation, we will get Δu = AΔ σa +Δ σr (1-A). That is
by rearranging and for this specific case Δ σr = 0.
So, now, we have Δu = AΔ σa. So, as σa changes, the pore water pressure will change. Once
we know the pore water pressure parameter A, then we can always determine what is the
change in pore water pressure.
So, now, it is easy to plot ESP, but is it that easy? One, we have got the pore water pressure
estimated. So, now, you can find out what is the effective stress. Δσ’a = Δ σa - Δu. We have
425
the expression for Δu. Substitute that we will get Δσ’a = Δ σa(1-A). So, Δσ’r = Δ σr - Δu. Δσ’r
= 0 – A Δ σa .
So, Δσ’r = – A Δ σa. So, the slope Δσ’a / Δσ’r = -(1-A) / A. So, you divide this by this, -(1-A)
/ A. So, now, we have we are left with an equation for the slope earlier it was a number so, it
is not changing here it changes the effective stress path keeps changing with shearing. And
that change will be with respect to the parameter A.
So, ESP varies with A. So, and A is dependent on stress history. This much we have already
seen in our previous lectures. Now, consider a case of normally consolidated because once
we have stress history coming into picture, let us first see what happens for NC. Now, let A at
failure Af is equal to 0.5 as approximate value for NC, this we know. If A is considered
constant, then ESP will be a straight line.
So, we have an expression in terms of A. Now let us say that we assume that A is constant
and we take the final A value. This also we have seen. A value at failure that is Af. Now let us
say Af is equal to 0.5 for NC. Then we approximate it by a straight line. But if we presume
that let us first discuss what happens when A is equal to 0.5. So it will be always negative
because it is - A upon, in the previous slide we have seen the slope which is -(1-A) / A.
So if you substitute 0.5, this will be positive quantity with a negative sign so the slope will be
negative. So this is the negative slope, which you can see and M’ N’ represents the effective
stress path. For a value of let us say Af equal to 0.5. And in reality, that is for a case this
straight line is a case when Af is equal to 0.5 that is considered to be constant, but A changes
with shearing.
And hence, the effective stress path will be nonlinear. This every time it keeps changing. So,
if you consider the variation of A with respect to shearing, then it will be more like a
nonlinear change or nonlinear effective stress path. But most of the practical cases that we
discussed, we consider A value to be a constant and we take the value of Af and hence, we
can consider the effective stress path for un-drained condition also to be a straight line.
426
So, the same thing that we try to do in t-s, s’ plot. Starting point is there. Δ σR =0, Δ σa= + Δ
σa. Everything remains same, u naught exist. So then it will be shifted. Now, this also we
have already seen in our initial stress path discussions like s’o = so - uo. So, in case if you
have any doubt please refer back to the earlier lectures.
So, this is the point M’. So, it is separated by initial pore water pressure uo. So, a total stress
path we need to compute. It is fairly straightforward and it will be same as that of the drained
stress path. So, Δt/ Δs = 1. This we have already seen. So, it is a it is a straight line at an
inclination of 1 is to 1. And this is U1. So, this is the total stress path.
Now, for effective stress path, we need to compute the pore water pressure expression and
that we have already seen, We have already got the pore water pressure expression. So, Δt’
now, we already know that Δt’ or Δt = Δt’. It is not getting influenced, but then still in this
particular case, I am just representing it as dash only to make you understand.
So, Δt’ = Δ σ’a - Δ σ’r / 2. Substituting so, (Δ σa - Δ u) - (Δ σr - Δ u)/ 2. So, now, we have the
expression for Δ u. So, substituting for Δ σr = 0, because that is the condition we have it is Δ
σa /2. Because - Δ u minus of minus is + Δ u, so, that gets cancelled. So, it will be σa /2.
427
Δs’ = Δ σ’a + Δ σ’r / 2. Substituting so, in terms of total stress and then you will get Δs’ = Δ
σa /2 - Δ u. Why? Because here it is minus of minus. So, it is minus 2 Δ u by 2. That is minus
Δ u. So, Δs’ = Δ σa /2 – A Δ σa. Because Δ u = A Δ σa which we have already found out.
Now, Δs’ = Δ σa /2 *(1-2A). So, rearranging this so, Δt’/ Δs’ = 1/ (1-2A). Everything remains
same as that of the previous derivation the same we need to find out what is the slope. So,
slope is 1/ (1-2A) and it is again a function of A. So, if A is considered constant it will be a
straight line.
Now, you need to see that this particular line can vary with A and it can be of any slope in
this direction. So, N’ M’ N’ gives the effective stress path when A is considered to be
constant and if A changes, then it is a nonlinear stress path. For a specific case in this
particular exercise, if A is equal to 0.5, you will see that what happens it becomes infinite.
Δt’/ Δs’ is equal to infinity. So, it will go in the upward direction.
The slope of the effective stress path will be in the upward direction. In fact, these
informations are very important, when we try to understand the critical state soil mechanics.
For NC a typical A value will be 0.5, if you are representing the stress part in t-s, s’ plot, we
know that it is moving in the upward direction.
428
Similarly, for compression in q-p, p’ plot it is one at the same, only thing is the expression
changes. So, we are just trying to find out the expression all other things remain same. So, p’0
= p0 - u0. So, this is the starting point, total stress path remains same. So, we can say that it is
at an inclination of 3 because it is a compression stress path. Δq = Δ σa and Δ p = Δ σa /3.
So, the slope will be 3 and this remains same as that of drained stress path. Now, for effective
stress path Δq’ = Δ σ’a - Δ σ’r. Substituting for total stress in terms of total stress, so, we have
Δq’ = Δ σa - Δ u. Minus of this becomes - Δ u plus Δ u zero. So, that will be equal to Δ σa.
So, Δ p’ is equal to, again substitute in terms of pore water pressure. Δ p’ is equal to Δ σa - Δ
u + 2 (- Δ u). So, that will become - 3Δ u / 3, that goes. So, Δ σa / 3 - Δ u. So, Δ p’ is equal to
Δ σa / 3 - substitute for Δ u, A Δ σa. So, Δ p’ = Δ σa / 3* (1-3A).
429
So, Δq’/Δ p’ = 3/ (1-3A). So, again if A is considered constant, it will be in this. This is only
an example direction it can be in any direction depending on A. So, M’ N’ is the effective
stress path and if it is considered to be A is considered to be varying, so, then it becomes a
nonlinear stress path. Now, for example, if Af is equal to 0.5, then in this case you will have
an inclination of - 6. So, it will be even more steep.
So, you will have the effective stress path. Now, what we have done for un-drained
compression, we have determined the effective stress path for all the three cases. Now, the
compression it will just like we discussed for drained stress path, the compression can also be
brought about in another sequence. So, un-drained compression test with constant actual total
stress, so, that is represented by U2. Initial stress conditions are considered to be isotropic.
So, it is all remain same, release of radial stress causes compression. So, this we know that,
when actual stress is kept constant, we need to release the radial stress so, that there is a
compression. So, that is what we are discussing now. So, rate release of radial stress causes
compression. Again the starting point remains same.
430
So, Δ σa = 0 and Δ σr = - Δ σr. why because it is release. So, we know that in TSP is same as
that of the drained stress path and that is for constant σa, σr is reducing So, it goes and meets
the failure line. Let us presume in this particular case that there is no initial pore water
pressure, so, u0 = 0.
So, what is the implication of this statement? The total stress path and the effective stress
path it starts at the same point. So, M = M’ So, that is what it means. So, ESP and TSP will
start at the same point. So, we have M’. Now, what is the next procedure we know that is un-
drained. So, we need to have the pore water pressure measured or estimated. So, let us
estimate u based on Skempton’s pore water pressure equation.
Is it going to be same or different? It will definitely be different because it will be a

corresponding to radial stress variation. Δu again the same procedure. It is written in terms of
Δ σr, Δ σa, B equal to 1. So, Δu is equal to Δ σr. Rearranging this plus A into Δ σa - Δ σr. So,
Δu = A Δ σa - Δ σr (1-A). So, now, Δu = Δ σr (A- 1).
431
So, earlier the expression was different. Now, we need to find out what is the effective stress
path. So, effective stress Δ σ’a= Δ σa - Δ u. Substitute for Δ u. So, that will be Δ σa is already
zero because that is the condition what we are dealing with. So, - Δ σr (A-1). Δ σ’a = Δ σr (1-
A).
And Δ σ’r= - Δ σr - Δ u. So, Δ σ’r= - Δ σr - - Δ σr (A-1). So, Δ σ’r= Δ σr (-1 –A+1). So, this
minus plus 1 goes away. So, it gives - A Δ σr. So, we will take the slope Δ σ’a / Δ σ’r = -(1-
A)/ A.
Now, have you seen this expression before? Yes, we have seen because the slope of the
effective stress path remains same as in the previous case even though the sequence of
loading has changed. Definitely it has impacted the determination of pore water pressure. The
expressions were different, but the slope is going to be same. So, whatever be the manner in
which the compression has been done or performed the effective stress path slope is same.
So, that is a typical example where in there is a slope and if it is a is varying then it will be
nonlinear. So, M, M’, N’ is the effective stress path. For both cases of compression, the slope
of effective stress path, this expression remains the same.
432
Same thing we need to repeat with t-s, s’. So, it is clear now, for a since the slope for σa –σ’r
plot effective stress path was same in both the cases of compression. We know that it is going
to be same for t-s, s’ and q-p, p’. So, let us see how and whether it is or not. So, TSP is Δt = Δ
σr /2, Δs = - Δ σr /2 because that is the condition.
So, Δt/ Δs = -1. This remains same as that of our drained stress path. So, U2 is in this
direction. So, this is the total stress path. Effective stress path the initial point is same,
because u0 = 0. Δt’ = Δ σ’a - Δ σ’r /2. Substitute in terms of total stress. Yes, Δu the
expression is known. So, we have substituting for Δ σa = 0, we get minus Δu minus minus.
So, that will become plus Δ σr and this will become plus Δu that gets cancelled off. So, Δt’ =
Δ σ’r /2 and Δs’ = Δ σ’a + Δ σ’r /2. Substituting it for in terms of Δu, we have Δs’ = [(0- Δu) +
(Δ σr - Δu )] /2. So, that will give – (Δ σr / 2) - Δu.
433
So, if we substitute for Δu, so, then we will get Δs’ = – (Δ σr / 2) - Δu. So, Δs’ = – (Δ σr / 2) –
(A-1) Δ σr. So, this is the expression for Δu. So, Δs’ = (Δ σr / 2)(-1 – 2A+ 2). So, that is 1-
2A. Δt’/ Δs’ = 1/(1-2A) is the same expression. This is the effective stress path and if it is
varying then nonlinear.
So, similarly, with q-p, p’ we will quickly glance through the all the steps remains same. TSP
is equal to Δq = Δ σr. So, we have Δp = -2 Δ σr / 3. So, that gives the slope of – 3/ 2 is the
total stress path. And effective stress path starting same in the same procedure, we get Δq’ =
Δ σr and Δp’ = Δ σ’a + 2 Δ σ’r / 3.
And [(Δ σa – Δu) + 2(Δ σr – Δu)]/3. So, Δp’ = [(0 – Δu) + 2(-Δ σr – Δu)]/3. And that is equal
to (-2Δ σr /3) – Δu.
434
So, Δu gets substituted. So, Δp’ = (Δ σr /3)(1- 3A). And the slope is 3/(1-3A) and that
remains same. There is the M’ N’ is the ESP and the nonlinear variation.
So, in all the 3 cases, we have got these slope of the effective stress paths same in both the
cases of compression. In both ways of compression it is same. Now, obviously, the curiosity
arises whether this will be true for extension, because we have discussed 4 cases in un-
drained as in the case of drained stress path. So, is it going to be same? So, let us see the case
for just the plot of σa – σ’r, what will be the slope of un-drained effective stress path for both
the cases of extension.
435
So, un-drained extension test with constant radial total stress. So, we know that, if it has to be
extension, then the actual stress has to be released. So, - Δ σa and Δ σr= 0. So, Δ u = B Δ σr +
AB(Δ σa - Δ σr). So, that gives substituting Δ σr= 0. We have Δ u= -A Δ σa. Now, we can find
out Δ σ’a.
Substituting for Δ u, we get Δ σ’a = Δ σa (A-1). And Δ σ’r= 0- Δ u that gives Δ σ’r = A Δ σa.
So, we are substituting it for the pore water pressure and that gives again Δ σ’a / Δ σ’r. If you
rearrange it will give –(1-A)/ A. So, the same slope as in the case of compression.
So, even the extension case the slope is going to be same and it is unique. Now, the fourth
case is un-drained extension test with constant axial total stress. So, we need to create
extension if it is axial total stress. Then Δ σa = 0. Then if extension has to be created then it has
to be increased in the radial stress. So, that is the kind of squeezing effect.
Δ σr substituting now, the same procedure Δ u = Δ σr (1-A). Δ σ’a we can find out that is 0- Δ
σr(1-A). So, Δ σ’a = - Δ σr(1-A), Δ σ’r = Δ σr - Δ σr(1-A). And that will give Δ σ’r = AΔ σr.
So, Δ σ’a / Δ σ’r = - (1-A)/ A. These are identical. So, whatever be the manner in which the
stresses are varying, if it is compression or extension the effective stress path expression is
unique.
So, different TSPs and unique ESP. So, we have different TSP and this TSP variations for
different cases of compression and extension. We have seen that this is very much similar to
436
that of the drained stress path but for all these TSPs we have the unique ESP. Here is just an
example, this particular slope, so it is unique ESP that is what we have understood from our
previous discussion.
So, A value influence on ESP, how A is going to change the ESP. So, let us, this is a t-s, s’
plot, the starting point M. So, this is the TSP for a typical case of compression and extension.
Now, if A is equal to 0.5 in t-s, s’, we know that it is infinity. So, it moves in the vertical
direction. Now, for A equal to minus 0.15. That is this has a minus 0.15 means it refers to a
heavily or consolidated soil.
So, in that case, the effective stress path moves towards in this direction. That is in the right
ward direction and touches the failure envelope. Now, if it is different value other than A,
then we will see that this effective stress path will move leftwards. So, if it is a kind of, so,
these are the different cases of what how the A value would influence the effective stress
path. So, more or less, we have discussed all the different aspects of stress paths for un-
drained testing.
So, let us summarise today’s lecture. So triaxial un-drained stress paths are discussed. ESP
and TSP is separated by pore water pressure developed during shearing. ESP in un-drained
triaxial test is unique for identical soil specimen and initial state. ESP is same irrespective of
the manner in which shearing is imposed whether It is compression or extension mode. In
both all the cases we have seen that the the expression for effective stress path remains the
same.
437
ESP is governed by Skempton’s A parameter and exhibit nonlinear variation during shearing.
And ESP is assumed to be straight line variation from starting to failure by assuming A value
to be constant. In σa, σ’a, σr, σ’r stress path plot the slope of un-drained effective stress path is
–(1-A)/A. Now, in t-s, s’ plot it is 1 /1- 2A and q-p, p’plot it is 3/ 1- 3A.
So, these values, these expressions are unique. So, in un-drained test failure is entirely
defined by ESP when it approaches the failure line. So, we do not have anything to do with
the total stress path and this particular aspect will become very clear as we move further. So,
that is all for this lecture. Thank you.
438
Lecture – 39
Stress Path-Additional Undrained Case
Welcome back all of you in the last lecture, we have seen the stress path for un-drained test.
Now, there are a few more aspects which we need to discuss related to un-drained stress path.
So, we will see those in the today’s lecture.
So, stress path and un-drained strength, so, this is the first one where we will try to obtain the
expression for un-drained shear strength based on the stress path and this we have already
done this for drained stress path in the towards the end of those lectures we have already
derived it for drained stress path. So, similarly for un-drained stress path, let us try to derive
it. Slope of un-drained ESP, let us say, we are taking t-s, s’ plot is 1/ 1 - 2 Af.
Now, these are familiar to all of you now, considering constant A value at failure. So, that is
how we obtained this expression. Let the initial condition t0, s’0. So, let us consider t-s, s’ plot
and this is a general case where even the cohesion is also accounted for this intercept is
nothing but the cohesion in t-s, s’ plot. This will be a’. So, this will be a’cotα’ and α’is the
inclination of the failure line and all these things are familiar to us now.
439
So, the starting point or the initial condition is t0, s’0. This P is the initial condition. So, this is
s’0 and it is at t0. So, this accounts for all the conditions like the influence of cohesion, the
influence of k0 value and we have already seen how k0 influences the initial starting point, it
is not starting from the isotropic line. So, that is what we need to keep in mind. This is a
general case.
Now, for TSP or we do not have much confusion, but presently we are discussing about ESP.
Now, ESP, this slope is governed by this. So, this is a typical value of Af, let us say that it is a
positive slope. Now, what is the un-drained strength? This is the slope 1/ 1 - 2 Af. Let the
point where the effective stress path meets the failure line P. And this is the case when A is
considered to be constant, let us say take the value of Af.
And here this one is when a changes with shearing. Now, depending upon A value, we will
have different slopes of the effective stress path. We have seen one particular case where a is
equal to when Af is substituted by 0.5, we see that it is vertically upwards. Similarly, this
some value of Af , what we need to see is whether this is a positive slope or a negative slope.
So, the moment Af takes the value of 1, for example, we will have 1 – 2, this – 1.
So, there will be a negative slope in this particular direction at. So, here let us say that Af is
such a value that it gives a positive slope whatever it is going to be the same. Whatever we
are going to study or whatever we are going to explain here that remains more or less the
same procedure. So, now this is tfU. Now, what is the implication of tfU. tfU, we know that it is
the radius of Mohr circle and radius of Mohr circle gives the un-drained shear strength sU.
I hope you remember this point. And that is the reason why we have spent a lot of time
describing shear strength characteristics. So, if you do not understand that, it is very difficult
to correlate how tf represents un-drained shear strength. So, it is one of the same tf is σ1 – σ3/
2, which is nothing but the radius of the Mohr circle and radius of the Mohr circle gives the
un-drained shear strength as per trescas criterion.
So, please refer back to the lectures if you have not followed this particular point. So, let us
say QR is what it is represented by that is tfU is given by QR. Now, if you extend this
particular stress path from P downwards, it will meet s, s’ axis at capital S. Now, we know
440
from this particular geometry that QR / SR, it gives the slope 1 / 1 – 2 Af. Why? Because it is
the slope of this. So, 1 / 1 – 2 Af.
So, we can write SR. What is QR? QR is nothing but tfU. So, this QR is substituted here. So,
tfU / SR is equal to 1 / 1 – 2 Af. So, SR = tfU *(1 – 2 Af) So, that is what has been given here.
That is what is marked in this figure SR. Now, let us say if you drop a line from P vertically
downwards, it will meet at point t on the s, s’ this particular axis.
So, there is a dropping of the vertical it meets here. Similarly, the way we have discussed for
QR/ SR, we can also write PT/ ST and that is also equal to 1 / 1 – 2 Af, because it is an
extension. If that is the case, ST can be written as this distance is nothing but t0; t0 *(1 – 2 Af).
So, ST also we got.
So, we have got now SR and ST. Now, we can also write tan α’ = QR/ UR again from the
geometry. Now, if you substitute it, QR is tfU this and so, tan α’= tfU / UR. Now, what is UR?
UR, you can see that this is, this distance UO + OS then + SR. So, if you submit, you will get
this whole distance UR. Now, this is known a’ cot α’, but the next distance is up to s’0 that is
up to the point t.
So, s’0 is known and this ST is also we have got the expression. So, that – ST will give you
the point where this s + SR will give you UR. So, that is what is done here. So, a’ cot α’
comes from here plus s’0, it reaches here minus this distance t0 into (1 – 2 Af)+ SR that is tfU
441
into (1 – 2 Af). So, this distance is also known. Now, if you multiply denominator to tan α’,
we get tfU is equal to a’ into cot α’in tan α’is 1.
So, that will give you a’ + s’0 tan α’- t0 (1 – 2 Af) tan α’+ tfU (1 – 2 Af) tan α’. If you
rearrange again, we can take tfU to one side. So, that will give you tfU into, if you take it
outside, one will come from here minus when you take this on the other side, it will be – (1 –
2 Af) tan α’ that is equal to this particular term which comes here.
So, we will get the expression for tfU which is what we want and that tfU is then equal to a’ +
s’0 tan α’- t0 (1 – 2 Af) tan α’divided by this term come on the denominator that is 1 – (1 – 2
Af) tan α’. So, this is the required expression for tfU. Now, if you substitute for, we already
know that a’ and α’is correlated to Mohr coulomb shear strength parameters.
So, let us see what expression we get. Now, this is the same expression for tfU. We know tan
α’= sin φ’ and a’ = c’ cos φ’. So, this is known to us. So, then tfU is equal to, let us substitute
for a’ which is c’ cos φ’ + s’0 sin φ’. So, tan α’gets replaced by sin φ’. This is what -t0 (1 – 2
Af) instead of tan α’ sin φ’ divided by 1 – (1 – 2 Af) instead of tan α’ sin φ’.
Now, if you divide this the numerator and denominator by sin φ’, because that is a common
term, so, then we will get tfU is equal to c’ cot φ’ that is cos φ’/ sin φ’+ s’0 – t0 (1 – 2 Af). 1 by
sin φ’is cosec φ’ – 1, if you expand the bracket, + 2 Af. So, this is the expression for tfU in
terms of Mohr coulomb failure envelop parameters that is what is the expression for un-
drained shear strength tfU.
442
Now, let us take a case of a sample which is sampled from the field soil which is sampled
from the field, we know that we have already seen that there will be a negative pore water
pressure of – ui. This we have already seen when we discussed about sampling of soil from
the field. Now, this initial pore of pressure, what is the effect of this? This result in initial
effective stress s’i = – ui.
What is the implication of this? And this initial pore water pressure helps the soil to be intact.
So, - s’i that is the initial effective stress is equal to – ui and ui is generally taken as the mean
effective stress when it was sampled. When in the field, what is the mean effective stress that
is equal to ui. So, ui = - (σ’vo + 2 σ’ho)/ 3.
So, there is a mean effective stress. So, ui expression is equal to mean effective stress. For a
saturated sample, which is subjected to cell pressure, there will be a corresponding increase in
use, so, that what it means is that. We mount the sample. So, there is an initial effective
pressure which is attributed to the sampling in the field and when you raise the deviator
pressure under un-drained condition, there will be an equivalent increase in the pore water
pressure.
And hence, it is not going to change the initial effective stress conditions. So, that is what this
particular statement means. So, there is no change in mean effective stress. Why? Because it
is an un-drained condition, but the initial effective stress can be taken equal to s’i = – ui,
443
which is isotropic. Now, this we need to understand a bit carefully, because there is an initial
effective stress of s’i = – ui, this we have seen.
Now, since it is initial pore water pressure, this will act equally in all direction and hence the
condition is considered to be isotropic. So, but in the previous slide, we have seen that it is at
a different starting point. So, t0 is also there; s’0 is also there. Now, it is not on the isotropic
line. But since in this particular case, when we are sampling the soil from the field, we know
that it is at an initial negative pore water pressure and hence the condition has to be enforced
to isotropic condition.
Now, if it is isotropic condition, then it has to start from s, s’ axis, which we know that it
represents the isotropic condition. We have seen this all these concepts by this time should be
very clear to all of you. Now, if it starts from the isotropic stress condition, we have to take t0
= 0 and that is what since it is isotropic t0 = 0 and s’0 = s’i that is the starting point.
Substituting this in the previous expression, that is tfU equal to c’ cot φ’. This expression what
we just discussed, we substitute these values put t0 = 0 and s’0 = s’i. So, then it will give the
expression tfU = (c’ cot φ’ + s’i) / (cosec φ’– 1 + 2 Af). Now, this is a condition which is
specifically useful for soil sample from the field and also for an un-drained condition where
the initial effective stress condition is not altered.
Now, if it is subjected to further consolidation before shearing, the condition would alter
again. So, this you have to understand that it is mostly prominent for UU condition where the
initial effective stress is not altered. It is a purely a case of un-drained shear strength.
444
Now, let us discuss specifically on the stress path for normally consolidated soil under un-
drained conditions. For NC soil, we know that u increases as the deviator stress is applied.
Again, it does not need any sort of explanation for this. This is a typical stress strain variation
for NC soil, you can note that it is strain hardening. Now, if you see that there is a initial pore
water pressure of u0 in the sample during sharing, we know that it is a saturated sample and
hence pore water pressure would increase.
Now, increasing u, increases the difference between ESP and TSP. There is nothing great in
this particular sentence, because we know that as pore water pressure keeps increasing; ESP
keeps reducing from the total stress. So, let us again understand this in terms of t-s, s’ plot, let
us say P is the starting point. Now, P and P’; it is one as total stress, the other one is effective
stress; is separated by u0 because that is the initial pore water pressure as you can see in this
particular plot.
Now, for total stress path, that is if you draw a total stress path from P, now, what will be the
slope in t-s, s’. This is a case of conventional compression test and we know that in
conventional compression test, the slope is going to be positive one that is 45 degrees. So,
that is what is shown here and this is TSP that is the total stress path. PQ represents TSP.
Effective stress path we know that because of 1 / 1 – 2 Af; this is the typical condition where
it is moving towards the kf line.
For example, in the case of if Af is equal to 1 it is equal to negative one; will be the slope. So,
P’ Q’ represents the effective stress path. Now, let us try to understand this particular figure
445
more in detail. Now, let us say like this line is parallel to PQ and this is separated by u0. Now,
then only we will be able to know like there is an initial pore water pressure of u0. During
shearing, what will be the kind of pore water pressure that gets generated?
Now, u is the horizontal distance between ESP and TSP that is the total pore water pressure
that is always it will be u0 which remains constant. You can see here plus this particular
whatever is these differences that will be the Δu. Now, why did I stop? The total stress path
here, why it is not keep on increasing? Why? Because the failure has already happened when
the effective stress path has just meets the Kfc line.
So, that is this particular point and there is no soil beyond this. So, why? There is no point in
increasing or extending the TSP beyond this point. So, that is why this dashed line has been
drawn. Now, u constitute that is the total pore water pressure constitute u0 and Δu, which is
due to the deviatoric stress application. Let us see how it looks like. Now, this parallelogram,
this one, it represents u0.
For any instance of shearing, it is going to be this u0 as this going to be there and when we
start the shearing from here, when it reaches here, so, this particular, this one, is the Δu. So,
you can see here at any particular point, it is Δu + u0. So, this part of Δu comes from shearing
and u0 is the initial pore water pressure. At failure still that u0 is there; you can see that at
failure the pore water pressure is more and that is given by Δuf.
Now, Δuf + u0 gives this particular term u. So, that is uf at failure. So, ESP, it shifts towards
left and approaches towards failure line. And why? Because this is a case of normally
consolidated soil and we can always take A value equal to 1 which gives a negative slope.
Now, we know that Δu = A (σ1 – σ3). Please note that if you refer to the expression for
Skempton’s pore water pressure we have used Δσ1 – Δσ3.
Here, we are just writing it in a general term just to make us make understand like where t is
going to come here. So, Δu = A2t. So, this will be 2 of t. So, A = Δu /2t. So, it is just for our
understanding like Af = Δuf / 2 t so, where t is going to come because when we discussed
Skempton’s pore water pressure equation, we did not have to actually tell about t parameter
there.
446
So, here we are just writing for our understanding and Δσ1 – Δσ3 still holds good. Here, it is
only to make you understand how the t comes into picture.
Now, if you see this in q-p, p’ plot, I am telling about un-drained stress path for normally
consolidated soil. Slope of TSP is 3, which we know which we have already seen. And that is
not going to change. This is the TSP which is given by PQ with an inclination of 3. Why it is
3? Because it is conventional triaxial compression test. And slope of ESP is equal to 3 / (1 – 3
A). We know this.
Now, there is something interesting fact which comes out of this expression. Now, for elastic
behaviour of soil A is equal to 1 / 3, this we have already seen. Now, if you substitute A
equal to 1 / 3 in this expression, we know that that is going to be infinity that means Δ p’ is
not going to change during shearing. For what condition we need to understand this very
carefully, 3 / 1 – 3 Af or 3 A.
In general is the slope of effective stress path in q-p, p’ because this particular understanding
is useful when we discuss critical state soil mechanics. So, 3 / 1 – 3 A is the slope of effective
stress path when it is drawn in q-p, p’ plot. A is equal to 1 / 3, we know for elastic behaviour
that comes from the derivation of Skempton’s pore water pressure equation.
If you substitute A equal to 1 / 3, we know that the slope of effective stress path is going to be
vertical. Is it going to be vertical till the failure line? No. It is going to be vertical, so, long as
the soil behaves elastically that is the initial portion of shearing. At some point of time, soil
447
would yield. So, the behaviour it changes from elastic to plastic behaviour after yielding. So,
the value of A equal to 1 / 3 yield is there.
So, it is vertical only till the point the soil yields. So, that particular aspect we will see here.
From p’, so, this till yielding ESP will be vertically upwards. This information is very
important for critical state. After yielding, ESP deviates leftwards towards failure lines. So,
once it yields, then the response is the same. So, here now, how to define the yield point
when it is going to shift leftwards. This things will be very clear when we discuss 4th
module.
So, here this is the ESP. You can see here it is vertical till this particular point and here the
yielding has occurred and then it moves towards the left, towards the failure line. So, P’ Q’ is
slightly different in q-p, p’ plot and this P’ Q’ gives the effective stress path. Now, this is the
point of a yielding and up to yielding, it is elastic and hence it is vertically upwards which
comes from this again Δu = A(σ1 – σ 3).
You can replace A into q. So, A = Δu/q. Af = Δuf /q. So, this is again where the q comes into
picture or where t comes into picture which we have already discussed.
Now, let us see un-drained test on over consolidated soil. Now, we have discussed normally
consolidated and we know where the difference is going to come, it is basically dictated by
the value of A parameter and this we already have seen. So, let us see this in more detail. So,
448
this is the typical stress strain response you can see that it is peaking and then exhibits a strain
softening behaviour.
For OC, what happens because of this particular aspect? u decreases, now, NC, u increased.
Here, it is decreasing as the shearing happens why, because of dilation. ESP is governed by
decreasing u. Now, this decreasing u will have its effect on effective stress path. Let us say
that u0 is the initial pore water pressure and this is u equal to 0. Now, this is positive. Here,
upwards is positive and here, it is negative. Now, from u, we can see that it initially increases
and then starts reducing.
Now, this particular point at this particular point Δu start becoming negative that means then
you can see that it keeps on reducing drastically to a final value. And if you see this particular
point, this is the point where pore water pressure actually becomes from a positive value to
negative value. So here from here, this is the zero, so, where it crosses; it crosses from here.
So, this particular portion exhibits negative pore water pressure even though there are 2
things here. One is, disagreement increment is positive or negative and here, the actual value
of u is negative or positive. Now, from here onwards, it starts becoming negative. And from
here, you can see that u actually becomes negative, the magnitude and towards this will be
plus u. A value changes during shearing. Now, because of this, we know that A value is not
constant.
So, at some point, A value becomes negative. So, it basically from here onwards, A will start
becoming negative and finally, it will reach a negative value for OC. So, Af for heavily OC
soil is negative. Final Af will always be negative. So, decreasing u and Af becoming negative.
This would influence the ESP of OC soil. So, both because of decreasing u, there will be a
condition of Af becoming negative; both these influences the effective stress path of OC soil.
Now, we again consider t-s, s’ plot. This is the TSP which is PQ as like in the previous case
and that is not going to change. The starting point is because there is a initial pore water
pressure, it is at p’. ESP of OC soil shifts towards right. Why? Because of decreasing u. So,
let us say, this parallel line P P’, this line is dashed line it is extending and from P’, we can
see that there is a ESP curve that is shifting towards right.
449
Initially, it was for NC it is towards left. Now, it is towards right. This is the ESP. You can
see that at one particular point, this ESP crosses this particular point. So, this is the initial
pore water pressure which is there in the sample. Now, what is actually happening the final
pore water pressure is becoming less than the initial pore water pressure. So, point R is the
zero u change.
So, here this particular point R means, there is no change in pore water pressure because it
comes from here and it reaches here and then comes down. So, there is the change in pore
water pressure. This particular point is zero and then it becomes negative. So, here, it will be
positive. From here, it crosses and then here it will be negative. So, this particular portion,
this final u is less than the u0. So, point of zero Δu or where u start becoming negative.
So, that is a point R. P’ Q’ represents the effective stress path. So, this particular value is the
Δuf and that is negative Δ uf. In the case of NC, what has happened? Since, this is joining on
the other side, we have u0 + Δuf whereas, in the case of OC, we have u0 – Δuf. So, this u0
minus this value. So, that is the final point Q’. So, let us say, this particular curve is for
heavily over consolidated soil. Now, for ESP likely over consolidated soil; may not give a
negative view.
So, in the case of HOC, it gives negative value whereas for LOC, it may not be negative, but
it is still, it is inclining towards right but it is not crossing u0 that means this particular value
the final u value is u0 plus this value, not minus. So, final u of LOC does not fall below u0.
So, it is. In this particular case, it is falling below u0 whereas in lightly over consolidated, it is
generally; it is u0 plus this particular increment.
So, finally u of HOC fall below u0. So, this is the comparison between HOC and LOC how it
changes with respect to pore water pressure.
450
Now, we have discussed for a typical case of OC and NC specifically, I just want to highlight
what is the stress path for UU test on NC soil. We have already discussed this. These are the
total Mohr circle for 3 different combinations that is T 1, T 2, T 3. Mohr circle for total stress
this we have seen. And along with that, I have just added UCS which is a special case of UU
test. So, all the 4 Mohr circles are total Mohr circle.
Now, for all these we have seen that for UU test there will be only one effective stress circle
and that is given by this dotted circle and that is the Mohr circle for effective stress and this
gives the Mohr coulomb failure envelope which is φu = 0 this is the total stress envelope and
Su is the un-drained shear strength. Now, let us see in t-s, s’ plot. Now, this particular radius
of the circle is represented by t.
So, in kfc is this one and sin φ’ is the slope. You have total stress envelop. Su is the same
thing. So, it is Su. This corresponds to the Mohr circle T1. So, this is the stress part, which is
total stress path which is T 2 and T 3 and in the case of UCS, it has to start from zero and this
is at 45 degrees because it is in t-s, s’ plot and this gives T 0 which has to be starting from the
origin.
And this is the effective stress path. You can see that the effective stress path is at the
meeting point of un-drained strength and this drained strength that is kfc = sin φ’. You
remember when we discussed about UU test, we have shown that one cannot draw the
effective stress envelope from UU test because we have only one effective stress circle, but if
we know φ’, then one can always draw a tangent to the effective stress circle.
451
For example, if you have it here, we can always draw a kind of failure envelope
corresponding to this, but for this, we have to know what is the value of φ’. Now, if the φ’ is
known, kfc can be drawn and the effective stress path will meet at this particular point. This is
the effective stress path corresponding to UU test, but that will be at the meeting point of kfc
and the total stress envelope. And this slope is 1 / 1 – 2 A that is all related to some additional
cases of un-drained shear strength and the stress path representation.
We will just do a quick example of finding out the un-drained shear strength from whatever
we have learned. What is given? Undisturbed soil is sampled from a depth of 5 metre in a soft
lightly over consolidated clay for which coefficient of lateral earth pressure at rest is 0.7. Unit
weight of soil is 17 kN/ m3. Cohesion is 0. An effective angle of internal friction is given to
be 22 degrees.
Water table is at a depth of 1 metre. Unit weight of water is 10 kN/ m3. A parameter at failure
that is Skempton’s A parameter is 0.8. We are asked to determine the un-drained shear
strength. We know the expression for un-drained shear strength. We have derived it which is
tfU = (c’ cot φ’+ s’i) / (cosec φ’– 1 + 2 Af). It is sampled from the field undisturbed.
So, here, we need to use s’i and we know that s’i = - ui. So, this relationship also we know.
And ui = - (σ’v0 + 2 σ’h0) /3. So, we need to find out σ’v0 and σ’h0. σ’v0 is equal to 16 * 5
because it is sampled at a depth of 6 and the unit weight is given as 16. So, that gives 80 kPa.
452
σ’v0 is equal to 80 minus that is pore water pressure, static pore water pressure is 10 * 4 that
will give 40 kPa.
You can see here, it is 10 and the groundwater table is at 1 metre. So, the remaining 4 metre
will be the water table. So, that will give 40 kPa as σ’v0. So, this is obtained. σ’h0 will be 0.7
into 40 kPa that gives 28 kPa. So, ui can be determined minus 40 + 2 (28)/3 that gives minus
32 kPa. So, s’i is now known. This is known.
φ' is given. c’ is zero, because it has been already given and then all other Af is also known.
So, s’i substituting it, we get tfU is equal to 10 kPa. Just to give an example of how do we use
tfU expression.
So, let us summarise today’s lecture, the expression for un-drained shear strength was
determined. For a soil sample from the field, there will be a negative pore water pressure of
minus ui. In fact, this we have already summarised before as well. The initial state will be
isotropic because it is pore water pressure. Expression for un-drained shear strength modified
for initial isotropic condition. So, this because of the sampling and minus ui, the starting point
will be at isotropic line.
Un-drained stress path for normally consolidated and over consolidated soil is discussed. For
NC, ESP shift leftwards towards failure line. In q-p, p’ plot, ESP is vertical for elastic
behaviour and deviates left after yielding for normally consolidated soil. For heavily over
consolidated, ESP shift rightwards towards the failure line due to decreasing u and negative
453
Af; we have seen that. Final u of HOC is less than the initial pore water pressure because of
this negative pore water pressure.
For likely over consolidated soil, even though the ESP shift rightwards that we have seen.
The final u is greater than u0. So, the final u always be u0 + Δu. So, that is all for this lecture
and we have completed the effective stress path for un-drained condition. Now, we are left
with ESP for some specific field problems. So, that is all for now. Thank you.
454
Lecture – 39
Stress Path-Field Cases - I
Okay, welcome back all of you. In the last few lectures, we were discussing about stress path,
we have seen some common cases of stress path and we have understood how to plot stress
path for laboratory cases. We have discussed it for drained and un-drained cases. Now, in this
lecture onwards, we will see how these stress paths can be applied for actual field problems
and we will see how to interpret distress path for a given field situation.
So, let us first take the case of an embankment. So, here, it is a manmade embankment and
hence, what is the consequence of the construction of this embankment is that the stress
acting on the soil increases. So, immediately below the embankment, if we consider a
cylindrical soil mass, then what happens is because of this embankment there is an increase in
the stress acting on this.
Now, this is very much similar to a typical compression of this particular element. Now, what
is the consequence of this? On a cylindrical soil element which is away from the
embankment, let us say the case B wherein, there is a stress increase that because of the stress
acting here that will be delegated in the horizontal direction, because of which the radial
stress increases in the case of sample B that is a cylindrical sample B.
455
So, these 2 are very much similar to the cases that we discussed. Let us take the case of
typical compression to actual compression. And in the case of the sample B, we are
increasing the radial stress. So, when we are increasing the radial stress, it is a typical case of
extension. So, these 2 cases are very much similar to what we discussed. Now, let us take
another case where we talk about excavation and there is a slope formation. Because of this
excavation, there is a release of stress.
Now, if you consider a soil mass on the slope, what happens is the lateral stresses are
released, whereas the actual stress remains more or less same. Now, actual stress is same and
the radial stresses getting released means, it is a typical compression that we discussed in the
lecture that we dealt with laboratory triaxial testing. Now, case D is a case where lateral stress
remains same, but then the actual stress gets released.
So, that is also a typical case that we have discussed and that creates a kind of extension. So,
let us summarise these cases A, B, C, D into a table. So, here, what is the stress increment,
whether it is Δσa or Δσr is stated and once we know Δσa, Δσr, we can find out what is Δt/ Δs.
And now, we will be typically using t-s, s’ plot.
So, for case A, what is happening is actual stress increases and hence, it is + Δσ and that is
what Δσa and Δσr is 0. And this is a case where the stress path Δt/ Δs will have + 1 that is
around 45 degrees in the positive direction. Now, in the case of case B, where Δσa is zero,
there is no change but the lateral stress is increasing. So, Δσr is equal to + Δσ.
Now, when you substitute this in Δt/ Δs, we will get minus 1 as the slope. Now, case C is the
case of excavation, where Δσa here is 0. Δσr is released. So, it is - Δσ and the slope will be
again minus 1 and in the case of D, the actual stress is getting released. So, Δσa is equal to -
Δσ. Δσr equal to 0. There is no change and hence, Δt/ Δs is equal to 1.
Please try to work it out whether Δt/ Δs is correct or not that will give you an experience of
the determination of slope and what we need for plotting this stress path is the slope of the
stress path.
456
So, let us try to understand this on t-s, s’ plot. This t-s, s’ plot and let the starting point be
same, let us assume. So, the first stress path that is corresponding to case A is typical drained
stress path. So, let us first talk about drained stress path and those that an inclination of 45
degrees and Ad the presence the case A. Now, case B that is lateral stress increase. It is
towards extension now, when the stress path moves towards the blue line, that means it fails
in compression.
And when it moves towards red line, it means that it fails in extension and we know that the
case B is a case of extension and hence, with the slope 1, it fails at extension failure line and
that is given by Bd that is B drained condition. And for C, it will be negative one, but the
cases compression and hence, it moves towards the compression failure line and the last case
D is towards extension.
Why? Because is that case of degrees in actual stress. So, that represents Dd and all the 4
cases were discussed in detail when we discussed about the triaxial testing. Now, we have
drawn drained stress path. What about un-drained stress path? Now, we have already seen
this that un-drained stress path is basically governed by A value and whatever be the
condition the expression remains the same.
So, this is a typical un-drained stress path and depending upon A value, the effective stress
path would either shift towards left or towards a right. So, this is the un-drained stress path
for A and C Why it is written A and C? Because that represents the compression. And the
457
effective stress path for cases B and D which represents extension is given by this. Now,
where does it go and fail that is governed by the A value.
Now, if the A value after substitution that is 1 /1 – 2 A is the slope of effective stress path.
So, when you substitute a value depending upon the A value, if you can decide effective
stress path is in which direction. So, un-drained stress path will depend on the characteristics
of the soil, stress history and the manner in which the stress changes is all these governs the
A value and hence, that gives you an appropriate effective stress path.
So, we need to understand that it depends upon the soil behaviour, it is very much dependent
on the stress history. And it also depends on the sequence and the manner in which the
loading is happening. So that is all about brief introduction on how the effective stress path
and the total stress path is dependent on the various field situations. Now, let us take some
common examples in the field and see how the stress path changes.
So, let us first take the case of infinite slope. And what is an infinite slope? Long slopes that
are treated which are long are treated as infinite slope. For example, a mountain and the
planar area gets extended over a long distance. That can be treated as an infinite slope.
Whereas the smaller slopes for example in the case of embankment, these are finite slope. So
in long, huge hills and big mountains, the slope can be treated as planar and it can be treated
as infinite slope.
458
The stability analysis is based on planar failure surface parallel to the slope. I hope you
remember about how we understand the stability of infinite slope. So, it fails, basically a
longer planar surface which is parallel to the slope. So, let us take a representative infinite
slope, the plane of failure is indicated by FF. This is the plane along which the failure
happens. So, let us presume that this is the failure surface.
Now, let us incorporate water table into the slope condition where WW is the water table
within this slope. Now, let us consider element of slope A, B, C, D and let us try to analyse
this. So, the length of AB is l along the slope. Angle of the infinite slope is β. hw is the height
of water table that is this particular distance is hw from the failure surface. γ is the unit weight
of soil. γ w is the unit weight of water.
Now, if you want to find out the weight of ABCD, this particular element ABCD; ABCD is
nothing but a parallelogram. And hence, this ABCD area multiplied by one that is
perpendicular to this particular plane. So, that will give you the volume. So, weight of ABCD
for unit width. So, that is what I told it is unit width in the direction perpendicular to this
plane.
So, the weight of ABCD is equal to γhl cos β. Now, what is hl cos β? Now, this is the
distance h into this perpendicular distance is lcos β. So, that gives you the area of
parallelogram. So, hl cos β multiplied by γ that is what is given here, into hl cos β gives the
area; into one gives volume; into γ gives the weight of ABCD.
So, forces acting on AD and BC are considered equal and opposite. So, it cancels off and that
does not come into the analysis. Shear stress along DC. Now, let us resolve the forces on this
failure surface. When I say forces, all these stresses acting on any failure surface or plane, we
understand that it is normal and shear stresses. So, we have understood this in our earlier
lectures.
So, let us try to understand what is the normal stress and shear stress acting on the plain, so,
plain DC. So, shear stress along DC is that is the component of weight acting along DC. So,
we know this how to resolve. So, if you have DC, this is weight acting and this will be β. So,
this will be w sin β. So, w sin β will be along the C direction. So, that is what is written here.
459
So, w is gamma γhl cos β sin β/ l * 1. Now, why it is l * 1? Because we have we are
discussing about stress. So, the force component in sin β is in DC direction divided by l * 1 is
the area So, that gives shear stress along DC that will give 0.5γhsin2β. l and l gets cancelled
off; cos β sin β is 0.5 sin2β. So, substituting that we get what is the shear stress.
Now, total normal stress along DC that is what is the normal stress acting that is this is DC,
so, this stress. So, we know that DC the component has been taken, so, it will be weight
component in cos β direction. So, (γh l cos β * cos β) / l into 1 that gives γh cos2 β. All these
exercise we have already seen during our undergraduate. So, we are just brushing it up and
see how we will understand this in the context of stress path.
Now, shear stress and normal stress is assumed to be t and s for the time being. So, let us
assume that the shear stress what we have determined and the normal stress what we have
just found out, this can be considered as t and s, s, s, s’ respectively. So, t is equal to that is
equivalent to shear stress 0.5 γh sin 2 β and S is equal to γh cos2 β.
So, initial u, we also have a static pore water pressure now, that is initial u acting on DC will
be same as normal stress. So, always the pore water pressure is influenced in the normal
stress. So, we take the normal component that is ui = γwhw cos2 β. Same as this only change is
height and the unit weight corresponds to that of water. So, we can find out what is s’ for
effective stress.
But we need s’ = s- u that is γh cos2 β - γwhw cos2 β that gives s’ = cos2 β (γh - γwhw).
460
So, initial condition P, P’, so, let us try to plot these conditions on t-s, s’ plot. So, let us say P
is the initial condition and there is an initial pore water pressure of ui. So, this difference will
be ui, so, P and P’ is separated by ui. Let the soil above and below the water table is saturated.
Let us presume that it is because of capillary rise or whatever the soil sample above the water
table is also saturated.
We are just trying to simplify it because this means that when there is any change in the unit
weight that is going to change the t component as well, I mean t component. So, in this case,
if we presume that soil is more or less saturated, then there is nothing much that is going to
change in t. So, that is the reason why it is assumed that the soil above and below the water
table is saturated.
Therefore, rise in water table will not impact t because t is not influenced by pore water
pressure. So, whether it fluctuates up and down, it is not going to get influenced. What will
get influenced? s component will get influenced. Other s’ will get influenced by variation in
pore water pressure. But, if the soil sample is not considered to be saturated, then t can also
vary because change in unit weight.
When you are calculating t, you know that it is governed by unit weights. So, if it is not
assumed to be saturated, then it will change. If the soil above water table is not saturated,
there will be change in P due to change in unit weight as the soil starts saturating because of
461
the infiltrating rainwater. The increase in water table due to rainfall infiltration would alter u.
So, what is the effect of infiltrating rainwater?
Infiltrating rainwater means, it will rise the water table. And when the water table rise, hw
changes and when hw changes, u will change. So, increasing u will decrease s dash because
of the assumption P remains same, but increasing u will decrease s’, so, no change in point P
that is total stress point is not going to change. ESP follows P’Q’. Now, there is no change in
t but s’ decreases which means that the stress path will be in this direction and that P’Q’ gives
the respective effective stress path.
Now, there is no requirement that Q’ has to be always on k. So, it is a progressive failure

condition. From P’ as in when there is change in pore water pressure, it moves towards Q’. At
some point of time, it fails.
So, if the slope is steep and pore water pressure generated due to infiltrating rainwater is high
enough, failure of the slope would happen when ESP approach Q’. Now, when it is steep, it
will influence the weight and when rainwater infiltrates, u will increase. These 2 components
together would alter the effective stress path and it will proceed towards Q’ to fail. P’ Q’ does
not represent any drained stress path normally performed in the laboratory. This is why we
are discussing this in that much detail.
Whenever we want to do a slope stability analysis, we need to have the shear strength
parameters. Now, if the soil is brought to the lab, we normally do a triaxial compression test.
462
This stress path which is followed in triaxial compression test may not be same as that of the
condition that is actually existing in the field. Now, this stress path P’Q’ is not simulated in
the lab for determining shear strength parameters.
So, that differences is what we have to appreciate in this lecture. So, P P’, this points P and
P’, it can be simulated by applying appropriate σa and σr in the lab. So, the initial conditions
can be mimicked in the laboratory. Now, for simulating this particular effective stress path,
what we have to do? So, we need to increase the pore water pressure because that is what is
happening in the present case.
The slope is not failing because it is getting loaded. It is failing because pore water pressure is
getting altered. So, the same thing if we want to, the same condition has to be simulated in
the lab, then we need to increase the pore water pressure. Now, we know that there is a way
of increasing pore water pressure which we have known for saturation that is by increasing
the back pressure.
At that point of time, it did not destabilise the soil because we increase both confinement and
back pressure together, but in this case, if we do not disturb the stresses, that is axial and
radial stresses and keep on increasing the pore water pressure in terms of back pressure, then
that will simulate a situation which is very close to the infinite slope condition that we just
discussed. So, increase u by increasing uBP, keeping σa, σr constant. So, that will give us this
appropriate effective stress path.
Now, whatever failure and the failures strength parameters that we get corresponding to that
will be realistic condition very close to the field 1. So uBP is increased till failure occurs that
is till the stress path reaches Q’. For accurate stability analysis, the laboratory tests should be
carried out at representative stress level and correct stress path. Representative stress level
means simulating the initial condition P and P’ and correct stress path means how the stress
path moves towards the failure.
Now, unless we discuss this in this manner, that is discussion of P’ Q’, one will not be able to
appreciate what actually is happening in the field. So, this gives a very good opportunity for
us to understand better I mean to say the stress path plot gives us an opportunity to
463
understand better what is realistic in the field. And if one really wants, we can try to simulate
the same condition in the lab and then get the appropriate strength parameters.
So, now consider the conventional triaxial compression ESP that is U1 performed on 3
identical samples. We are just going to see how things will be different if we perform a
triaxial compression test. Now, U1 is familiar to you, because we have already discussed this,
when we discuss un-drained triaxial stress path. So, please refer to U1 is a typical un-drained
stress path.
Now, it is governed by the slope 1/ 1-2A in t-s, s’ plot. Now, what we will do? For finding
out the strength parameters, we consider 3 identical samples and we apply 3 cell pressure and
then get the deviatoric stress at failure plot the Mohr circle or we plot it on t-s, s’ and obtain
the shear strength parameters that is what we generally do. So, let us see, if we do U1 test in
the lab, how it will look like visa- versa this stress path, which is the actual stress path in the
field which is P’Q’.
So, consider the samples to be highly heavily over consolidated, what is the implication of
the statement? The only implication is, A value will be negative. Let us say that A value for a
heavily over consolidated sample will be minus 0.5, it is an example. So, 1 /1 – 2 A 2 into –
0.5 – 2 into – 0.5 will be plus. So, this will be a positive slope. So, we need to understand
now only that the effective stress path in t-s, s’ will move in a positive direction.
464
So, this gives effective stress path of P’1 R’1, P’2 R’2, P’3 R’3. Let us say the first point
simulated in the lab for U1. U1 case of compression is P’1. So, the initial point is P’1 and the
stress path is in the positive direction effective stress path which is P’1 R’1. For P’2 R’2 that is
at a higher confinement and P’3 R’3 at even higher confinement. So, this gives the effective
stress path.
The failure points are defined by R’1 R’2 R’3. So, these are the failure points, you can see that
these points are much higher than Q’, which means, this strength that we anticipate from
laboratory test is much higher than the actual stress strength which is available in the field
when an infinite slope fails due to infiltration of rainwater. So, that is what that difference is
what we need to understand here.
So, R’1 R’2 R’3defines the kfc line. Now, if we try to extrapolate this for the present case, I am
telling about this present case, if we tried to extrapolate R’1 R’2 R’3 downwards, then in the
lower stress region, it may not be representative for over consolidated soil. Why? We have
seen that the failure lines are always non-linear in the case of over consolidated because of
the variation in dilation.
A lower stress region, the dilation will be more or much more free because the confinement is
less. As we increase the confinement, the dilation gets suppressed because of this, there is a
non-linearity that creeps in. So, kfc line may be non-linear in the lower stress region due to
dilation associated with low confinement but I am not showing that in this particular plot
because we know this much earlier.
So, we just want to understand here what is that a normal compression test in the lab may not
simulate the actual field condition. So, you may naturally have a question that we are
working with the highest strength, then how we get an appropriate result. Now, if any kind of
work which we execute on a given slope, we always have a higher factor of safety. We take a
factor of safety margin. So, it is mainly because of that.
And please understand, we are not dealing with a manmade slope when I say infinite slope, it
is a natural slope. So, natural slope normally fails during rainfall or due to any other forces,
but generally a lot of failures happens during rainfall. Now, this rainfall increases the pore
water pressure. So, the actual cause of failure is P’Q’, P’1Q’.
465
For the case when water table is well below the failure surface, now, we have considered the
case where water table is above the failure surface. Now, here let us try to see what difference
will it make from P’Q’, if the water table is well below the failure surface. Now, what is the
implication when water table is below? When the water table is below, the pore water
pressure in the along the failure surface will be initially negative.
So, initial u will be negative because it is an unsaturated state. So, let us say t is the initial
point, total stress point and there is a negative pore water pressure so, it will shift towards
right; if it is positive, which shifts towards left. So, the effective stress point the initial point
will be t’. So, during rainfall, the water infiltrates. Now, when the water infiltrates, t will
increase due to change in unit weight of soil.
Now, in this particular case, the unit weight initially was bulk as it gets saturated, it becomes
gamma set which is more and hence, t will increase. So, we know now the direction in which
the stress path would move because of increase in t, it has to move towards in the upward
direction. Now, due to increase in u, what happens to s’. As u increases, s’ starts decreasing,
so, the stress path should move towards left and towards upwards.
So, the total stress path remains same it is in this direction whereas, the effective stress path is
T’U’ that will be in this direction. So, this satisfies what we have told that is t is increasing
and s’ is decreasing, because of the increase in pore water pressure. So, T’U’ represents the
466
effective stress path. So, this total stress path is drawn only up to this because the soil sample
fails here at U’.
So, that is why this dashed line shows the limit of total stress path. So, effective stress path is
T’U’. In laboratory triaxial test, initial field stress condition has to be simulated which means
to say it is T and T’. In shearing stage, how to simulate this particular condition? In shearing
stage, we also need to increase the actual stress as well as we need to increase the pore water
pressure.
Here, in the shearing stage, both the actual stress, actual stress increase means, we are
increasing the deviatoric stress. This deviatoric stress increase is increase in t because t is
same as σ1 – σ3/ 2. So, σ1 – σ3 has to increase because that is the case discussed here. t is also
increasing and the s’ component is reducing because of the increase in pore water pressure.
So, when we simulate this in the lab, we need to make sure that the actual stress is increased;
at the same time the back pressure is also increased, so, that u increases that gives the
effective stress path P’U’ if we want to simulate that in the lab. So, that is all about infinite
slope.
Now, let us see a small example related to this and related to how to use stress path for a
given field condition. An infinite slope with slope angle 18 degrees has failure surface at a
depth of 3 m. Soil has unit weight of 20 kN/m3. Water table is that a depth of 2 m from the
467
ground surface. Using stress path, evaluate whether the failure would occur if the ground
water rose to the ground surface.
Friction angle is 30 degrees. γw is equal to 10 kN/m3. These are this given condition in the
field. We just need to understand by plotting the stress path whether the failure would occur
or not, if there is a groundwater fluctuation from the given depth to ground surface. So, let us
try to solve. It is on t-s, s’. t = 0.5γhsin 2β, substituting for β will get 17.6 kPa.
And this total stress path point, initial point is γhcos 2β which is equal to 54.2 kPa. So, the
initial point, we can mark, let this be the initial point P. Now ui = γwhwcos 2β, which is 9 kPa.
Please substitute this and see for yourself. So, s’ = s- ui, which is equal to 54.2 – 9, which is
45.2 kPa. So, let us say that this is P’, P remains same. Now, this is the ui that is 9 kPa.
Now, let us assume that the stress path reaches Q’ when there is groundwater fluctuation.
Now, actually, whether it is touching the failure line or not that is what we need to evaluate in
the given example. Now, if it is touching, then it means that the failure would occur. Now,
this is a given ESP for this infinite slope condition which we have just seen. Rising of
groundwater to the ground surface, let us try to understand this particular case.
t is equal to 17.6 kPa which is not changing. u changes to 27.1 kPa. So, s’ = s- u which is
equal to 54.2 minus 27.1 which gives 27.1 kPa that is if the groundwater table is rising, then
the s’ becomes even less that is 27.1 kPa. Now, let us try to understand whether this is a
failure state or not. So, let us again re-plot t-s, s’. sin 30 which because the failure line is
defined, we know that friction angle is 30 degrees and tan α for this is equal to sin φ’ which
we have already seen.
So, we know that it is sin 30 which is equal to 0.5. Now, let us say that this distance. Now, s’
is 27.1. Where it is and what will be the strength available at this particular point? Let us try
to see. So, tf is the strength which is available at s’ of 27.1. Here, it is 27.1 So, tf divided by
27.1 equal to 0.5. So, tf is equal to 13.6 kPa. Now, what is the strength available at s’ of 27.1?
The strength available is 13.6 kPa, but, what is the stress t at 27.1? The stress t is 17.6 kPa.
Now, the stress t 17.6 is greater than this strength. Hence, the failure would occur. So, that is
how we try to analyse the given situation. So, the pore water pressure increase led to this
468
failure which means that there is no scope for the pour the pressure to rise up to the ground
surface before that the failure would happen for the given situation. So, this is a small
example to show how to use the stress path for understanding and analysing the field
situation.
We can anticipate when the failure is going to happen. So, all these exercises can be done
using a stress path plot.
Now, let us deal with un-drained slope excavation. Manmade slope construction in place
results in un-drained condition during excavation. So, here also we are dealing with slope but
a finite slope instead of an in finites. Now, this is a given finite slope, this has been created
because it got excavated from this end. So, consider initial isotropic condition with initial u0.
For simplicity, water table is kept constant and why we want to maintain this?
Normally, it is not maintained in this manner. First of all, the water table will be slightly
lower and because of that the phreatic line changes and we do not want to complicate this
problem with that. So, we are considering water table to be constant and hence the
equilibrium pore water pressure after excavation, there will be change in pore water pressure
and again the pore water pressure reaches to an equilibrium value.
Now, if you consider water table to be seen, then after equilibrium, whatever pore water
pressure attains will be same as the initial pore water pressure because the height of the water
table is not changing. So, let us try to understand. Since the initial condition is taken to be
469
isotropic and with an initial u0, let us say that the point will be on the isotopic line. So, P P’ is
drawn now. I hope you are familiar how to at least mark the initial points. So, P and P’ at a
difference of u0.
There is a release of radial stress because of excavation. Here, a typical case of release which
is happening. So, TSP that is the slope of total stress path Δt / Δs will be – 1. So, it will be in
the reverse direction with a negative slope. Let it be PQ which is the TSP. Immediately after
excavation, there will be change in pore water pressure. Now, this is a kind of loading in post,
but here it is a release that is happening.
But whatever P it is one of the same, there is a stress change which is happening because of
this there is a change in pore water pressure. This change in pore water pressure can be taken
as excess pore water pressure change. So, that is why there is ui excess becomes negative.
Why? Because there is excavation, it is released and whenever there is release, there is a
tendency for negative pore water pressure.
So, ui excess becomes negative. So, ui is less than steady state final uf. Now, this ui will
whatever at the point of Q, when the excavation is made, that ui will be generally less than
the final pore water pressure. So, with time, negative ui excess will dissipate and slowly u
will increase. So, initially immediately after excavation, there will be negative pore water
pressure generation, but with time, this negative pore water pressure will dissipate.
And slowly the pore water pressure will again regained and come to its original value of u0.
So, uf which is the equilibrium pore water pressure in the final stage will be more or less
equal to u0. So, that is why it is told at the beginning the pore water pressure ui will be less
than the earlier pore water pressure that is u0. Why? Because there is a negative pore water
pressure and it gets reduced.
The positive pore water pressure gets reduced. So, this is what it means. So, when you have
this, this is the point where there is an initial pore water pressure ui. So, this pore water
pressure which is marked as ui, so, this is ui, this ui is generally less than u0 or maybe it is this
is the final point of effective stress path, this ui will be less than the final uf. Now, this P’Q’ is
the effective stress path and Q’R’ is the point when it reaches equilibrium.
470
So, P’ to Q’ is immediately after excavation and Q’ R’ when the negative ui excess start
dissipating. So, when negative pressure start dissipating, the pore water pressure increases
and Q Q’ R’ is the corresponding effective stress path. So, this is the point. So, that is why a
horizontal line is drawn. So, TSP point is Q. So, steady state final uf that is when it reaches
equilibrium is same as u0 when the condition is un-drained.
So, this is the final pore water pressure, which is the steady state or equilibrium pore water
pressure uf and you can see that ui is less than uf. Now, this uf is equal to u0 because of our
assumption, but provided the condition is un-drained. When the water table start changing
because of drain condition, then the final pore water pressure will not be same as that of the
initial pore water pressure. So, this is the condition where it is un-drained condition and water
table still is uniform.
Now, we will see 2 different cases that is steady state final uf will be different from u0 as the
water drains off the excavation. So, in the previous slide, we have assumed that there is no
change, but now if it starts draining off, then the final pore water pressure uf will not be same
as that of the initial pore water pressure u0. Slope would fail if the ESP that is P’Q’ reaches
failure during un-drained excavation.
Now, it can so happen that from P’ to Q’, it just extends and join the failure line. So, it can
fail in that manner as well that is during the un-drained excavation itself, it can fail. P’Q’, it
represents the excavation. So, during excavation itself it can fail or it can also fail, because Q’
R’ can reach towards the failure line that is after some time completing the excavation.
471
So, here the excavation is over. So, Q’ R’ is the manner in which the pore water pressure
increases and it can extend and it can fail at this point. Now, that is what it means. So, there
are 2 ways the failure of this particular slope can happen. The distance of Q’ or R’ whichever
be the case either this case or this case is the factor of safety of the slope. Now, if it is here,
then this is the margin of safety.
If it is here, this is the margin of safety and factor of safety decreases with time. Why? Here
the factor of safety is this much, but as the time progresses, u increases and Q’ R’ is the stress
path. So, the factor of safety reduces and as the u increases to the equilibrium value. So, the
most critical time of a slope is in long term when u reaches equilibrium at steady state
seepage that is u is increasing. So, the critical path of a slope under excavation is in long
term. Why?
We know that the pore water pressure keep on increasing with time. So, if the equilibrium
value is less than the failure state, then the slope is safe or it may free. Slope should be
designed for long term fully drained condition when it comes to the excavation. So, dashed
line P’R’ represents the effective stress path for drain case that is when u is constant. So, if it
is drained, we know that is the same slope and P’ R’ represents this condition.
Now, slow failure due to instability soon after excavation. So, when the slow fails, the total
stress would change due to reduction in height and slope angle. So, what is happening? When
there is instability, there is a failure which happens to this slope and because of this failure,
472
both the height and the slope angle changes and it tries to achieve an equilibrium condition.
So, if θi is the initial angle, after failure the slope will look like this because it needs to shut
off this much portion and that will give the new slope angle which is θc.
So, effective stress path for failures P’Q’ and total stress path is PQ. So, we will see again
what is happening. So, the initial condition is same and this is the total stress path and P’Q’ is
the effective stress path that is immediately after excavation, so, the excavation stops here.
TSP and ESP and ui is the initial pore water pressure at the point when the failure takes place
due to excavation.
Since, a failure happens during un-drained condition, this strength can be considered equal to
un-drained shear strength because it is completely under un-drained condition. TSP would
have continued to s corresponding to θi. So, let us say that this total stress path would have
continued to this point corresponding to this particular height of θi, but it is unstable due to
excavation and hence, it falls to, it cannot sustain θi corresponding to the θi.
So, it cannot achieve this particular state s rather slope fails and take a new slope geometry of
θc for which total stress path correspond to PQ. So, due to failure, it brings down and hence Q
is the point which corresponds to the new slope geometry of θc.
Now, if the slope fails after some time of excavation, so, what will happen? Slope does not
fail immediately after excavation as ui does not cause the ESP to test the failure line. So, this
case is different, but the failure happens after some time of excavation. So, the initial
473
condition is same; total stress path is same; effective stress path you can see that the point Q’
is not touching the failure line. So, it is not failing immediately after excavation.
Now, ui has not caused the failure. A failure occurs R and R’. When u becomes the stable
value of uf, this is the point R and this is the point as the pore water pressure start increasing
that is the final equilibrium value, it reaches at this particular point. This is not final
equilibrium value as the pore water pressure increases, it touches the point failure line at R’
which corresponds to the value uf.
So, uf is the value which causes the failure at R’ and then the slope changes its geometry from
θi to θc due to failure. Total stress now would reduce after failure. Why the total stress now
has to reduce? Because the slope height is reduced and the failure has occurred, so, the slope
angle got changed. So, definitely the total stress path also would change because the total
stress is reducing.
Earlier cases, total stress did not change. So, TSP moves in RS. So, it moves downwards due
to unloading as θi reduces to θc. But, u continues to increase because it is a gaining of the
pore water pressure till it reaches a steady state equilibrium value which is denoted as maybe
ufc. So, ESP moves in R’ and S’. Now, what is the effect that happens to effective stress path?
So, it is not terminating here, now, it has achieved a new geometry. Now, due to unloading
because of the θi reducing to θc, what is happening to ESP? ESP moves towards S’, so, R’S’
corresponding to a reduction in total stress path. So, slope will reach a stable state. Now, θc
will become a stable state when u reach a final steady state value which is equal to ufs.
So, ufs is the stable final value of pore water pressure. So, where S which means that S’ So,
R’S’ at this point, this slope will be stable corresponding to θc and ufs is the final equilibrium
pore water pressure. Change in u with time till it reaches equilibrium decides the stability of
slope under excavation. So, 2 cases we have discussed. One is immediately after construction
it fails; after some time it fails.
474
So, let us summarise today’s lecture. The lab ESP is identified for different field conditions in
the very first slide what we have discussed. Stress path variation for infinite and finite slope
is discussed. For infinite slope failure condition due to groundwater fluctuation has been
demonstrated in stress path plot. We have seen that as the groundwater rises the pore water
pressure increases resulting in a decrease in effective stress and that causes the ESP to
progress towards the failure line.
This ESP is different from the conventional triaxial compression test stress path and that we
have already seen in this lecture. The ESP to failure when water table is below the failure
surface of infinite slope is discussed. Under un-drained excavation that is for a finite slope, u
decreases initially during excavation because of release and then it increased when
equilibrium value. It gains the pore water pressure again.
And ESP for failure due to excavation, it is discussed for immediately after excavation and
after some duration. Slopes should be generally designed for long term fully drained
condition, which is a more critical condition because generally it fails because of the change
in pore water pressure. So, once the equilibrium pore water pressure is achieved that
determines whether the slope is stable or not. And this is particularly applicable for
excavation.
Change in u with time till it reaches the equilibrium decides the stability of slope. So, that is
all for today’s lecture. We will see a few more cases of stress path under field condition in the
next lecture. Thank you.
475
Indian Institute of Technology-Guwahati
Lecture-41
Stress Pathfield Cases-II
Welcome back all of you, in the last lecture we were discussing about the stress path related to
field cases. We will continue on that there are some more cases which we need to discuss. So, in
the last lecture we essentially focused on slopes. So, in this lecture we will see some more
examples relevant to field cases.
First is the stress changes beneath foundation, a very common case which that we encounter. Let
us say let us consider the undrained loading of foundation, now is a typical foundation where it is
the soil is under a submerged state because the water table is at the ground surface. This is a case
where the foundation is resting on the ground for an example. Now the foundation transfers the
load onto the soil.
So, there will be these the soil element beneath the foundation is stressed because of this load.
Earlier it was stressed under it is own weight, now because of the placement of foundation some
additional stresses are imparted onto the soil. Stress conditions we will see again in terms of t-s,
s’, we can study the stress path in any form the most common forms are t-s, s’, and q-p, p’
476
because those are in terms of invariance. So, generally q-p, p’ is also preferred but since t-s, s’
and q-p, p’ plots are more or less similar because one this is a very marginal difference in terms
of slope.
The procedure remains the same, so we will discuss in terms of t-s, s’. You can try doing all
these and plot it on q-p, p’ plot as well. Foundation loading and the stresses imparted in the soil.
It is a typical case of compression and we have seen very specific stress paths in our previous
lectures where we discussed about compression, extension and different manner in which this
compression and extension can be created.
So, here we are discussing whatever is the foundation loading that is acting is a typical case of
compression. Assume the initial condition to be isotropic for simplicity. So, this t-s, s’plot, since
this condition is invoked we know that the starting point will be on s, s’, so let p be the total
stress point and if there is an initial pore water pressure of u0. So, the effective stress point will
be at p’.
Now we will have the total stress path from p, now the same total stress path when you draw at
p’ that we have seen in one of the examples earlier, so that is total stress path minus the initial
pore water pressure. So, it is one other same, it is a parallel shift that is happening and sometimes
that helps us to understand the pore water pressure and effective stress path difference better. PQ
represents the total stress path which is similar to U1.
Now U1 we have already discussed previously, so PQ is the total stress path. And loading from
the foundation increases the stresses developed in the soil. Effective stress path is from P’ it will
start P’Q’ corresponding to undrained loading. So, how it will look like? So, it will be P’Q’ and
this essentially would depend upon the slope of effective stress path and the A value. This is how
it will look like and if we draw this parallel total stress path here, so this total stress path is TSP -
u0 plot, this is u0 and this is the actual increase in the pore water pressure, this one.
I think this we have already discussed in the previous lecture. Now how this ESP would look
like, it will depend upon the soil characteristics and stress history. You will see that, this
477
sentence is getting repeated in all the examples, this is intentional because we should not forget
that ESP is essentially governed by A value and as well as the soil characteristics.
Now stress changes beneath the foundation again the remaining aspects. Let us say that ui is the u
immediately after construction. Now there are 2, 3 types of pore water pressure that we are
discussing; one is the initial pore water pressure that is u0. The other one is since it is an
undrained loading at the end of construction there will be an pore water pressure that pore water
pressure is initial pore water pressure that is ui, that is immediately after construction.
So, that is what is shown here ui. Now the initial excess pore water pressure which is denoted as
ui excess is positive because this is a case of compression. And hence we know that the excess
pore water pressure that gets generated in the beginning will be always positive. The positive ui
excess dissipate with time and then u reduces. So, the total ui comprises of this excess pore water
pressure as well as the initial pore water pressure.
Now what it says is that as the time progresses we know that the positive pore water pressure has
to dissipate. Now after dissipation the u starts reducing, so it will no longer be ui, it starts
reducing. So, what is the implication of this reduction in u on ESP, this ESP moves towards R
dash which corresponds to compression of the soil mass. As the soil gets compressed because of
dissipation of pore water pressure, it moves towards R’.
478
And that is in towards this direction that is given by R’. Because there is no more shear stress
getting added because the foundation load has been fixed, that is the maximum that is Q’. But
because of reduction in pore water pressure that gets affected in s’. So, now you can see that why
this pore water pressure? That is initial power pressure at the end of construction period why it
comes down?
It comes down because this much component is the excess pore water pressure that gets
generated because of the foundation load. So, that gets dissipated and then the stress path is Q’
R’. And Q’ R’gives the idea about the compression that is taking place within the soil mass. So,
what is this then left out that will be more or less equal to u0 that is the final equilibrium pore
water pressure?
So, there is an increase in s’, final equilibrium pore water pressure is uf, now this is the final
value that is where the effective stress path stops. So, uf is the equilibrium pore water pressure
which happens after the dissipation of excess pore water pressure.
Foundation will fail; we know that if Q’ touches the failure line, what does that mean? Because
up to Q’ it represents the undrained loading condition. So, if the foundation has to fail, it has to
fail during this short term undrained period, why? I am talking with respect to a specific loading
479
condition, because if Q’ touches the failure line then only it is bound to fail, I am talking
specifically with respect to undrained condition.
And if the foundation crosses this period like undrained condition and the pore water pressure
starts dissipating you can see that the effective stress is moving right wards, that is away from the
failure line, so does it fail? No, it can fail only when it touches the failure line. So, in long term
condition a foundation which is subjected to compressive load and for a specific load will not
fail, if it does not fail in the initial period that is during undrained condition.
But definitely it can fail under drained condition if the stress becomes more and more. So, at
drained condition the slope of the effective stress path is I mean to say effective stress path and
total stress path both are same, that slope is different and that will be at an inclination of 3. So,
distance of Q’ from the failure line, so this distance, I mean to say this distance represents the
fact of safety.
That is, it gives the margin by which this foundation is safe. So, factor of safety of foundation
under undrained loading increases with time as you can see here. So, this distance goes on
increasing, so it is moving away and hence factor of safety increases with time. As ui excess
dissipate there will be settlements due to consolidations. Now it is moving away, why? Because
it is progressively densifying and the strength is getting enhanced.
So, that is the reason why it is moving away, but what is the cost we have to pay? The cost is the
excessive consolidation, because this densification happens because of the excess pore water
pressure dissipation and void ratio getting reduced. Void ratio getting reduced means settlement
is more, so it is not that the foundation is performing absolutely well. In certain cases it can go
wrong because the settlements may increase even though it is moving away from the failure line.
The soil may not fail but foundation will undergo distress because of this settlement. So, drained
loading of foundation is given by P’, R’, S’. Now if suppose that we have discussed about
undrained loading where pore water pressure is there. Now from P’ if there is a drained loading
that is happening then it will follow P’, R’, S’. Now this is the slope which I talked about.
480
You can see that the margin towards the failure line is this much, that is a huge margin. So, lot of
load can get added up to the soil before the actual failure takes place, but what is the cost? From
here to here it will densify. Now failure is delayed in drained loading, yes fine, that is fine and
that is good also the soil is not going to fail but then it is at the expense of the settlement. So,
more loading possible from P’ to S’ but settlement will be high.
Now due to submergence, now we are considering this it to be submerged condition. Now since
there is submerged condition the pore water pressure change is not going to take place. Now
because of this assumption of submergence the final equilibrium pore water pressure will
become equal to the initial power pressure u0.
Now let us say that it is not submerged condition, then there is a possibility that the final pore
water of pressure is not actually equal to u0, it can be even less than that. And that is generally
governed by in what manner the pore water pressure got dissipated, how it sieved out.
So, that is all about the condition or the stress variations that happens beneath a foundation. Now
let us see another example wherein we talk about stress changes near the retaining wall. Now
retaining wall all of us know, it can generate movement and because of this movement there can
481
be failure due to active or passive condition. Please refer back to the undergraduate portion for
understanding this better.
Now TSP and ESP of soils near retaining wall, it will help to assess the failure condition. Help to
understand whether short term undrained or long term drained conditions are more critical. Now
by discussing these stress paths in the last lecture and just the previous slide. It is very clear that
there is a kind of stability is determined either under short term or at long term. Now plotting
stress path helps us to understand this much better, why?
In the previous cases, in the previous lecture we have seen that with time the effective stress path
moved towards the failure line. In the last example that we have seen the foundation we have
seen that effective stress path moves away from the failure line after construction. So, then that
determines whether it is moving towards or away. That determines whether the given structure or
the given condition will be safe during short term or at long term.
Now what determines this? Essentially during undrained condition the kind of pore water
pressure that gets generated determines this. It can be positive pore water pressure or it can be
negative pore water pressure. Now that determines in what direction the effective stress path
would move further. For retaining wall, shearing caused by excavation is different from shearing
caused by filling.
Now when you say retaining wall, there are 2 possibilities, they are in one case it can get
excavated and in the other case you can have progressive filling of the backfill. So, in both the
cases the kind of stress path will be different. So, let us first consider retaining wall subjected to
excavation. Now we have already seen an example of slope excavation, so it is more or less very
similar to that.
Excavation is kept full of water, now this is what it means, to make sure that the final
equilibrium pore water pressure is comparable to the initial pore water pressure u0 So, that is
what this same statement uf, the final equilibrium value will be equal to initial u0. If the
482
excavation is dewatered, that is during construction of a retaining wall if there is a high ground
water level; definitely the water has to be pumped out.
So, this condition what we have assumed is no longer valid when it comes to actual field
implementation. So, but what is the implication of that? If the water is drained out then what
happens is the pore water pressure that is the equilibrium pore water pressure it is dependent on
steady state seepage flow net. We have already seen what happens if water is pumped out, so
there will be seepage from higher water level to lower water level.
Now this will be governed basically by the flow net under this structure. Now this will determine
what will be the equilibrium final water pressure. Now to avoid that complication for this
example we are considering this water level to be unchanged, so it remains the same. So, uf will
be less than u0, if this condition happens I mean to say the steady state seepage is happening then
definitely the final part of pressure will be less than the initial pore water pressure u0.
For active and passive cases irrespective of in what direction the retaining wall is moving that in
both the cases the shear stress increase would be there, whether it is active or passive it is bound
to fail at some point of time.
483
Now PQ represents the total stress path for both active and passive cases, why is it so? For active
and passive case slope of TSP is negative, you can see that if you substitute for active and
passive conditions, that is how Δσa Δσr changes. You will see that in both the cases you will get
the total stress path to be negative, please verify yourself. Because this we have already
explained in our previous lecture while considering one is constant the other is increasing.
So, P this is the starting point and there is an initial pore water pressure of u0. Now this is the
manner in which the total stress path changes, why? Because the total stress path the slope is
negative, so that is why it is moving in this direction. The shear stress component is increasing
but s’ or s component is decreasing. Now nature of ESP will depend upon soil type and soil
behavior and stress history, this again I am repeating.
TSP and ESP, that is ESP is represented by P’, Q’, R’ is similar to undrained slope excavation,
which I have already told. Now this is the typical response of the effective stress path where this
is P’, Q’, R’. Now P’ to Q’ it represents the construction stage where Q’ is at the end of
construction, where there will be an initial pore water pressure of ui, and with time the excess
part dissipates and then achieves a final equilibrium value.
So, this is the point Q’, so this is the initial pore water pressure at the end of construction period
and this is the final pore water of pressure. Now what is happening here, immediately after
excavation the excess pore water pressure because of the excavation that gets it becomes it
achieves a negative value. So, the initial excess pore water pressure is negative, what was the
kind of water pressure in the case of compression for foundation?
Initially it was positive, here initially it was negative, so it looks fine, since it is negative power
pressure we know that it will improve the strength. But with time what will happen? This initial
negative excess pore water pressure would dissipate. So, ui is less than steady state final uf, here
you can see that, why uf is becoming more than ui ? Because initially the excess pore water of
pressure was negative and that will dissipate.
484
And in the process pore water pressure has to increase from the initial ui value. With time the
negative pore water pressure excess pore water pressure dissipate and that is why u has to
increase from the value of ui to a given value. So, let uf be the equilibrium final value that is the
steady state uf. And this steady state uf, since we have assumed a condition of submergence will
be equal to initial pore water pressure u0. For drained condition uf will be less than u0, and that
will depend upon the seepage flow net.
Now retaining wall for this particular case of excavation it would fail once the effective stress
path touches the failure line. Now this is moving in this direction, you can see that it moves up to
here and Q’ to R’. Now if R’, it extends towards the failure line then once it reaches the failure
line it would fail. If Q’ reaches the failure line, now there is another possibility that is from P’ to
Q’, it is the construction phase that is at the end of construction.
Now it can also fail during the construction that is at the end of construction when Q’ it goes and
touches the failure line. So, then we say that the failure happens during undrained excavation, so
it has not started dissipating, so it fails at the short term undrained condition. If R’ reaches the
failure line, now if this portion reaches the failure line then failure happens sometimes after the
excavation that is as the negative pore water pressure dissipates.
485
Now why there is negative pore water pressure? Because here it is a kind of excavation that is
happening and there is a release that is happening. The distance of Q’ or R’ from the failure line
is the safety margin or the factor of safety. Now the factor of safety of retaining wall supporting
excavation decreases with time. So, in the case of this particular retaining wall, which is
supporting the excavation, then the factor of safety decreases with time why because effective
stress path is moving towards the failure line with time.
Now let us see what is the case of retaining wall subjected to backfilling or filling? Due to staged
backfilling both the stress parameters increase. Now it is more like a compression, so here it is
the starting and the filling takes place, so this is the fill soil. Now PQ is the total stress path due
to increase in backfilling. So, now P, P’ and with a difference of initial pore water pressure u0,
you can see that it increases and it was positive slope, so PQ is the total stress path.
Now ESP is given by P’ Q’, and it will depend upon soil characteristics and stress history. So, P’
Q’ is the effective stress path, ui is the pore water pressure immediately after construction. Now
in this case since it is getting loaded, it is a more like a compressive load, so the initial excess
pore water pressure will be positive, so that is the difference. So, this will be positive and this
positive pore water pressure dissipates with time and u reduces.
486
Now here it is just the opposite than the case that we discussed before, u excess which is positive
when it dissipates then the ui starts reducing. So, ESP will move away from the failure line and it
moves in the direction of R’. So, you can see that it is departing away from the failure line and
this is similar to the case that we discussed for foundation imparting compressive load. So, ui
excess, so this much excess pore water of pressure gets dissipated and then it reaches a final
equilibrium pore water pressure which corresponds to this value.
So, that uf is the final equilibrium pore water pressure which is equal to u0, because we have
considered a submergence condition.
There is an increase in s’, s it is increasing, so final equilibrium is u pore water pressure is uf, ui
you can see that it is greater than uf because it is excess pore water pressure is positive. With
time u decrease, soil consolidate for ESP Q’ R’, so from Q’ to R’ definitely the soil will undergo
consolidation under this load of the fill soil. Wall will fail if Q’ touches the failure line.
So, definitely in this particular case where we have backfilling we know that the soil will fail in
when it is under undrained condition, that is immediately after the construction only under
undrained condition. With time as the pore water pressure dissipates it moves away, so it is not
going to fail in long term. So, distance of Q’ from the failure line represents the factor of safety,
with time R’ is moving away from the failure line, so factor of safety increases with time.
487
So, long term behavior of wall supporting excavation and filling these are different. So, wall
supporting excavation it becomes less safe in long term and vice versa for filling, so that is what
we have seen. In this case factor of safety increases, so safety is increasing which is reverse in
the case of excavation.
So, these are some of the additional cases which we will see in the field, there is an embankment
load which is acting on the soil. And there is a possibility of slip failure that can take place; it is a
typical case of compression. So, it is an axial compression, this is the initial stress σ’a0 and σ’r0
before the construction of embankment. And when the embankment is constructed, we have an
additional Δσ in the axial direction.
So, this is the case of axial compression and we have discussed how the total stress path effective
stress path would look like in the case of axial compression. So, these are cases which we have
already seen because we have seen if σr is constant σa increases, those type of cases we have
already discussed, we are just correlating these field cases to those cases. Now in the case of
excavation what is happening?
You can see that it the actual stress is getting released, so in this, this is the initial condition when
the axial stress is getting released. So, that results in axial extension, so this is another case which
488
we have already seen. So, I am not going into the details of stress path, because that will be then
repetition.
Now this is the retaining wall which undergoes an active condition and active failure. So, in this
particular case is active pressure which comes into play. And here we can see that because of the
excavation there is a release in the lateral stress that is happening. So, this is the initial condition
under it is own weight and because of the excavation there is a release that is happening in
radially. So, this is a case of lateral extension that is it is moving in this direction.
Now just the reverse of it that is a typical case of passive earth pressure that is when the retaining
wall is rotating this mass will undergo a passive condition. So, hence what is happening? There
is a more compressive condition which is getting added up, that is lateral stress is increasing, so
that is the case of lateral compression. So, these are some of the cases, field cases which we have
already completed. Now let us try to summarize what we have learnt in this particular lecture.
489
The stress path for undrained loading of foundation which is similar to U1 is discussed. Effective
stress path will depend on soil characteristics and stress history, which we have already seen in
couple of examples before. Foundation loading under undrained condition can be unsafe during
short term and it becomes safe in long term. We have very clearly seen this how the stress path
departs away from the failure line.
Now this is basically due to the positive pore water pressure that gets developed. And reduction
in u with time that is after construction and that results in more stability. But I would like to
underline here it is progressively densifying and hence becoming more stable in terms of stress
path. But what is happening to the deformation or the settlement response of the foundation? So,
this densification is at the expense of settlement, so we need to also look into the serviceability
part of the foundation.
For retaining wall, stress path depends on whether the process is excavation or filling the
response is totally different in both the cases. The response for excavation is similar to the stress
path of excavated slope. Factor of safety decreases with time, that is after construction for
excavation and it increases for filling. In stress path plotting for any case in the field determining
the slope of the stress path becomes important, now what does all these means?
490
We need to identify or we need to determine the correct slope of stress path. And for that what
we have done the slope of stress path for σa σr variation is discussed keeping one constant. So,
we have discussed the very critical cases, what are they? We first kept σr constant and σa
increased, and all other similar there are 4 cases, now accordingly what we have got?
We have got if this is the failure line, we have got a response something like this, so we have 4
stress path. Now in the field depending upon how σa and σr varies? The slope of the stress path
would change and that can change in any direction in this 360 degree. So, what we have
discussed is a critical case where one is constant the other one is changing, this we have done for
simplicity.
But it can so happen that σa and σr changes in a particular manner, and for that we will have
different slope of the stress path. And this is relevant for both drained as well as undrained stress
paths. So, other drained and undrained stress path variations in the field, it will depend on the
corresponding changes in σa and σr. So, depending upon this variation the slope of the stress path
would change.
It can be in this or it can be in this direction, so the actual magnitude of the changes would
determine the stress path slope. So, whenever we get a field problem what we need to understand
is, how the stress path is moving depending upon the relative changes in σa and σr. Now most of
the examples that the cases in the field that we have discussed, it is mostly from the reference
textbook of Atkinson.
So, you can refer to this textbook if you get access to it because most of these examples have
been taken from that and also from the book by Parry. So, you may refer to it for further reading.
So, that is all for this lecture, so we have completed the stress path for various field cases. Now
in the next lecture we will see some worked out examples related to stress path, thank you.
491
Lecture-42
Stress Path Problems
Welcome back all of you, in the last lecture we have finished this module 3. And we have
discussed various aspects of stress path. In this lecture we will work out some representative
problems related to stress path.
So, the first question is a soil sample is isotropically consolidated to σ’0, there is no cohesion and
angle of failure line is α, slope of effective stress path is β for over consolidated state of the soil
and that is in compression. Prove maximum shear stress at failure
So, this is a representative problem which helps us to interpret the question and draw the stress
path.
There are a few conditions which are given and the most importantly it is over consolidated,
alpha means it gives a clue that it should be in t- s, s’ plot, there is no cohesion. So, A’ is not
492
going to come into picture, so first it is isotropically consolidated that means we know where is
the starting point. And we need to first find out what is the slope of ESP, so effective stress path,
so it has to be in a more general manner, had it been total stress path we know like there are clear
specific leads 1 is to 1 is the slope.
But since it is effective stress path, we also need to take care of other aspects as well, now let us
try to solve this problem. So, it is clear that it is on t-s, s’ and the slope of the failure line is α, so
that is marked here. Next is the initial point, soil sample is isotropically consolidated to σ’0. So, it
gives a clear indication that it has to be on the isotropic line and which is the isotropic line? So,
this is the isotropic line.
So, the initial point that is p’ which corresponds to σ’0 will be here, now how does this point
become σ’0. Because we know that s’ = σ’0+ σ’0 /2, why is it so? We know that s’ = σ’0+ σ’r /2,
so since it is isotropic σ’0+ σ’0 /2 that will be, so s’0 that is the initial point s’0 = σ’0.
Now how did I draw the effective stress path with such an inclination? Now nothing is told only
thing is it is told that the effective stress path is at an inclination of β. Now this β can be in this
direction as well. So, how did I plan to draw ESP in this direction? ESP in this direction, now we
have to look back to the slope of ESP in t-s, s’ plot, that is 1 /1 - 2A. Now for over consolidated
state the A value at failure if that is to be considered as constant.
Let us take Af to be constant and we know that for heavily over consolidated soil Af will be
negative. So, if you substitute for A equal to a negative value whatever be the value, this whole
of the slope will become positive. So, we are shear that for OC state this ESP is going to be
positive and that is what I have drawn with a positive slope and that is equal to β. And what is of
maximum shear stress at failure? This is the maximum shear stress at failure.
Because where effective stress path touches the failure line, now this is tf. And now the problem
is very simple, the only aspect we need to keep in mind while solving stress path problem is that,
we should be able to plot it correctly. Once we get the geometry then it is very easy to solve the
493
problem. So, what is important is to determine the slope? The certain basic knowledge that
failure is always brought about / effective stress path.
So, these bare minimal conditions we should keep in mind and that will help to solve the
problem. Now the problem is simple, it is only geometry now. Let us consider this distance to be
b, so tan α considering this particular triangle tan α = tf divided / this is σ’0. So, this distance is
σ’0 and this distance is marked as b, so that is σ’0 + b, tanβ, this triangle, we have tf / b, so that is
what is written here.
Then I can replace b, so tan α = tf /( σ’0 + (tf /tanβ)). Now further rearranging we have
So, hence it is proved. This expression you can use it for solving further problems as well, so that
is all about the first problem.
494
Now let us move on. So, the next problem is as follows, there are 2 cylindrical samples A and B,
now both the samples were isotropically loaded / a stress of 200 kPa under drained condition. So,
there are 2 samples A and B, let us consider it to be a triaxial sample, both of them are
isotropically consolidated. So, isotropically consolidated, it is very clear that it is under drained
condition otherwise it will not consolidate.
And the initial isotropic consolidation pressure is given that is 200 kPa. Further, let us take the
case of sample A, for sample A radial stress was kept constant, radial stress means σr is kept
constant, and axial stress is increased to 440 kPa under undrained conditions. Now you
remember that it is a straightforward case of compression. So, radial stress was kept constant and
actual stress increased / 440 kPa.
And now we have to keep in mind that it is under undrained condition. For sample B axial stress
was kept constant and radial stress reduced 250 kPa under drained condition. So, this is also
compression because axial stress was kept constant and you are releasing or reducing the radial
stress. Now earlier radial stress was 200 it is reduced to 50 kPa under drained condition, now that
also we have to keep in mind.
Now we need to plot total stress path and effective stress path in q-p, p’ plot, for A and B
assuming soil to be linear, isotropic and elastic material. Calculate maximum excess pore water
pressure for sample A. So, it is a straightforward problem, we are not asked to find out any
failure conditions rather we just need to find in what manner the total stress path and effective
stress path would be in this particular problem.
So, let us say A and B is isotropically consolidated to 200 kPa, and before that I think we also
need to take the value of A param. That is the pore water pressure param also, so we will do that
or we will assume that in the due course. So, A and B is isotropically consolidated to 200 kPa, so
p’ is, we need to find out what is the value of p’ and q’, because that q because that is where we
have to plot it.
495
And Δ q = Δσ’a - Δσ’r, so this is known to us. So, the initial condition is 200 + 2 into 200 / 3 this
200 and Δ q is 0. Let us plot this initial condition, so we are starting from 0, we are consolidating
it isotropically, here you can see q is same. So, Δ q is 0 and Δ p’ keeps increasing to 200 kPa, so
this is p, p’ because it is isotropically consolidated, there is no pore water pressure existing in the
beginning. So, that is 200 and this initial point is p, p’. Here effective stress path is equal to total
stress path, and these things are very clear to all of us.
Now specifically for sample A, let us start it is an undrained condition. Undrained loading, we
know that it is at constant radial stress, so Δ σr = 0. So, interpreting the question is very
important, Δ a is then increased to 440 kPa, so Δ σa will be 440 - 200. Because earlier there is an
isotropic stress condition further which it is increased up to 440 kPa. So, do not misinterpret that
440 is in addition to 200, it is increased to the axial loading is increased to 440.
So, the total axial stress increment will be 440 - 200 which is 240 kPa, so these data is known.
Now for total stress path it is pretty straight forward Δ p = Δ σa + 2 Δ σr / 3, Δ σr is 0, that is
substituted Δ σa is 240 and hence Δ p = 80 kPa. Where you can go wrong in this question? If you
interpret that it is 440 is over and above there is a possibility of misinterpretation, in fact it is
loaded to the final value.
496
So, those terminologies and English usage you have to keep in mind Δ q = Δσa- Δσr which is 240
- 0 which is 240 kPa, Δ q upon Δ p is 240 / 80 which is equal to 3. Now this slope we already
know, in q-p, p’ we have already seen for drained condition that is total stress path equal to
effective stress path, the slope is equal to 3. Now in the undrained case for total stress path, this
slope will always remain same.
So, that is what we have just found out and that is equal to 3. So, this is P Q is the total stress
path at an inclination of 3 or a slope of 3. So, here this value will be 280 and 240, now the final
value will be because this is Δ p, so increment is 80. So, up to here it is 200 + 80, so that is 280
on x axis and y axis which is represented / Δ q which starts from 0. So, whatever it is 240 that
appear as 240, so this is the point q.
Now, we also now need to find out what is ESP? So, the moment for undrained case we need to
plot ESP, we also need to determine what is the variation in pore water pressure? So, Δ u is the
excess pore water pressure which is B * Δσr + AB (Δσa - Δσr), B = 1. And Δ u = 0 + (1 / 3) (240
– 0) which is equal to 80 kPa. Now what is the catch here? It is already given in the question;
assume the soil to be isotropic and elastic.
Now if Skempton's pore water pressure equation for triaxial condition is not clear to us, it is
difficult to assume the value of A. Here for elastic condition we know that the value of A = 1 / 3,
497
and that is what we have done here A has been substituted to 1 / 3, so Δ u we will get to be 80
kPa. So, this is the condition that we have used, so slope of effective stress path is equal to, this
also we know 3 / 1 - 3A.
Now A, the value is 1 / 3, so here it becomes 0 and hence it will be infinity. So, that means it
grows vertically upwards because it is Δ q upon Δ p. So, Δ p there is no change in p, so it has to
go vertically upwards, Δ q = Δ q’ = 240 which we have already seen in the previous slide. So,
maximum value for this particular case is 240 and so q’, I have just taken q’ just to make sure
that it is for effective stress path.
But we already know q’ and q are the same, so q’ = q0 + Δ q that is 0 + 240, so it will be Δ p’ = Δ
p - Δ u which is 80 - 80 which is equal to 0, so that is also now known. So, there is no change, so
p’ = p’0 + Δ p’ that is 200 is the initial value + 0 which is 200 kPa. And hence Δ q’ / Δ p’ =
infinity, and this information we already have here, is just for your understanding we have
worked it out.
So, here is the effective stress path, so this we have understood. Now this is Q’ 200 and 240, now
what is this distance? It is already there Δ u which is equal to 80, so 280 - 80 that gives 200, q
value is 240, as we have determined here. So, that is all about sample A for undrained loading.
498
Now sample B, so we know that it is a drained unloading, so here it is release of radial stress, so
it is unloading. So, drained unloading Δσa is 0, axial stress kept constant, σr decreased to 50 kPa.
So, there was an initial value of 200 from there it is decrease to 50 kPa. So, Δσr = 50 – 200, that
is - 150, all these sequence of loading we have clearly discussed while discussing the stress path.
So, you can cross refer to those for understanding this better. So, TSP, now TSP and ESP both
are same because it is drained unloading. So, Δ p = Δσa+ 2 Δσr / 3, Δσa given 0 + 2 into - 150 / 3,
it is -100 kPa, Δ q = Δσa- Δ 0 minus of minus it is +150. So, Δ q upon Δ p = 150 / - 100 which is
equal to -1.5, so it is very clear that the slope is negative.
p = p0 + Δ p, so p0 is 200 and Δ p = -100, substituting that we know that it is 100 kPa but in the
left ward direction. And q = q0 + Δ q q0, initial value is 0 + 150 Δ q is +150 that will give you
150. So, 150 in this direction 100 it has to go in this direction, so that is the slope it is negative
slope. So, TSP for B, so this is the P R is the required TSP for B and all other interpretation it
comes from this calculation.
Because it is minus, so we know that it is moving in this particular direction, and apart from that
the slope is also minus, so we know that it has to be in the left ward direction. Now should it be
in which direction that it should it be in this or in this, so that information you will get from the
actual values of P and Q. Here it is looking positive but please remember that it is going in the
reverse direction. Earlier value was 200; since it is negative it is decreasing. So, we know that in
which direction it will be moving, so that is 100 and 150.
499
So, ESP of this Δ u = 0 drained, so q’ = q0 + del is 150, p is 100 kPa and hence it is the same. So,
there is no difference for ESP and TSP, so since it is drained we do not have to explain it further.
So, that is all about the second question we will move on to the next one.
So, the next question is, an embankment is to be constructed on soft clay as shown. So, this is the
embankment the top width is 11 m, bottom width is given it is 17 m, the height is 2.75 m and the
bulk unit weight of the material is 20.6 KN/m3. And this is the soft clay below with the given
properties, there is a water table at 1 m depth, above the water table gamma or the bulk unit
weight is 12.1 KN/m3.
500
Below the water table the saturated unit weight is 12.7 KN/m3, initial water content is 140%
below, liquid limit is 160 and plasticity index is 105% and the lateral earth pressure at rest K0 is
0.6, that is given. A parameter before failure is 0.35, and A parameter that is Af at failure is 0.5,
so Skempton's A parameter is given, triaxial compression test in the lab you can add gave the
effective friction angle = 23 degrees and c’ = 0.
Determine whether the soil element at depth 5 m would fail or not? So, this is the reference soil
element which is at 5 m below the ground surface. Now we need to understand based on stress
path whether this particular element would fail or not. Consider the vertical stress increment at 5
m depth is 0.9 times at the ground surface and lateral stress increment is 0.33 times of vertical
stress increment. What is the importance of last statement?
So, there are 2 cases, first is the initial condition which is due to self weight, now under self
weight it is not going to fail. Now when you are placing the embankment load whether the
element would fail or not, that is the question. Now when you add the embankment there will be
some amount of stress that gets added up at the element A. Now if you studied the stress
distribution we know that there is a way / which we can calculate it.
Here we are adopting this particular information for finding out what will be the stress increment
at point A, that is consider it is at0.9 times, now if you consider what is the stress because of this
embankment at the ground surface then at point A it will be 0.9 times of what is the stress due to
embankment at ground surface. So, first we need to find out what is the stress increment due to
embankment at the ground surface.
It is 0.9 times will be the stress at the element A, and the lateral stress increment is 0.33 times of
vertical stress increment at A. So, those information are given, so let us see how to solve this
problem. So, initial condition of A, it is because of the sulfate, so σv0 = 12.1 * 1 + 4* 12.7, so
that is the total vertical stress which is equal to 63 kPa and that is the initial condition.
u0, if you consider a piezometer here it will rise up to 4 m, so 4 into 9.8 which is 39 kPa. So, σ’v0
= 63 - 39 which is 24 kPa, so σ’h0 = 0.6 because lateral earth pressure at rest is given to be 0.6, so
501
0.6 into 24, that will give us 14 kPa. So, σ h0 = 14 + 39, that is a total horizontal stress 14 + 39
which is equal to 53 kPa.
So, initial condition is known. So, t0 we have plotting this in t-s, s’ plot, so we can find out what
is t0? Which is σ v 0 - σ h0 / 2 all the values are given 63 - 53 / 2 which is 5 kPa, s0 the same + 63 +
53 / 2 which is 58 kPa. So, the initial condition is now known. So, s’0is s0 - u0 which is equal to
58 - 39 which is equal to 19 kPa, because u0 is 39. So, all these initial conditions are known now.
So, t-s, s’, kfc is the failure line and you can easily denote what is kfc because φ’is known. tan α =
sine φ’, φ’ is 23, so we get0.4, so sin 23 = 0.4 is the slope of the failure line without this, this
problem cannot be solved. Now the starting point is the point P, we know that it is t0 and s 0
which is 5 and 58. So, here it is somewhere 58 is s 0 and t0 is 5, so 58 and 5, we have s’ 0 which
is s 0 - u 0 that is p’ is around 19 kPa and t0 remains same.
So, at 5 height we have 19, so it is close to here 19, so p’ that is 19 5. So, you can see that the
initial point itself is very close to the failure line. So, you can make some judgment right at this
point itself, like there are not much of elements for failure. Like it is very close and factor of
safety is substantially low. So, factor of safety how do you be in the last lectures we have seen
that when it is close to the failure line the factor of safety is less. So, accordingly we can easily
make out the factor of safety or the safety margin is quite less in this particular case, this is 39.
502
Now after the construction of embankment what will happen? Whether it was going to fail or
not? So, the incremental stress is Δ σ v which is equal to 2.75 which is the height of the
embankment into 20.6 is the unit weight given for the embankment material which is 57 kPa at 5
m depth, the entire stress is not going to be delegated it is 0.9 times according to what is given in
the question.
Δ σ h = 0.33 into 51 which is equal to 17 kPa, this is given and this is as per what is given in the
question. Now Δ u what will be the change in pore water pressure because of embankment
construction which is B* Δ σ h + AB* (Δ σ v - Δ σ h). Substituting this we have all the
information with us, we have Δ u = 34 kPa. So, σ vf is 63 + 51, the incremental load that is 114
kPa, σ hf = 23 + 17 which is 70 kPa.
503
Now with this we can find out what will be the final condition of total stress that is tf = σ vf - σ hf /
2 which is 114 - 70 / 2, which is 22 kPa. So, here it is 22, I am drawing a dashed line here, what
is s f? s f = σ vf +σ hf / 2 114 + 70 / 2 which is 92 kPa, so somewhere here. So, I am drawn a
vertical dashed line s f - u 0, that is 92 - 39 which is equal to 53, so we already know that it is 39
which is 53 kPa.
So, u f = 34 + 39 because in the previous slide we have seen that Δ u because of embankment
construction is 34. So, initial is 39, so total pore water pressure at final is 73 kPa, why did I use
this particular s f - u 0? This is to just show that the total stress path here is translated to a total
stress path here that is TSP - u 0 stress path. So, t - u 0 stress path, we have total stress path TSP
and from the total stress path whatever is the initial porter pressure we minus that.
So, that we draw the total stress path at this particular point, so that is another way of
representing total stress path, t - u 0 stress path. That also we have discussed in the last lecture,
you may please refer back. So, s’ f = s f – u f that is 92 – 73, from here you need to minus the
final pore water pressure 73, that is 19 kPa. So, it is at the same point here s’ f is at this point, so
this is the total stress path P Q that is 92, 22.
So, this is 92 and here it is 22, that is due to the construction of embankment. Now when the
embankment is getting constructed the total stress path is progressing in this direction, and that is
504
92, 22, that is total stress path. Now this is t - u 0 stress path, it is parallel to stress path but it is
drawn at this initial point. And that is initial pore water pressure 39 is minus from this and this is
the t - u 0 stress path.
Now the effective stress path is the s’ f is 19, so it is at this point, and what is t f? It is 22. So, it
moves in this direction and touches here, why it is moving in the vertical direction? It is already
given in the question that the final pore water pressure at failure will be the A parameter
corresponding to this will be 0.5. And we know that the total effective stress path of t-s, s’ plot is
1 / 1 - 2A.
According to that, that is slope is 1 / 1 - 2A, so A f is given as 0.5. So, if you substitute A f = 0.5,
this will be infinity, so that is why it is moving vertically upwards, so that is what is shown here
A f = 0.5 and that gives infinite. So, it is in the vertical direction and this is 19, 22. Now we do
not need to discuss further to say whether it will fail or not, yes the stress state according to the
final construction of embankment is well above the failure line which is an impossible state.
So, the soil would have failed by the time it reached here, it cannot cross the failure line in any
circumstances. So, the embankment construction would fail the soil element which is at 5 m
depth. So, ESP indicates the failure of soil elements, so how beautifully the effective stress path
helps us to understand a lot of facts whether it will fail, at what time it will fail, and how it is
going to fail?
All these information we can derive / plotting stress path. So, this is a very good example which
clearly demonstrates for a field problem how we can interpret various situations based on stress
path plotting. So, now whatever we are asked to find out, it is already done, we have found out
what is the excess pore water pressure, we have found out whether the embankment construction
would fail the soil element or not, all these things are known.
Now we are curious to find out what is the possible strength available if the condition is to be
safe. There is a strength available that is t fs corresponding to s dash f, because this is a starting
point is equal to 19 kPa, that is s’f = 19, that is this particular point what is the available strength?
505
That is tfs, so this is what is shown as tfs. Now tfs is equal to we know that the inclination is 0.4,
0.4 into 19 this distance, that is 7.6 kPa is the strength available here.
So, maximum shear stress increment which is possible due to the construction of embankment if
the soil has to be safe is equal to 7.6 - 5 which is equal to 2.6 kPa. So, 2.6 kPa is the only
permitted increment, why it is 2.6? Because we know that Af is 0.5, and it is moving in the
vertical direction. Now if it is moving in the vertical direction the only possible increment
available with us is 2.6 kPa.
So, we can construct only that much of the embankment that will give an incremental stress of
2.6, beyond that it is going to fail. So, we have understood like where it is going to and how it is
going to fail. Now here already the initial condition is at 5 kPa above now let us say that if it is
isotropically consolidated. Then we would get an allowance of 7.6 kPa before failure that is this
point will come down.
So, total possible increment would be 7.6, it is only an added information which we can get from
the stress path plot. For example if the starting point is again towards left, we will have even less
allowance or if this point is slightly here then we have this much of strength available. So, that
much amount of embankment construction is possible. So, this particular plot help us for various
possibilities and understanding related to a given field condition.
This is what we have discussed related to embankment. Now this can be any problem, it can be a
foundation, it can be retaining walls like the one we have discussed in the last lecture, so that is
all. So, now we have almost completed module 3 and some representative problems we have
done.
506
The problems which are discussed in this lecture, these are very similar to the one reported in the
textbooks of M. Budhu, Holtz and others. So, please refer to these textbooks for more examples.
So, that is all for now in this lecture, the last lecture in module 3 is summarizing the whole of
module 3 what we have learned, so that is all for now, thank you.
507
Lecture-43
Summary of Module 3
So, welcome back all of you, we have now finished module 3 we have discussed different
aspects of stress path which I feel and hope that it is sufficient for going on to the next or the last
module on critical state soil mechanics. So, here in this lecture we will try to summarize what all
we have learned in module 3.
So, this module focus on the stress path and it is importance in soil mechanics which I hope you
would have appreciated by this time how importance stress path is. Stress path helps to
understand better the stress variations in the soil during loading and unloading and how the soil
mass approaches failure. It gives a very clear representation which we have try to demonstrate in
few of our lectures.
This understanding helps to design the most suitable strength testing protocol. Now if we do a
kind of analysis prior to the actual driving of the shear strength parameters, it also helps to
understand how different sequence of loading that happens in the field would give different
stress path. So, we can try to project or we can try to use the laboratory testing, if is possible, like
508
stress path controlled triaxial testing if it is possible we can always do that in the lab to get the
best representative shear strength parameter.
So, it helps us in planning a suitable strength testing protocol. It would also help to plan the
observation and instrumentation plan for monitoring the performance of the structure founded in
soil. So, let us say that we are interested in doing some sort of monitoring, so we the stress path
analysis in the beginning helps us to focus on what exactly and how and at what time we need to
monitor?
So, the instrumentation plan or sensors placement and it is operation can be planned with the
help of initial stress path analysis. Now stress path representation in three different formats we
have understood σa, σr, t-s, s’ and q-p, p’. Factors affecting total stress path and effective stress
path has been discussed. Importance of failure line in stress path for both compression and
extension we have understood.
And we have also defined the failure lines for different cases, different cases means the different
stress path formats. Stress path for isotropic and one dimensional consolidation we have
discussed. Stress path for unloading and re-loading and the associated elastic behaviour, because
we are talking about the over consolidated state when we actually say it is unloaded, so re-
loading also it follows the elastic behaviour.
Stress path for isotropic and anisotropic initial consolidation, so the influence of k, like
depending upon the initial k value how it is going to change and where it actually gets place the
initial the starting point. So, we have discussed for both k 0 = 1 and k 0 less than 1, k 0 greater
than 1.
509
Elastic stress path for undrained triaxial compression, we have already seen that. Stress path for
laboratory drained triaxial test for both compression and extension, we have discussed all the 4
cases with different permutation, combination. Influence of initial stress state that is in terms of k
value and what is the influence of cohesion on the stress path, we have seen that it gets extended
above, so the available strength increases.
Derivation of drained strength in t-s, s’ plot, stress path for laboratory undrained triaxial test for
both compression and extension, just similar to what we have discuss for drained test. But here A
parameter, Δ u determination and the uniqueness. So, ESP in undrained triaxial test is unique for
identical soil specimen and initial state. Unique ESP irrespective of undrained compression or
extension.
Slope of undrained ESP and the role of Skempton’s pore water pressure parameter A, we have
seen that. Importance of effective stress path in defining the failure criteria, so we have to keep
in mind, it is not the total stress path; it is the effective stress path for undrained case that
determines failure. So, where the effective stress path touches the failure line that is our point of
failure.
Derivation of undrained shear strength in t-s, s’ plot, expression for undrained shear strength for
initial isotropic condition that is soil sample from the field. That is extension of this particular
510
condition where we know that due to sampling there will be initial pore water pressure of
negative initial pore water pressure. So, since it is pore water pressure it becomes isotropic
condition, so it gets shifted to the isotropic line. So, it is a just modification of what is discussed
in this.
So, undrained stress path as a function of stress history, lab effective stress path is identified for
different field conditions that was the last part of our lecture where we discussed about how the
stress path can be identified for various field conditions. So, total stress path and effective stress
path discussed for various field situations like infinite slope, finite slope excavation, shallow
foundation, retaining wall supporting either excavation or filling. So, all these cases we have
discussed.
Now to end this module 3, the major reference books that I have used for this module is Holtz
and others, J. Atkinson, R.H.G. Parry and M. Budhu. All these reference textbooks are there in
the reference list, if you want to know more or if you want to enhance your reading you are free
to do that and these are the textbooks which I have referred a lot for developing this module. I
have referred to others as well but for basic textbooks you can refer to these.
511
So, this is all for now, so we have finished our module 3 and we are fully prepared now to move
onto the next module and the last module where we will discuss about critical state soil
mechanics. So, thanks a lot for now.
512
Lecture-44
Introduction-Critical State Soil Mechanics
Welcome back all of you, in the last lecture, we have completed module 3 on stress path. We
have discussed various aspects of stress path which will become handy for module 4, which is
the final module, which will be starting today. And the module is all about critical state soil
mechanics. Now, the terminology critical state is not like you have already gone through, it is not
new. So, are we going to study altogether a different concept in this module 4?
Not necessarily, what we will be doing is, we will be using the information that we have studied
in our last few lectures basically related to module 2 and module 3, we will be placing it at
appropriate locations and that is all about module 4. Of course, there will be some new
terminologies that come into picture for understanding this framework better. So, with this, let us
start the module 4 on critical state soil mechanics.
So, this concept, please remember this is a concept and which is put forth by Schofield and
Wroth in 1968 at Cambridge University. It is about critical state soil mechanics, short form CSSM.
Now, why this concept was evolved? The concept essentially integrates whatever we have
513
learned that is the essential or the essence of soil mechanical behaviour, how do we define it is
Mohr Coulomb failure criterion, the effect of stress history, the concept of soil yielding, drained
and undrained behaviour.
So, it integrates, it tried to integrate these aspects together. This particular framework or the
concept, it laid the foundation for sophisticated soil constitutive models used for simulating
complex geotechnical problems. This we have already discussed a few of these models, I mean,
we did not discuss into the depth, but conceptual understanding of these models we have done in
our first module.
Now, there are a lot of models related to soil, which got developed to essentially cater to specific
requirements or specific complex geotechnical conditions. In fact, we will not be discussing
those in this; dedicated cause on geomechanics would help you to understand those, starting from
the linear model to very complicated nonlinear models. You can very well go through the books
by Potts, which is there already in the reference Haltz and Kovacs and many other books which
discuss about these models.
So, critical state soil mechanics offers some foundation for understanding these models better.
Some of these constitutive models include modified Cam Clay, cap models, nested models. This
concept of CSSM was originally developed for saturated reconstituted clays in which clays were
remolded and consolidated to a normally consolidated state. Essentially it was developed to
understand triaxial behaviour of clays essentially remolded ones.
So, reconsolidation that is going to an NC state can be followed by over consolidation to a

specified OCR. How do we generate over consolidated state by unloading it? CSSM offers a
relationship between two 2-dimensional variations. In fact, there are 2 different variations which
CSSM accounts and further it integrates these 2 different variations, what are these two 2D
variations? The first one is void ratio versus effective stress.
Which again is something which we have already learned during consolidation, not in this
particular course but then we have already have this basics from our undergraduate soil
514
mechanics. A void ratio versus effective stress, it represents the consolidation condition. The
second variation is shear stress versus effective stress, so effective stress is common here. So,
this critical state soil mechanics combines these two 2D spaces to a single 3D space that defines
the behaviour of clay of specific stress history during shearing.
So, whatever be, so there will be a particular stress history attached to that particular soil. So,
critical state framework it helps to understand the soil behaviour during shearing in a three
dimensional space which is basically obtained by integrating the two 2D variations. Now, for
this, it is considered the volume change for drained test and pore water pressure for undrained
interest. So, everything is accounted into this model be it the influence of stress history, be it the
influence of drainage conditions all into one. And we try to understand the soil behaviour in a
unique manner. So, that is the whole purpose of this development.
So, in essence CSSM integrates consolidation and shear strength behaviour. According to this
concept, all soils will fail on a unique failure surface which is defined in q, p’ e or q, p’ v space.
Now, q is deviatoric stress which represents the shear stress, p’ is the volumetric stress which
represents the normal or effective stress condition and e the void ratio or v the specific volume
that indicates the kind of pore structure or what is the kind of deformation behaviour.
515
So, the change in e helps to understand what is the kind of volume change that is happening
within the soil. So, what we are trying to do, initially we talked in the stress part if you remember
we have talked only in terms of two stress parameters. In fact, all our failure criterion is basically
based on stress parameters only τ versus σ in Mohr Coulomb. We do not discuss about the
volume change it has undergone during any kind of shearing.
So, critical state framework helps to bring in this new element. Now, in some textbooks you will
find the volume change is accounted in terms of e and in some others it will be in terms of
specific volume v, whatever be the concept remains the same, there will be some change in the
parameters that helps to define the critical state framework that is all. So, in fact, we will try to
discuss both side by side even though the prominent one is in terms of q, p’ and v.
It provides a relationship between shear strength, principal stresses and void ratio under drained
and undrained condition, as is visible in this q, p’ e or q, p’ v. CSSM incorporates volume change
behaviour in failure criterion, which was not considered in Mohr Coulomb failure criterion
which is the most prominent model for soils, we do not talk about e or v at failure. So, that was
purely based on maximum stress obliquity, I hope you understand this term σ’1 / σ’3 at failure.
CSSM helps to estimate the mechanical behaviour of soil subjected to loading during and after
construction. Later, you will see that you can predict based on this framework, what will be the
kind of behaviour in drained, undrained or with respect to stress history. It essentially depends on
basic soil properties and it is equally applicable for cohesive and cohesionless soil. If you
remember in the earlier lectures, the concept of cohesive and cohesionless was treated
differently.
When we discussed shear strength in module 2, we discussed separately for sands and clays,
dense sand, loose sand against normally consolidated or over consolidated clays. Now this
framework it does not actually need to see into the nitty gritties of these, we can all of them have
been encapsulated into one particular model. This idealized concept, please remember this is an
idealized concept.
516
Whenever we learn certain things related to mechanical behaviour of soils, somehow we tend to
expect that soils should always behave the way we have studied. This may not happen and that is
where we need to expand our understanding in the sense that whatever quantifications we are
studying from various textbooks, these are for engineering applications for us to design, for us to
understand the behaviour of soils.
So, all of them involves some sorts of assumption and idealization, here in the case of critical
state soil mechanics also, this is an idealized concept. So, if soils behave in a more ideal manner,
this is how it will behave, but there will be always departure from this concept. Now that we also
need to keep in mind, so this is a well idealized concept which is time tested and we understand
that in general soil behaviour can be explained using this.
So, this idealized concept may not fully be applicable to real soil behaviour, but assumed to be
applicable. Because this assumption is needed for us to predict the behaviour better the data
when we monitor the behaviour, sometimes it may be different, but prior to any testing program,
if we want to study the various scenarios this framework helps us to predict the behaviour better.
So, please do not take into account that, this is the final. So, soil is highly complex material and
in nature this material behaves the in a most uncertain manner. So, every aspect may not go as
per our technical understanding, it gives necessary insights related to soil behaviour. So, that is
what is more important, for an engineer when he is planning for any sort of analysis or design
certain insights related to soil behaviour is important and for that we need a framework.
And that framework is offered by CSSM. For all practical purpose, critical state can be considered
as a fundamental soil property dependent on the effective stress and not on the initial conditions
like density, stress history, drainage conditions. So, this is one of the most powerful aspect of
critical state soil mechanics. Earlier when we discussed Mohr Coulomb failure envelope, we
discussed about friction angle and cohesion, but we could not associate it to be a fundamental
soil property, because depending on the conditions these may change.
517
Now, here in this framework, it offers some solution partial solution to this problem by stating
that this is more or less a fundamental property of soil which is independent of the initial
conditions such as initial density, the stress history and what drainage condition does it account
for. Finally, the state of the soil reaches a particular state called critical state, so that can be
considered as fundamental behaviour.
So, when the soil is sheared, it tends towards an ultimate condition called critical state which is a
concept which we already know. So, that is what I told like we know these concepts only thing is
we will be putting these in the right shelf, which will be easy for us to understand. Wherein the
plastic shearing where we call it as shear strain would continue without changes in volume or
effective stress.
So, any soil which has sheared it reaches a final state what is known as critical state, wherein the
shear stress would continue but at constant effective stress and constant volume. So, the final
state has constant volume, so shearing happens or the plastic shearing happens at constant
volume when the soil reaches this critical state. This state of perfect plasticity is known as the
critical state.
The soil flows as a frictional fluid and the flow is considered to be turbulent, where the soil
continues to undergo shear strain without any change in shear stress, effective normal stress and
518
void ratio. Please remember, we are just repeating the same concept again and again, so it does
not matter that we are repeating, what is more important is that concept goes into our head very
clearly. So, that is why even though it is getting repeated, every step we are adding some new
information.
So, it is like this, at this particular critical state soil behaves more like a frictional fluid. Now, it is
very difficult to understand like how it looks like a frictional fluid, why it is called frictional?
Because there are solids which are involved, wherein the moment is happening and the friction is
getting mobilized So, possibly that is the reason why the terminology frictional fluid, and at the
critical state it is assumed that the flow is more like a turbulent manner.
So, it is not in an orderly manner, it is in a turbulent manner where the soil continues to undergo
shear strain at constant shear stress, effective normal stress and void ratio. So, it is more than a
distortion under constant state. All soils attain a well defined critical state that is a turbulent
condition irrespective of it is initial state. Again this is a reputation but still it is important to
repeat these statements.
At critical state, the effective stress, shear stress and volumetric strain, that is ε v with respect to
shear strain is constant. That means, ∂ p’ / ∂ ε s is equal to this is p’ is the effective stress or the
volumetric stress equal to ∂ q / ∂ ε s, where q represents the shear stress or the deviatoric stress, ∂
ε v upon ∂ ε s that is volumetric strain which is equal to ∂ v upon ∂ ε s where v is the volumetric
component of the soil that is specific volume and that is equal to s.
So, all of these with respect to differentiated with respect to shear strain ε s = 0, and where p’ just
for completeness I am adding, we know this, we have studied this σ’1 + σ’2 + σ’3 / 3. And triaxial
condition it is σ’1 + 2 σ’3 / 3. q, the general expression for deviatoric stress is this that is equal to
σ 1 - σ 3 for triaxial condition.
And specific volume v = 1 + e, so it is the volume of solids that contains unit volume of solids,.
So, it is a specific volume is defined as the volume of soil containing unit volume of solids. So, 1
+ c, if you remember the unit solid volume model, 1 + c is the total volume, in which the soil
519
volume is considered to be 1. So, that is why it is 1 + e is the total volume of soil that constitutes
unit solid volume that is 1.
So, any soil undergoing shear should finally attain a constant state, why because if not, it would
continue to compress that is strengthen or dilate that is weakening, which is not possible. This we
have seen these two actions are common whether it is cohesionless or cohesive when the soil is
getting sheared, it will either undergo compression or it will undergo dilation. Now can this be a
continuous process? No, it has to stop somewhere and that is where the critical state offers that
bound at which it can stop.
So, this is taken care of by defining critical state concept where soil reaches a constant state,
which is neither compressing nor dilating, it reaches a constant volume state. So, critical state in
general is attained at shear strain greater than 10%. Soil gets reconstituted by shear strain; due to
this reconstitution soil loses the track of initial conditions and achieves a unique critical state. So,
that is what is more important, why it is not dependent on?
The concept says that the reconstitution that happens during shearing enables soil to lose it is
memory from where it has come. So, it is more like it is forgetting it is initial state of conditions
what is the stress history to which it was subjected to, finally it reaches to this surface or line
520
which is called critical state. Since it is independent of initial state, it is reasonable to consider
critical state as a fundamental soil behaviour.
All effective stress paths finally reach a line in a 2D representation or a surface in 3D

representation, which is unique for that effective volumetric stress and soil type. So, what is
more important is soil type and the p’ that we consider. So, what is the p’ that we are starting
from that is important, initial state means we tend to focus on what is the influence of stress
history or what is the influence from the drainage condition or what is the influence of density.
Here p’ is something to do with the confinement and we know that confinement do influence the
shearing behaviour. So, here, it is basically dependent on what confinement or the initial state p’
it starts and also what is the soil type. It may be possible for clay soils to attain higher shear
displacement beyond critical state. Now this is one concept that can induce some sort of what
you say doubt or confusion.
When we discussed our earlier lectures in shear strength, we also discussed about something
known as residual strength. Now to alleviate those confusion, we are discussing this particular
point in fact, it is not really relevant in this module. But still we would like to discuss just
because to avoid that confusion. Now there are certain cases where soil can undergo very large
displacement, but we need to understand that the soil has already been in a state of failure.
Still it has not completely failed, so there will be some sort of shear strength that is residual in
that soil, that is what we discussed about residual strength as we discussed in module 2. Now this
strength is when it is surpasses the critical state condition and it is still able to sustain without
failure. So, there the constant shear strength condition is achieved at residual state, but critical
state is a common terminology even if it surpasses.
You can consider the failure to have happened at critical state, but when you want to actually
discuss slip lines of slope or when we consider the slip circle failure in the case of slope, there
will be some amount of movement that happens. That means it is the soil is subjected to very
521
large displacement before failure. To account for a very large displacement, the concept of
residual strength is also there.
So, that is the specific zones of intense shearing is formed, and shear stress drops to a
considerable low value, and that is what is known as residual strength. For materials like sands,
we can notice that the critical state and the residual state are the same, for them there is no that
such excessive movement that is possible. It is only valid for some certain kind of clays under
certain situations, in that case the residual strength get invoked.
So, there is some difference between the critical state as well as the residual strength that we
discussed. So, just to make to understand this concept better, let us try to understand this using a
figure.
So, this is what we have already studied, shear stress, τ peak and then reaching a constant critical
state, so this is equally valid for sand and clay. Now at this particular critical state, it is conceived
there will be at kind of turbulent flow. And this critical state in general is attained at shear strain
greater than 10% in general, so this what is marked here. For sands critical state and residual
state are the same at this point. For soils that can go past the critical state can achieve the residual
state, so that is what is shown by this blue line.
522
So, this is applicable only for clays where we call this as residual state at very large
displacement. And here the flow is considered to be laminar because things have subsided there
more or less, and particles or the clay particles are oriented in a more orderly manner. And
hence, there will be the laminar kind of movement at residual state. Remember, whatever we are
telling these are all concepts. So, now let us go into the first step of understanding critical state.
That is relationship among shear stress, normal stress and void ratio. So, we are considering all
the three elements which are needed for critical state. So, first let us understand in terms of τ and
σ’, this is very well known to us is τ was a σ’ and the failure criterion in terms of critical state
friction angle, nothing to explain here. So, this is the critical state line, earlier we called it as
failure line.
Here, now specifically we can call it as CSL critical state line. Then the other responses e versus
σ’, you can see that σ’ which is the effective stress or the effective normal stress remains
common for both the figures, so e represents the volume change. So, e versus σ’, we can plot a
line where in the locus denotes all the failure state or the e and σ’ corresponding to the critical
state.
So, that line is critical state line in e versus σ’, when you start loading, we can find a particular
point where the soil would fail, and that particular point what is the e and σ that is the ecs and σ’
523
cs corresponding to critical state. So, the locus of all that point is what is denoted by CSL in e σ’.
So, for example this is σ’ cs which is relevant to both the plots where on this particular plot that is
e versus σ’.
We have e cs and here we have τcs, ecs, σ’cs these three are connected. So, if we plot e versus σ’ on
e log σ’ this is on base 10. We have critical state line which can be represented as a linear
variation. And we can also represent normally consolidated line wherein it represents the e and
σ’ variation of normally consolidated state. So, NCL is known as normal consolidation line.
So, we are now introducing certain important terminologies which are important for this
framework, one is critical state line, the other one is normal consolidation line. Both CSL and
NCL are parallel and with same slope of compression index Cc. Because now this is nothing but
e versus log σ’ which we have seen during consolidation. And we know that the slope of this is
going to be compression index Cc.
Now, when we discuss the next one that is q, p’ and v then the slope will change. So, that is the
only difference but the concept remains the same. So, slope is Cc for both, so for both CSL and
NCL, the slope will be Cc. And this normal consolidation line forms the right side boundary, why
right side boundary? There is no soil state which is possible beyond this line, why? For any
virgin soil that you consider in the field, it will undergo a consolidation along the normally
consolidated line.
We cannot generate a state beyond this line, so it is an impossible state which we will see later.
So for now we will understand that normal consolidation line forms the right side boundary.
Whereas on the left side of NCL will be over consolidated state, it can be lightly or heavily over
consolidate, why because all over consolidated state is created by unloading from the normally
consolidated. So, unloading means it will be on the left side of normally consolidated line.
And critical state line falls in between, and how do we interpret this? This we will understand as
we move further. So, lightly over consolidated will be somewhere around this, we will discuss
more about it later. At critical state there is a unique relationship among τ, σ’ and e, that is at this
524
particular line or this particular line there is a unique relationship in among all the three
parameters that is τ, σ’ and e.
For that, let us consider an initial point irrespective of any; let us start with a very low stress
condition, which is more or less corresponding to let us say zero condition because log of 1 is 0.
So, instead we take σ’ = 1 kPa, we have some reference point on critical state line as well as on
NCL line. Let us call the void ratio at 1 kPa, σ’ = 1 kPa on the critical state line to be eγ and void
ratio on NCL to be e0.
If that is the condition, then we can always write, τ failure which is same as τ critical state is
equal to σ’ cs tan φ’ cs from this particular figure. And this is very familiar to us, which is very
similar to what we call it as Mohr Coulomb failure envelope. And this is an new expression
which takes into account the volume change aspect of soil. And where we can write e critical
state, we are referring to void ratio along this critical state line, which is equal to which is a
starting point at 1 kPa, it is eγ.
So, eγ - Cc log σ’cs that is σ’ along critical state line. So, these two are very important
expressions, where e cs σ’cs corresponds to critical state, eγ we have explained that is e at σ’ = 1
kPa on CSL, e 0 is e at σ’ = 1 kPa on NCL. So, since critical state is independent of initial state,
φ’cs, eγ and Cc can be considered as a fundamental soil parameters, this is what we need to
understand.
So, if we are dealing critical state framework in terms of τ, σ and e, then we need to understand
that φ’ cs, eγ and Cc are considered as fundamental soil parameters because these parameters are
irrespective of the initial condition. Now we have discussed in terms of τ, σ’ and e, but in most of
the cases you will see that critical state framework is discussed in terms of q, p’ and specific
volume.
525
So, let us try to understand in that. Consider isotropic or normal consolidation and unloading. So,
that is represented by p’ and εv. So, initial state at O, so let us consider it O, isotropically
consolidated to A or O to A represents nonlinear bulk stiffness behaviour. So, O A is the
isotropical consolidation, then from A it is unloaded to B, as you can see it is unloaded to B. So,
this O and B are at the same point, so it is at the same level of same p’ level.
Then from B it is reloaded to C, where A and C is almost same and then further loaded to D.
Points A or C, this particular point can be considered as the yield point for this particular
behaviour. This is the point where it has yielded, why because this length C B is there, now this
is the unloading reloading path, so this represents the yield point. ABC will be relatively stiffer
because it is on the unloading path of the over consolidation part.
So, instantaneous bulk modulus at any point can be given as K’, there is the bulk modulus is
equal to dp’ / d εv, that is the slope of this particular curve. And which you can see that it is keep
on changing with the stress level. And hence it cannot be considered to be constant, it is not a
unique constant, it changes with different stress level.
526
So, let us further understand volume change behaviour in terms of v and p’, again isotropic
normal consolidation, which is plotted in terms of v and p’. So, here this is the reputation of the
figure from the previous slide, if this is translated to v, p’, now here this p’ ε v we are plotting in
terms of v p’ which has the normal way in which these parameters are plotted for critical state
framework.
So, let us start with the point O, it undergoes normal consolidation from O to A, then unloading-
reloading that is to B. And then to C, reloaded to C and then again loaded to B, whatever we
have discussed here, we have redrawn in v p’ plot. Let this B p’ 0, this is the yield point or the
yield stress p’ y which we call it as pre consolidation pressure, replotted in terms of v ln p’.
Now this is in terms of v and p’, what we want is a more general representation in terms of v ln
p’, whenever we use log or ln it is to account for a wide variation. So, here also v versus ln p’, it
is replotted and when it is replotted we know that it will become a straight line. So, both
unloading-reloading is represented by a single straight line, average straight line A, B, C and
further D, so this is what has been replotted.
So, this is normally consolidation line, and here you can see when it is represented in terms of v
ln p’, the slope is considered to be λ, what was our slope? Initially when we considered in terms
of the other parameter in terms of void ratio it is e versus log σ’, and that this slope is Cc, here it
527
is slope λ. And the slope of unloading-reloading line or the over consolidation portion is κ is in
terms of κ, this K represents Kappa.
So, and p’ y is the yield point p’0 is the point the stress level at O or B. Again here also we form a
reference point, that is p’ = 1 kPa, the way we have done earlier. And if you extend these lines
you will get what is the v corresponding to 1 kPa. And for normal consolidation line this is
represented as N0 and for over consolidation point, that is this slope κ, the initial point is
considered to be vκ.
Linear representation is a fair approximation for soils for wide range of loading. Unloading-
reloading is approximated as linear, line OACD is NCL, where you can represent this line NCL
by v = N0 - λ ln p’, same thing we have done earlier. Line ABC is swelling-reloading line, for
which v = vκ - κ ln p’, that is to represent this particular line. Line ABC that is A, B, C meets
NCL at C that is the yield point.
Now parameters λ, κ and N0 are soil parameters and depends only on soil type. So, these
parameters N 0, λ and κ this can be considered as a fundamental soil parameter. Now soil can be
unloaded from any point, it can be unloaded from here; accordingly the yield stress would
change. So, vk changes with the unloading point. Now since it has been unloaded from point A,
vk is this, but if it is unloaded from here vk would change.
528
So, vk cannot be considered as a fundamental parameter, even p’ y will depend on the unloading
point. So, if it is unloaded from here, this becomes the yield point or the yield stress. So,
equations can be used to calculate v corresponding to stress history, by knowing the previous
equations we can calculate what is the specific volume corresponding to different stress history
whether it is here or on this.
v = N0 - λ ln p’, whatever we have written earlier, if you differentiate this with respect to p’ we
will get dv / dp’ = -λ / p’, if you differentiate ln p’ it is 1 / p’. Now if you write -dv / v, what is -
dv / v? This is nothing but volumetric strain which is equal to λ / vp’ dp’. We are just adding v
here and bringing dp’ on the side. Now let - dv / v = d εv which is the volumetric strain.
And K’, we know that we have already written the bulk stiffness K’ = dp’ / d εv. So, dp’ / d, so
this is replaced / d εv, so dp’ / d εv will be K’ and this will go to the side. So, this will be v p’ / λ.
So, hence K’ this is the bulk stiffness can be written as vp’ / λ. Now this is the scenario of bulk
modulus on normal consolidation line because it is λ.
But for swelling-reloading line K’ = vp’ / vκ, that is this particular line. So, you can see that K’ is
not a constant parameter rather it gets changed depending upon whether it is λ or whether it is vκ
or with different position of v and p’.
529
Now let us discuss about we have already discussed about the right most boundary that is the
normally consolidated line we will try to add a bit more on this, so bounds in v ln p’ plot. NCL is
the right most boundary for all compression, so any compression that you do it gets terminated
on a normal consolidation line, this is the right boundary. Soil state is impossible top and right of
NCL.
So, any point soil related point on this part is not possible, so that is why it is marked as
impossible state. NCL is also yield line, since this is the maximum pressure soil is subjected to,
so it is otherwise also we understand because this is the maximum point to which the soil is
subjected to. Left of NCL is unloaded and hence over consolidation, so here this particular
portion represents the over consolidated state.
Where OCR is defined by p’ y / p’ B, where this is the yield pressure or the preconsolidation
pressure p’ y / p’ B. If this is the point of consideration, if it is this then appropriate present stress
has to be considered. So, here OCR is for B, point B is p’ y / p’ B.
530
Now let us try to understand another terminology which is known as yield stress ratio and
understand how yield stress ratio will be a better parameter than over consolidation ratio. For that
let us consider 2 points P and Q with same OCR. So, here it is represented by a normally
consolidation line and we have 2 unloading lines, one at A and the other one at B. Now consider
point P and consider point Q, now if you refer to P and Q they have the same OCR.
That is p’ y A / p’ p = p’ y B / p’ Q. So, the OCR is same for the points P and Q, so that is what is
written here. That is in terms of ln OCR we can rewrite this expression, line through P and Q for
which OCR is same is parallel to NCL. So, if you draw a line through P and Q that will be
parallel to NCL because these are of the same OCR, so which is shown here, this is the line.
Points A and Q has same p’ that is A and Q they are vertically below, so hence it has got the
same p’ but with different stiffness λ and κ, that is what an important aspect even for the same
initial pressure, you can see that if you consider point A, it is falling on NCL and hence the
stiffness of A can be considered as λ whereas Q is on the loading-unloading line and hence the
stiffness is κ.
Points B and Q has nearly same specific volume, why because it is on the unloading line, so the
amount of regain of v will be marginal. So, B and Q is practically are nearly same specific
volume but they can be considered of different stiffness, why? B is on NCL, so it will have a
531
stiffness of λ whereas Q has a stiffness of κ. Now points A and B on NCL will have different
stiffness corresponding to loading or unloading.
Now let us say that A and B are on NCL there is no doubt on that, if you consider a normal
consolidation then A and B refers to a stiffness of λ. But it is already on a loading-unloading line,
so at this point it can also come from the unloading line as well. So, unloading-reloading line as
well, if that is a case A can also assume a value of κ. So, these points are very important as we
move further when we try interpret the results, these informations are quite handy.
So, please keep this in your mind. Soil stiffness is not only dependent on v and p’, I mean to say
bulk stiffness but it also depends on λ and κ depending on whether it is on loading or unloading
line.
Now, let us come to yield stress ratio. Point P can move to Q by reloading, yielding at A,
normally consolidating to B and unloading to Q. I mean to say if point P has to reach Q, what is
the normal route it follows? It has to reload to A, so that is what is written here and further it is
again yielding at A and then it moves towards B. And on unloading it reaches the condition of Q,
but there is another possibility of P moving to Q and that is by creep.
532
Now here it is not actually creep because p’ is changing but if we consider a point directly below
then it represents creep. So, for understanding that better let us take the reference point via A and
Q. Now whether from A to Q it can move from A to B and B to Q, now if it moves directly that
is what is known as creep that is at constant stress condition. Consider a points A and Q, A and Q
which are vertically below where A is directly moving to Q and this is possible when there is
creep behaviour.
So, p’ yA = p’ Q that is already marked, when A is moving directly to Q, that is A is moving

directly to Q, we have OCR = 1 because stress has not changed. When A is moving to Q, OCR
remains 1 for both A as well as Q for both the points because p’ did not change, why? Because
this is the yield stress we are not accounting for the yield stress at this point even though it is on
the unloading line of B Q.
But here with reference to A to Q directly we will take OCR = 1, therefore OCR does not
represent the state of the soil adequately for creep, because it is ignorant of this particular yield
stress. State of soil can be better described by yield stress ratio which gives more clarity to the
soil state, where yield stress ratio is always defined in terms of yield stress. That is p’ y upon the
present stress that is p’0.
So, here B is the yield stress for point Q, so p’0 is the current stress state, p’ y yield stress which
represents the stress at the intersection of swelling line through Q with NCL. So, this is the
required p’. So, soil state of creep can be better defined by YSR than OCR, for example in this
particular case both A and Q had an OCR of 1. If we are not bothered about p’ y but if you are
defining in terms of yield stress ratio which is always with respect to p’ y.
For example if you are referring to point Q which is below A, you are referring to a swelling-
reloading line with a slope of κ which meets NCL at B. So, for this the reference yield stress will
be p’ y. so, p’ yB / p’ Q will be the yield stress ratio at Q which is always greater than p’ yA.
Because the yield stress at A is p’ yA / p’ A which is equal to 1 which is a normally consolidated

condition.
533
But if YSR at A will be 1 and YSR at Q will be more than 1 because p’ yB has been accounted.
So, here the yield stress ratio helps us to interpret this condition in a more better manner and that
is more relevant to conditions like creep. So, that is all for today’s lecture like we have
introduced critical state soil mechanics, a few more aspect also need to come in the introduction
part which will see in the next lecture, that is all for now, thank you.
534
Lecture-45
Introduction Critical State Soil Mechanics
Welcome back all of you. In the last lecture, we have started fourth module on critical state
soil mechanics, we were dealing with the introduction of critical state soil mechanics, we will
continue with that lecture in today's lecture as well. In the last lecture, we have understood a
few aspects of critical state which we already know initially, but we are trying to understand
in the context of a framework.
So, we have not actually discussed the entire aspect of critical state line. So, we will focus
more on that in today's lecture and see where it position itself with respect to the normal
consolidation line or the swelling reloading line. So, today's lecture will focus on critical
state, normal consolidation, heavily over consolidated state and likely over consolidated state,
why this is important is we need to understand what is the starting point of a given soil or
rather I would say state of the soil with respect to critical state.
According to critical state framework, we know that the failure would happen when the soil
state reaches a critical state that is how we define it. Now, where the soil position is initially
with respect to critical state will help us appreciate or understand better in what manner or in
what amount of what is the factor of safety that is available before failure.
535
We have already said that normally consolidation line is the right most boundary, it can be
normally consolidated or you can say it as isotopic consolidated, whatever be it remains the
same only thing is for tracks shell we are specifically interested in isotropic consolidation. So,
for all practical purpose our normal consolidation line we can treat same as that of isotropic
consolidation line.
So, in v ln p’, we can draw NCL, we have already seen that and that is the right most
boundary and this NCL has a slope of λ. Now, the space which is towards the left is always
the likely over consolidated or heavily over consolidated depending upon what amount of
unloading the soil has undergone with respect to its past maximum pressure, the past
maximum pressure lies on normal consolidation line or isotropic consolidation line from
where it is unloaded which creates lightly over consolidated state or heavily over
consolidated state depending upon how far the state of the soil is with respect to NCL.
So, this is the left most part is LOC and HOC. Even though we know all these facts, it is very
important that we visualize this in the critical state framework. Consider critical
overconsolidation ratio line which separates lightly over consolidated and heavily over
consolidated. Now, we have already seen in module 2, that the behaviour of heavily over
consolidated is fairly different from that of the normally consolidated response.
Understanding this response is very important to understand critical state framework as well.
So, first we need to understand what separates heavily over consolidated from normally
consolidated. Now there is another category which is likely over consolidated, which
536
resembles its behaviour more towards the normal consolidation, means it may not exhibit
peak. So, where should we draw a line, so that it separates the normally consolidated or
maybe lightly over consolidated from that of heavily over consolidated.
And that is given by critical overconsolidation ratio line. So, overconsolidation ratio helps us
to understand how much or what is the degree of over consolidation the soil is subjected to.
Now, we let us understand it to be a line that is critical overconsolidation ratio COCRL to be
a line which is parallel to NCL, because then the ratio will be the same. Now, this line also
has an inclination of or the slope of λ.
Now, the portion now, what it means is this particular line is the line between lightly over
consolidated state and heavily over consolidated state, this line separates it. So, any state
which is on this particular line on the left of COCRL will be HOC and towards the right of
COCRL will be lightly over consolidated and on this line it is NCL that is marked here. In
general now, where does the critical state line fall?
So, in general, CSL can be assumed to fall between normal consolidation and COCRL. So,
the critical state line falls somewhere in between NCL and COCRL, we need to understand
this that these are all certain concepts. Now, according to that, it is somewhere here. So, this
portion is lightly over consolidated, towards the side also likely over consolidated, towards
this side also lightly over consolidated. Now whether a lightly over consolidated soil will
exhibit peak or not these are all dependent on that specific type of soil, but definitely towards
the left of COCRL it is going to exhibit a heavily over consolidated behaviour.
The value of critical overconsolidation ratio depends on the soil type. So, what is the
overconsolidation ratio below which it is a LOC and above which it is HOC, it is dependent
mostly on it on soil type, but in general COCR that is critical overconsolidation ratio can be
considered as to in most of the cases for practical purpose we can consider the
overconsolidation ratio as to which is termed as critical overconsolidation ratio.
Now, below 2 if the OCR value is less than 2, we consider it to be LOC and above 2 can be
considered as HOC, these are all some bare ideas like it is not a very strict philosophy, it is
kind of a rule of thumb below which it is considered to be LOC, the region on the right side
537
of COCRL. Now, this COCRL is the reference towards the right side is called wet side. Now,
this particular definition of wet side, dry side we have seen in compassion curve.
What is towards the right of OMC we call it has wet side of optimum and towards left it is
called dry side of optimum. Similar to that it is defined that towards the right side of COCRL
it is wet side and towards left it is dry side. Now, wet side, dry side I am introducing because
the basic philosophy of critical state soil mechanics put forth by Schofield and Roth, they also
use these terminologies.
So, it is important, also it is good that we understand these terminologies better. So, that is
why I am introducing and as such wet side, dry side does not have much significance if you
understand it is towards right or left. So, the region on the right side is wet side and towards
left off COCRL is called the dry side. Now you need to keep in mind the difference between
COCRL and CSL is not much.
So, sometimes you can find certain discussions wherein the wet side and dry side is also
talked in terms of critical state line as well. So, we can see that dry side of critical state line or
towards wet side of critical state line. In fact, the overconsolidation ratio separating LOC,
HOC it is marginal. So, one can always say that towards dry side of critical state line, it
exhibits HOC behaviour.
So, wet side, dry side here, we are specifically discussing with respect to critical
overconsolidation ratio line, but keep in mind one can also refer to in terms of critical state
line as well. Now, here it is marked as wet side where specific volume that is v is greater than
v critical, v critical is the critical void ratio with reference to COCRL and if it is less than
COCRL then it is dry side.
Keep in mind we can also refer to vc as the specific volume along the critical state line as
well. So, please keep in mind this difference in general the definition is like this, but can also
refer vc as the specific volume on critical state line.
538
Now, let us see some important aspects related to undrained strength, we have discussed this
Su is the undrained shear strength of soil which we discussed in the module 2. So, we will see
in terms of critical state framework how does it look like. Now, we know the relationship τ f,
which is the failure state and for critical state framework, the failure state is same as the
critical state τ cs which is equal to σ’ cs tan φ’ cs is applicable for effective normal stress.
Now, this relationship it is familiar to us and we talk this in terms of critical state or the
effective normal stress. Now, undrained strength we need to represent this differently, which
is not similar to that of this particular relationship, this particular relationship we know it is
more like a Mohr coulomb failure envelope with inclination of φ’ cs. So, τ versus e how does
it look like?
So, this is shear stress up versus void ratio and this is the critical state line. Now, here this in
the case of undrained shear strength, we have τ cs that is the critical state strength is equal to
undrained shear strength and the void ratio corresponding to that is e cs. Now, we can see here
that strength decreases with increase in e as can be seen here. Now, there can be Su here,
there can be Su here.
So, as e increases the value of Su comes down. Now, we know that the relationship Gw = Sr
*e. Now, this is a saturated state, G is a constant. So, as e increases w increases that means, w
and e is related. So, it also means that with increase in e Su decreases, which also mean which
is very correct also as the water content increases, the Su value comes down.
539
So, strength is there by related to w as well, we are talking about undrained shear strength
very specifically and undrained shear strength is highly sensitive to its initial condition in
terms of e or w. Now, we also know Su refers to the strength when soil is sheared without
change in w. So, it is at constant w condition, we are doing the shear test. So, there is no
change in w why because there is no change in drainage condition.
And hence the e also remains constant. So, we can write τ f = τ cs = Su which is the undrained
shear strength, we also have the relationship τ cs = σ’ cs tan φ’ cs. Now, in this case, we can
write τ cs = S u. So substitute for τ cs = S u. Now, whether this equation holds good or not. So,
based on our previous understanding, please refer back to module 2.
What we have discussed when we discussed about UU test like we have only one effective
stress circle because of which we are not in a position to define the effective stress Mohr
coulomb envelope. But while discussing that, we also have seen if φ’ cs is known, one can
always draw the Mohr coulomb failure envelope for the very one effective stress circle that
you obtain from the UU test which means there is a point of intersection between the
effective Mohr circle envelope with the tresca envelope which is given by S u.
So, there is a point where it intersects. So, this particular equation holds good and that can be
explained in this particular manner. So, this is τ versus σ’ and this is the Mohr coulomb
failure envelope. Now, this is the tresca failure criterion wherein S u is the undrained shear
strength. So, this point is what we are discussing about. So, this point this particular equation
is very much valid.
So, S u = σ’ cs tan φ’ cs. So, we can write σ’ cs = S u / tan φ’ cs, we also have the relationship e cs
= e γ – C c log σ’ cs, this comes from our previous lecture. So, please refer back. So, we can
write e cs = e γ - C c log, this σ’ cs we can substitute it here. So, we have S u upon tan φ’ cs. So,
we can write this goes on the left side, we have C c log S u / tan φ’ cs = e γ – e cs upon C c.
So, here you are relating S u as a function of e cs and other parameters. So, here you can see Su
changes with change in e or w which we have already explained, since it is changing with the
state it is not a fundamental soil parameter and that is where we have started off with C’ and
φ’. So, for undrained shear strength, it still holds good that it is not a fundamental soil
parameter.
540
Su is a state dependent parameter. So, what is the state of the soil for undrained shear strength
it is more dependent on state. State means, how and where the soil initial point is positioned
with respect to the critical state line. So, that is defined by e or w. So, it is a state is defined
by void ratio in the critical state framework we define it using void ratio. So, initial void ratio
gives where it is with respect to the final state which is on the critical state. So, that is why Su
is known as the state dependent parameter. So, we will come to what is meant by state in the
subsequent slide.
So, critical state now, let us see in q, p’ format, we are explaining in 2 ways; critical state
directly we know it comes from τ σ’ response, whatever we know from module 2 and all
those concepts remains same in q-p’ framework, because you will see later the critical state is
mostly defined in terms of stress invariants which is q- p’. So, we will be discussing both.
First we will discuss in τ σ’ format and then we will come to q- p’ till we introduce all the
concept, further on we will deal only with q- p’ v space or you can also write it in terms of q,
p’ e. Because e and v is there is only a difference of 1. So, 1 + e gives me. So, both are valid,
but while understanding critical state framework, let us discuss both. So, whatever we have
discussed now, in terms of τ σ’ e we will see that in q, p’ and v. So, responses similar to τ σ’
e. Now, when we discussed q, p’ v in the previous lecture, we have not discussed the failure
state that is what is how q f and p’ f is correlated?
541
So, we are yet to discuss that, so that we will discuss now. Now, this particular response we
have not discussed in the previous lecture. So, it is q, p’ and critical state line is given by this;
this is nothing but the K f line that we have seen before. Now, how it becomes critical state
line because, here the φ’ considered as φ’ cs. So, that is why K f automatically becomes a
critical state line.
Let us assume p’ f = p’ cs. Now, critical state, the failure state is same. So, p’ f = p’ cs and for
which q f = q cs. Now, this is related. Now, the slope of the critical state line is given by
capital M. So, critical state line in q, p’ is defined by q f = M into p’ f very important
expression which you need to keep in mind. So, critical state in q, p’ v space is defined as q f
is equal to M into p’ f.
It is also good to write q cs = M into p’ cs. So, both are valid. What is M? M is the slope of
failure line in q- p’ plot and we already have defined or we have derived this expression for
the slope in q- p’ again please refer back we have already done it and for compression case it
is 6 sin φ cs / 3 - sin φ cs. Now, you can see here we are denoting it by φ’ cs.
So, that is why here the failure line K f is equal to critical state line. Now, let us see this in
terms of v ln p’. Now, this is the NCL with a slope of λ, the critical state line is position like
this, because we know the right most boundary is NCL and there is no state beyond NCL.
Now, this is the unloading reloading line with a slope of κ. Again we have already discussed
this.
Now, let us see here p’ f = p’ cs which is lying on the critical state line where v f = v cs. Now,
all the important parameters are defined that is p’ cs q cs and v cs. Now, we already have a
reference line or the reference point on every line which corresponds to p’ equal to 1 kPa. So,
here on NCL, it is N 0, for the unloading reloading line is v κ and for critical state line is
capital γ, which can be represented.
So, all the initial states of the line has been defined based on which the equation for this line
NCL and critical state line can be written. So, we can write the equation for critical state line
in v ln p’ as γ, that is this point - λ is the slope because this is also λ - λ ln p’ f. Now, p’ f
means p’ cs. So, every point on the critical state line can be defined using this expression v f is
542
equal to γ which is corresponding to 1 kPa - and then it depends upon where actually the
point is.
So, this is the p’ f. Here the parameters λ , M and γ are independent of initial state why,
because these are the parameters which represents the critical state line and can be considered
as fundamental soil parameters. Earlier we have understood in τ σ’ space, the τ σ’ e space that
C c which corresponds to λ and the other parameters it can be used for defining the
fundamental parameters of soil.
So, similarly in q v p’ space you can write λ M and γ these are the fundamental parameters of
soil and it can be considered in the case of critical state framework. M and γ that is the slope
M and γ measured in compression is different from extension that we have already seen. The
slope of K f line in extension is different for q- p’ plot.
So, similarly, M is also marginally different in the case of extension case. So, now, we are
basically discussing with reference to competition line. So, we will not bother about that right
now. Also, we know that in the previous slide we have discussed about undrained shear
strength. Now, here also we can very well define undrained shear strength q S u = q f / 2.
Now, q f is σ’1 - σ’3 / 2 is the radius of the circle which gives you tresca criterion.
So, Su is very much equal to q f / 2 and q f = M into p’ f / 2. So, we can also write ln 2 S u / M
= γ - v f / λ. The way we have written previously in terms of log. So, how this is possible?
This is possible because you are writing the expression for ln p’ f, which is equal to γ - v f
upon λ which is equal to ln p’ f and p’ f is nothing but to S u / M which from here. So, that
gives the expression for S u in terms of critical state parameters.
543
So, now, let us further try to explain every point in q- p’ plot has the corresponding point in v,
p’. So, that is true whatever we are having here wherever we are taking we are taking a point
here can be correspondingly marked in the v ln p’ plot as well. Failure is defined by a single
surface in v, q, p’ which is the critical state of that particular soil. Now, unloading reloading
line crossing critical state line in v- p’ plot does not mean any failure condition.
Now, this is something which I am explaining, because when we discuss this in the normal
classroom, we have found that some students always have this doubt that you can see here
critical state line is placed here. Now, from here the unloading line it goes and it crosses the
critical state line. Now, sometimes there is a misconception that it is crossing the critical state
lines. So, the failure has already occurred, it is not true.
Critical state line is only a relative position of the failure condition, the unloading reloading
line it only determines the stress history of the soil. So, crossing this critical state line in
please underline in v, p’ or v ln p’ plot is not a failure condition. Obviously, when it crosses
or when it touches the critical state line in q-p’ plot it is the failure condition not in v ln p’
plot.
So, please do not confuse yourself this with respect to v ln p’ it is only the relative position or
the state of the soil when I say that it is on unloading reloading line or on NCL, it is only the
initial state of the soil and during shearing it moves towards the critical state line, it is not that
merely this crossing does not mean that the failure has occurred. It only indicates the initial
state of the soil with respect to stress history; you know that different points on the unloading
544
line with respect to the given NCL gives different OCR, whether the soil is on wet or dry side
of the critical OCR line.
So, that is the only purpose of these points. Failure is entirely governed by effective stress
path when it touches the critical state line which comes from this particular figure. So, when
the effective stress path when it starts from a particular point and when it reaches the critical
state line, then that explains whether the failure has occurred or not. For failure to take place
soil has to yield first, so for this also we know for failure to take place, the soil has to yield
first.
That means we need to introduce another concept, which we will see shortly that we need to
define what is soil yielding? Now, when we discussed in the case of module 2, we never
discussed this point explicitly, but it is there, but we are not discussing that explicitly, but in
this module, we need to clearly state when the soil is going to yield. So, without a yielding
failure cannot take place.
So, I am not going into the details right now, you can underline this fact that for failure to
take place soil has to yield first. This particular understanding is extremely important when
you try to understand the behaviour of OC. So, let us hold on for the time being about this
yielding, but make sure that we understand that for failure to take place soil has to yield first.
So, that is all the preliminary discussion about critical state in both the format's that is τ as
well as q, v, p’ format.
545
Now, we will discuss a bit about normalization of v, q, p’ plot with respect to some initial
condition or some specific parameters, so, that the comparison becomes easy. Consider a
point P. The state of P is defined by so, this is the point P. Now, state of P can be defined by
what is you can assume a line which is parallel to NCL with slope λ passing through P. So,
the state of this particular point P can be defined by p’ p, v p as well as q, p in q, p’ plot.
Now, corresponding to this particular v p, if we draw a horizontal line, we can also find p’
corresponding to that on the critical state line and that is p’ pc. Now, the value of specific
volume corresponding to 1 kPa on the line through P is given by v p λ. The effective stress
corresponding to v p on NCL line is given by p p’, p’ p M. Now, p’ represents it is effective p,
mean stress.
This p represents that it is with reference to point p and n represents that it is on NCL
similarly, here c represents it is on critical state line. So, there are certain reference points
now, v p is the specific volume for point P. Now p’ corresponding to v p has been marked and
that the normalizing parameter is p’ pc and v p λ. So, these 2 parameters now, for every point
in this space, you can find a corresponding v p λ and p’ pc.
Now, you may be wondering why the normalizing parameter is not p’ p, because this p’ p is
dependent on the position of the point p, whereas, p’ pc, it is a more stable point, because for
every p you are considering the point or the location of this point with respect to critical state
line. So, that is why p’ pc can be a good normalizing parameter.
Obviously, this 1 kPa reference is also a very good normalizing parameter. P’ pn is also used
as normalizing parameter. So, what is the distance of point P from NCL that is p’ pn. So, this
can also be used as a normalizing parameter, but rather than normally consolidated line
critical state line is of more importance, why that distance actually represents the factor of
safety.
So, for all practical purposes the normalizing parameter can be considered as v pλ and p’ pc,
we can write v p that is specific volume along this line is equal to v p λ - λ ln p’ p. Now, when
we substitute ln p’ p that is this particular point we are discussing v p of point P. So, v p of
point P = v p λ - λ ln p’ p. Now, when you know the initial point p’ p is known, v p is known, v
pλ can be determined.
546
So, that becomes a reference parameter. So, v p λ = v p + λ ln p’ p. So, v p = γ - λ ln p’ pc. Now,
v p is this particular point. So, you have this v p on critical state line as well. We can very well
write v p = γ – λ, λ is the slope of this, λ ln p’ pc because this is the reference point. This
equation we have already discussed previously for critical state line.
So, v p is known now, v p depends on the position of p, γ is the critical state parameter which
is known, λ is known, p’ pc can be determined. So, ln p’ pc = γ – v p / λ. So, both the
normalizing parameters can be determined. Now, when you normalize the whole of the v, q,
p’ plot how does it look like? So, let us normalize with respect to p / p’ pc and p’ / p’ pc.
Now, how does it look? By normalization both critical state line and NCL will appear as a
point. Now, both instead of a line it becomes a point how? So, let us say that this is the point
which is the representation of that particular soil in q-p, p’ q / p’ pc p’ / p’ pc space, this is the
normalized plot. So, this is the point corresponding to CSL. Let us see how? Now, along
critical state line this line p’ = p’ pc.
So, any point we take, let us say take this point again this is p’ pc. So, p’ = p’ pc. So, p’ by p’ pc
= 1. So, this is not going to change. So, this 1 is represented by this particular point. So, you
have a line along p’ / p’ pc = 1. So, that is fixed for a critical state line. Now, q pf / p’ pc = M,
how that is q p / p’ f, that is q f / p’ f = M. Now, p’ f is nothing but p’ pc because it is only
critical state line.
So, we know that that particular ratio is M. Now, for critical state point, we know that q / p’ pc
- p’ / p’ pc is equal to M why? Because here p’ / p’ pc = 1 in this particular case and q / p’ pc or
this particular one is already equal to M. So, this becomes M. So, where you find a particular
point with a slope M which intersects this particular line is the point of critical state line.
So, that is how we can say that critical state line is a point in this normalized space. Now, for
considering NCL to be isotropic let us say that the initial point is isotropic. So, in that case,
σ’1 - σ’3 = 0 that is q = 0. Hence, q / p’ pc = 0. So, this q is 0. So, NCL will be represented by
a point anywhere on this particular line where it is 0. Also, p’ pn / p’ pc is a constant.
547
You can see that p’ pn is this and p’ pc is this; this remains always a constant wherever you
take, so, that constant is not going to change. So, always it can be represented by this
particular point. So, NCL is represented by a point. Now, depending upon the different type
of soils the relative position will change that is a long this it will change or along this it will
change. But these CSL and NCL can be represented by a particular point.
Now, we have discussed about state of the soil. So, we will discuss state of the soil in detail
and the state parameters. And NCL can be represented by a particular point. Now we have
discussed about state of the soil. So, we will discuss state of the soil in detail, and the state
parameters. Now, state of the soil is relevant for a given point. Now, here the reference point
is s peak P. State of soil also play a role in governing strength and stiffness, along with e, w,
p’ and OCR.
So, whatever be these are parameters that is going to define what is the strength and stiffness
of the soil? So similarly, it is important to understand how state of the soil also play a
governing role in defining the strength and stiffness. It is one of the same because once you
know this, you are defining the state, but the parameters in v ln p’, in terms of state
parameters, we will define.
State of the soil defines the initial point from CSL; you can see here, this point P, how
distant, it is from CSL is the kind of factor of safety. Consider point P. So, this we have
already done, vertical distance of point P from critical state line is Sp. So, when we define
state of soil we are defining what is the state of this particular soil initially with respect to
548
critical state line? So, that is why we are defining the vertical distance of point P from critical
state line, you can see this red line.
Now since these are parallel. This is also the same as this distance and this particular distance
is represented by Sv, which is the state parameter. So, that is what we are going to define.
This state parameter is what we are going to define. So, the vertical distance is Sv. Now, Sv
can be written as γ - v P λ, what is γ? This distance is γ and this distance is v P λ.
So, S v from the geometry you can very well write γ - v P λ. Horizontal distance of point P
from CSL is given by this one ln scale. So, ln Ss. So, this is the horizontal distance. Now ln Ss
can be written as ln p’ pc - p’ p. So, Ss can be written as p’ pc / p’ p, this p’pc / p’ p, you can see
that this is very much similar to the definition of OCR.
OCR is with reference instead of p’pc, it is with reference to the stress state on the NCL,
which is called as yield stress. So, here Sv / ln Ss = λ here, if you consider this particular line
and this. So, this is ln S s. This is ln Ss and this is Sv, so Sv / ln Ss is equal to the slope which is
λ. So, that is given as λ, so we can always write Sv = λ * ln Ss. So, this is the relationship
between the state parameter.
So Sv and Ss are called state parameters of that particular point or what the soil state is with
respect to critical state is given by the state parameters and the vertical and horizontal or Sv
and Ss. State parameters are related to critical state line, Ss is very similar to OCR, this
definition of Ss is very similar to OCR. In OCR is related to normally consolidated line,
whereas Ss is related to critical state line.
549
So summary of critical state parameters, now what we have done is in the name of
introducing critical state, we have discussed quite a lot of things in these 2 lectures, but what
is very essential as we move forward when we interpret critical state soil mechanics. So, that
summary, we are providing here. So, whatever we have learned finally, these facts, what
we're going to discuss now is very important.
So, only effective stress path is considered, you have ICL or NCL. ICL means isotropic
consolidation line or NCL with slope λ, CSL with slope λ, unloading reloading line with
slope κ, the initial point reference p’ = 1 kPa, N0γ and v κ. And we need to see only effective
stress path is considered for defining failure. So, then we have q p’ plot where q f upon p’ f is
given by slope M.
So, M λ κ they are important. CSM utilizes the index properties, determining C c from liquid
limit, because C c is the one which we get from the consolidation plot. So, this C c is very
important and we know that C c can be determined from liquid limit. So, CSSM utilizes index
properties in one way or the other. Then effective stress that is σ’n that is the normal stress or
the consolidation stress σ’ which is always in terms of effective.
Stress invariants, because q and p’ we are considering and stress path, whether it is drained or
untrained how it moves. So, CSSM utilizes these information as well. The primary
consolidation data that the C c ,mv, λ and κ. Shear strength q, p’, M and φ’ cs. So, all these we
have discussed and critical state framework utilizes these properties.
550
Isotopic consolidation are normally consolidation line, which is given as this, where it is
defined as N0 - λ ln p’. Unloading reloading line, which is defined as v = v κ - κ ln p’. All this
we have discussed, I am just summarizing it and critical straight line CSL is defined as v f = γ
- λ ln p’ f. So, you can see that with reference to this here, you are adding the subscript f in
for critical state line.
And the additional one what is needed is q f = M into p’ f. So, these are the essential
relationships, which we have to keep in mind for dealing with critical state framework.
We also need to understand the relationship of C c and m v with λ and v, because we need to
cross compute these as and when required. We know that C c = - de / d log p’, from our initial
definition of C c and that corresponds to normally consolidated portion, v = 1 + e, dv = de,
substitute for in the previous expression C c = - de is replaced by dv and d log p’ can be
written as 1 / 2.303 d ln p’.
Now d ln p’ is 1 / p’ into dp’, so that will give you C c = - 2.303 p’, that goes into a numerator
dv / dp’. Now we have the expression for normal consolidate line as v = N0 - λ ln p’, dv, if
you differentiate this, you will get - λ dp’ / p’. So, we can write dv upon dp’ = - λ / p’. Now
you can see, this expression and this dv / dp’ can be substituted.,
C c can be written as 2.303 λ. So, knowing C c λ can be determined. So, Cr that is for
unloading reloading line, the same equation holds only is instead of λ, it becomes κ.
551
C c and m v relationship with λ and v will continue, we need to find out the relationship for m
v. So, m v again from the basic definition, we have - de upon 1 + e 0, e 0 is the initial void ratio
upon dp’. So, m v is equal to again substituting we have dv 1 + e 0 is initial specific volume v
0 dp’, dv / dp’ we know that is - λ / p’.
So, m v = λ / v 0 p’. So, parameters λ, M γ are critical state parameters in v, q, p’ space. Please
keep these relationships what we discussed as summary in your minds, because these are very
important as we move further. These parameters can be considered as fundamental soil
parameters.
Some additional note on critical state before we wrap up the introduction part, the strain
required to reach critical state depends on the initial soil state, whether it is on wet or dry
552
state of critical. Soil sample strain without change in shear stress, effective stress and volume
at critical state. So, there is a shear straining happening at constant condition. Shear strength
tests are normally terminated at strain, less than 20%.
It depends upon the codal provision, what is the termination criteria? This strain may not be
sufficient to achieve critical state for soil samples on wet side of critical. Again I told that
these are all concept put in place. Now in reality what actually is happening to the soil we
really do not know? So, we have some termination criteria at a high strain and high strain we
consider it to be maybe less than 20%.
So, we presume that the state what we define as critical state has been achieved by the soil
during shearing, but we are not sure. Now, we also need to keep in mind that even at 20%
strain there are possibility that the soil has not reached its critical state. So φ’ that you define
based on this 20% strain actually may not be φ’ c as it may be even less, so what, actually it is
we do not know, but we presume it.
Now, for wet side of critical state it may so happen that it will not reach the required critical
state at 20% strain. Now wet side of critical means what it is mostly referring to normally
consolidated or lightly over consolidated state. Now for this there is a progressive
densification that is happening upon yielding. After yielding it is strain hardening. So, it can
build up more and more.
So, that is the possibility why we say that it may not reach to its critical state, whereas in the
case of OC state what happens with sharing, there is dilation, so the progressive softening
that is happening that soil particles are disintegrated or separated from each other. So, here
the critical state is fairly easy. The strain may be sufficient for whatever 20% what we discuss
this strain may be sufficient for soil samples on dry state, dry state means nothing but the
heavily over consolidated state.
It is ideal to conduct shear tests on lightly over consolidated samples for determining critical
state. Since the initial state is close to critical state. So, if you are planning to obtain the
critical state parameter one suggestion is that if the soil is lightly over consolidated then
fairly, it is easy for the soil state to reach critical state if you see that in terms of v ln p’ plot
you can see that LOC is more close to critical state than HOC and NC.
553
So, if we start the sharing it likely over consolidate state it is fairly, we can say that by 20%
strain yes there are all possibilities that the soil would reach its critical state. Critical state line
passes through origin in q, p’ plot, no strength at zero confinement. Now if there is cohesion,
definitely that is going to be there at zero confinement, that is what we know. Now in q, p’, if
we are representing the K f line in terms of critical state line, it passes through origin.
Now what is the logic here? If there is cohesion, it does not mean that it is not considered let
us see what actually does it imply? So, there is no strength is zero, we are not considering any
strength for the soil at zero confinement. That is we are discussing with respect to critical
state line that we have to keep in mind. In case of true cohesion, we are not discussing about
apparent coercion.
But if there is true cohesion soil will exhibit some sort of cementation characteristics and if
there is cementation characteristics, obviously, there will be some strength at zero
confinement. There will be strength at zero confinement, but in the case of critical state line
we do not consider this. Why it is because the strain required to reach a critical state is
enough to break the true cohesion or cementation.
So when the particle undergoes shearing, it is undergoing volume change and we know that
under critical state there is the movement of soil particles more like a fluid frictional fluid we
call it. So, there is a turbulent motion of particles. Now when we define such a situation for
critical state it is very implicit, that the particles bonding need not be considered at critical
state. That is a logic which states that there is zero strength at no confinement condition.
So, critical state line passes through the origin in q, p’ plot. It is hard to exactly achieve
critical state during shear test for the reasons we have already stated, it is often assumed that
critical state has reached for the strain level considered in the test. Now the concept of critical
state is also a hypothesis and in the testing in the actual practice for its implementation we
presume that the termination criteria that we adopted if the soil has not failed.
That particular strain level corresponds to critical states. So, whenever we are using this
termination criteria we need to keep in mind that it is based on the assumption that we are
following this.
554
So, let us summarize the last 2 lectures, where we have discussed or we have introduced the
critical state soil mechanics. Critical state soil mechanics is a concept that integrates
consolidation and shear strength behaviour for defining failure state which has not been done
in Mohr coulomb failure envelope. The 2D representation of void ratio versus effective stress
and shear stress versus effective stress is integrated to obtain 3D representation of critical
state we have not done this yet that will be towards the end of this module that we will see
how to integrate the concept.
According to CSSM all soils will fail on a unique failure surface defined in q, p’ e or q, p’ p
space. Critical state, in general is attained at shear strain greater than 10% why instead of
20% I have used 10% is in some codal provision it is 10%, in some it is 15%. So, in general,
less than 20% is the understanding. Critical state parameters can be considered as a
fundamental soil property.
Accordingly φ’ cs because we discussed it in terms of τ, σ’ e plot. So, φ’ cs e γ Cc can be

considered as fundamental soil parameters. Normal consolidation line NCL, unloading-
reloading line and critical state line are important for defining CSSM, we have seen that. In q,
p’ v space, slope of NCL and CSL is λ. And the slope of unloading-reloading line is κ. NCL
forms the right side boundary for the soil state to exist; we will be discussing this again in the
subsequent lectures. Left of NCL is LOC and HOC.
555
Importance of OCR and yield stress ratio discussed. Soil bulk stiffness is dependent on λ and
κ, depending on whether it is loading or unloading and hence it is not a fundamental
parameter. Critical overconsolidation ratio line is the boundary between LOC and HOC. The
region on the right side of COCRL is called wet side and left is dry side, which is heavily
over consolidated.
Undrained strength expressed in terms of CS parameters both cases for τ space as well as q, p’
space has been discussed. How to normalize v, q, p’ plot has been discussed. State of soil and
state parameters has been defined. Parameters λ, M and γ are critical state parameters in v, q,
p’ space and are fundamental soil parameters. Finally, we need to keep in mind that critical
state is a concept. And there can be departure of this concept from actual reality. It is
practically difficult to assess whether the soil has reached critical state during sharing.
In most cases, it is assumed that the soil reached critical state during sharing that is at high
string condition, which is practically greater than 10%. So, that is all for today's lecture. Now,
we have introduced critical state framework. In the next lecture we will see more about 2D
representation stage by stage we will progress. And then finally, we will discuss about 3D
formulation, how to integrate the concept. So, that is all for now. Thank you.
556
Lecture-46
CSSM-2D Representation
Welcome to all the participants, in the last lecture we have introduced critical state soil
mechanics, we have seen different elements of critical state. The various parameters and those
parameters considered to be the fundamental behaviour or properties of soil. Now in today's
lecture we will see the two dimensional representation of critical state soil mechanics, what we
initially introduced is that we are integrating the aspect of deformation or maybe volume change
characteristics along with stresses.
And hence it becomes a three dimensional representation. So, before moving on to integrating
into a 3D space, let us first understand the 2D space properly and what are the important factors
that come into play? What are those elements which we know but we have not projected in this
manner? That is what we will be seeing in today's lecture.
Basically the 2D representation of CSSM critical state soil mechanics. Now as I told in the
beginning for some lectures we will parallely understand both representation of critical state.
557
One is e versus log σ’ or e τ σ’, e void ratio, τ shear stress and σ’ is the normal stress, effective
normal stress. But we are interested in the representation of in terms of stress invariance.
So, we have v which is the specific volume q and p’, so we will be discussing both. So, in today's
2D representation also we will start with e versus σ’ plot for NC soil. Now when we do this
discussion, we have to keep in mind we have to understand CSSM discussion from the
perspective of stress history and from the perspective of drainage condition.
So, then we say, then we have to appreciate the point that this initial conditions in terms of stress
history and the drainage conditions is not going to affect finally the critical state parameter or the
critical state framework, so we need to discuss from those angles. So, first let us see e versus σ’
plot for NC soil, normally consolidated soil. We start with this particular curve e versus σ’,
please note it is not log σ’.
Let us consider three identical clay specimen, which is subjected to consolidated drained shear
test. So, now understand that we are first considering the drained test subjected to CD shear test
for three different σ’. This is what we generally do in triaxial testing also, we subject the soil to
three different confinements or maybe in the case of direct shear test, we subject the soil to three
different confining normal stress and then we find out the shear stress.
So, similarly we are considering now three identical clay specimen which is subjected to CD test.
And since it is normally consolidated, you can see the stress strain the shear stress versus shear
displacement it is not in terms of strain, it is in terms of displacement for three different normal
stresses where σ’n1 is less than σ’n2 less than σ’n3. Now for sigma n1 dash, we know that e is
going to change.
So, these the information that we have known in our previous lectures I was every time I was
hinting this point when we were discussing these points in module 2 as well as in module 3. Now
we need to keep those points in mind while discussing this module as well. So, what is the first
thing that we understand for normally consolidated soil, it is compression, now if it is
compression we know that the void ratio is going to decrease?
558
So, this is one information and that is what is shown here, e is decreasing. Now why it should
decrease in a vertical manner? Downwards, so downwards we have got the answer why vertical.
That means that σ normal stress, normal effective stress is not changing during the course of
shearing. And this is true, because this is a consolidated drained test, so shearing is under drained
condition.
And when it is under drained condition there is no development of excess pore water pressure.
So, whatever is the initial condition σ’n which is the confinement, please note this is an important
point here. Like what changes during shearing? The shear stress changes during shearing and it is
done at a constant confining pressure. And confining pressure here is σ’, so that is the effective
confining pressure.
So, that is not changing throughout the test because it is drained test. So, that is why
corresponding to σ’n1, we find that void ratio reduces and it touches the critical state line at this
point where it fails. Now let us say σ’n2, the same thing happens and σ’n3 also the same thing
happens. And wherever the path here it is not actually stress path, when the path of e variation
touches the critical state line, then the failure occurs.
Since the test is CD, total and effective normal stress is same, so that is why it is not moving on
left or right rather it is vertically down. There is no change in σ’n throughout the test. Now if you
plot this in terms of τ and σ’, how will it look like? So, σ’n1 is vertically upwards here, why
because this is not changing. And φ’cs and this is the critical state line, σ’n2 and σ’n3.
So, we have these points lying on the critical state line. Now critical state line represents the
failure state of all soil specimen irrespective of stress history with zero cohesion. You will hear
this sentence quite a number of times throughout this lecture. Because it is very important that it
gets into our head properly, why? Because that is the concept based on which the critical state is
portrayed, like at critical state irrespective of any condition.
559
So, we will have quite a few examples hinting the same point, and this statement will be in a way
repeating throughout this module. So, what it states is that whatever be the initial state, now we
are considering only the normally consolidated state, you will see towards the end of this lecture
this particular sentence will be even more clear to you rather it all falls on this critical state line,
this one.
And with zero cohesion, and this particular aspect we have discussed in the previous lecture,
what did we discuss? At critical state, the strain which the soil undergoes is significant, now at
that level of strain in general there cannot be any sort of cementation or cohesion. So, that is the
principle based on which it is told that critical state line always passes through origin. Now
knowing the NCL and the critical state line, it is possible to predict failure state and e changes
for a given σ’n.
So, given these conditions it is quite easy once we define the normally consolidated and critical
state line, it is very easy for us to predict this the failure and the change in which how the e
changes during the test but with some prior information.
Now let us see undrained shearing of normally consolidated soil. Let me again remind you that
the concepts that we are discussing it is not new rather you know this and you have studied this
in one or the other form. If you have not studied during your undergraduate at least you have
560
studied this in the previous modules. So, we are just placing those information at the appropriate
place, so please follow the previous lectures very carefully.
So, that this explanation does not need much time for you to grasp. Again we start with e σ’. Soil
sample is initially consolidated to σ’n0, so this is the point to which it is initially consolidated, so
it is the starting point. Now this is reminding you it is an undrained shearing, so it is kind of
consolidated undrained test where the soil is initially consolidated to σ’n0.
For undrained test, so this again becomes a quite a handy information. So, if you do not
appreciate that point that void ratio remains same during undrained test, you will not be able to
understand this better. So, for undrained test e remains constant throughout shearing. So, we
know that this is the starting point; it is very easy to predict, what will be the path towards the
critical state line at constant shearing?
Definitely the soil the state cannot move in this direction, this is impossible state, we have seen
that, NCL becomes the right boundary. Now during shearing e remains constant, so how it
should move? It has to move in this direction, so this is the only possible way by which the state
of the soil would change. That is this is the starting point and during shearing for NC at constant
e, this is the path till it reaches the critical state line.
Now here you can see in comparison to the previous slide the starting point and the end point is
different. In the earlier case this was constant, that is σ’n0= σn0 whatever is the starting point that
remains same, because e was changing. But in this case e is constant because it is undrained and
due to the development of excess pore water pressure the initial confining stress keeps reducing,
the effective confining stress keeps reducing.
For undrained shearing positive u excess is generated for NC soil. Now this again becomes an
important information from our previous modules, what? During undrained shearing or during
drained shearing the tendency of normally consolidated soil is to compress. So, corresponding to
that in an undrained test there will be positive pore water pressure, so that is what is written here.
Final effective stress σ’nf will be less than σ’n0, so this is the point of σ’nf.
561
So, let us see this on τ σ’ plot, so this is the point where it starts, now the stress path will be. So,
the σ’nf will be this, now this is the point from where the σ’nf comes down and this is the stress.
Now this is the starting point, it has to end on the critical state line, so this is the point. So, this is
the effective stress path in τ σ’ plot and this we have already seen it moves leftwards and this is
the pore water pressure.
So, that is about the undrained response of normally consolidated soil. Now e versus σ’ plot for
OC soil, now let us see what will be the implication of stress history. We will start with the same
discussion e versus σ’, now what is happening? For over consolidated state what has to happen?
The soil has been consolidated to this particular stress level and then it is further unloaded, so
this is the unloading path.
So, this is unloading path and hence the final point reaches here. Now what is the over
consolidation? From here to here is the over consolidation. So, soil consolidated to maximum
stress or yield stress what we call σ’ y, this is the maximum stress to which it is subjected to in
the past. Soil unloaded it to point P, so let us say this is the point P, where the stress is σ’ p, so σ’
p, so OCR is σ’ y / σ’ p.
562
Now this is represented on e log σ’ plot. All these representations can be considered for all sort
of representation, you should not get confused why e versus log σ’ is discussed only for OC not
for NC, it is applicable for all. It is like I am using different framework for representation all of
them are equally valid for all the cases. So, σ’ y, the point P σ’ p marked same.
In drained test e changes σ’ p remains constant that is what we have already seen. For HOC e
increases due to dilation, now this is again another important fact which we should be knowing.
Otherwise from here how the stress state would change, we will not be able to understand better.
So, for OC we know that it is dilation, now for dilation what happens? e has to increase, so how
the stress state will move this is represented by PQ.
You can see that during shearing it is moving upwards, and why it is moving vertically upwards?
Because σ’p remains constant because it is a drained test with excess pore water pressure 0. So,
σ’p is not going to change, initial and final states are the same, now e instead of moving
downwards in it is moving upwards, now what is this? This is due to dilation which the OC soil
undergoes during shearing; same is represented in e log σ’ plot.
In τ σ’ how it will look like? So, it is again the same it is moving in the upward direction from σ’
p and it touches at this point where it fails. In HOC, now this is another important aspect we need
to keep in mind. You know that in heavily over consolidated soil it would first exhibit peak and
then come to ultimate state or the critical state. Now where is this shown here?
If we do like this it means that from this point the stress path moves and touches the critical state
point. Now where is the question of peak coming into picture, in fact this representation is not
fully correct, so what has to be done? We need to have understand that the stress path would first
hit the peak line or the peak state and then return to critical state, and that too we have to
remember it said σ’ p constant because it is drained test.
So, there is a peak state because this representation is also very clear to us in module 2, when we
discuss the shear strength. Because corresponding to every critical state we have the peak state
which is above the critical state. And there we have discussed effective peak angle includes both
563
the component that is a basic critical state component and the dilation component. So, that is why
we have this particular state, and this is called peak state.
So, first the stress path has to reach this peak state. So, here it is vertically upwards, the stress
path first moves, touches the peak and then returns to the critical state line, so this is the correct
method of representation. HOC has to first, the state of HOC soil during drained shearing would
first hit the peak state and then returns back to critical state which has been shown.
So, first it goes, touches there and then comes down, so that is what the representation is. Now
this brings the relevance of yielding of soil before failure. Now you may ask why it is crossing
critical state line, it would have failed by this time, it is not like that. For any soil to fail should
yield first from elastic the point of yielding and failure, this is the sequence of events. Now in the
case of over consolidated, it exhibits an elastic characteristics during shearing.
Now we know that at a very low strain itself it picks up and it reaches the peak. So, close to the
peak it starts yielding, so once it start yielding strain softening happens and then the stress comes
down, this much we know, this is very clear to all of us. So, initially the stress path goes towards
the yield line sorry not yield line, the peak line yield there and after yielding it comes down to
failure to critical state.
Anyway we will be repeating this in the following lectures as well where the same concept will
be told again and it will be pretty clear. As of now we have not started the concept of yielding of
the soil. Once we start discussing yielding this particular aspect will become very clear. So, as of
now we need to understand that concept of yielding is important, only after yielding the failure
can occur. And that is the reason why for a heavily over consolidated soil the stress path moves
to peak line yield there and then returns back to critical state line.
564
So, now e versus σ’ plot for OC soil for undrained shearing. So, we have discussed about the
drained characteristics, now what will happen in undrained shearing of OC? Now since OC has a
tendency to dilate the corresponding nature in the undrained shearing will be the result will be
negative pore water pressure, this we have already seen. We know that e remains constant, σ’
changes due to change in u during shearing.
Now pore water pressure keeps changing during undrained, in HOC since the dilation is the
tendency the corresponding pore water pressure will be negative. Now during negative due to
this negative pore water pressure the initial effective normal stress would change. Due to
negative u, now what is the implication of negative u? σ’ f which is the final or the failure normal
stress will be greater than the initial normal stress, this we know.
Because it is getting added up, u is getting added up, now this is represented by PQ, now what
will be PQ? It is very easy to understand undrained, because there is no change in e, and we
know the relative position of critical state, so from here it is should be horizontally to Q, so that
is σ’ f, so this is the final normal stress, same is marked in log σ’. And this is the way in which
the stress path moves, and this we have seen.
You remember in the case of stress path representation like we have marked like this that is
moving towards right. Now at that particular point of time we did not have this complication of
565
touching the peak and then coming down to critical state after yielding. We did not intentionally
discuss those points there but then actual manner of stress path is this; this is the correct way of
representation.
Like the stress path moves above the critical state line, it is not that it is moving from critical
state to peak, do not consider ever like that. It is going towards peak yielding and then coming
down to the critical state and that is very clear from the strain softening response of OC. And
here this stress path moves and it moves towards right and it is touching at this point Q, so this is
Δ u which is negative.
So, because of this negative your normal stress is greater than the initial normal stress. In
undrained test there is a point P that is σ’ P for which Δ u is 0. Now we need to understand that,
in fact I have not shown the total stress path here. Total stress path, now whatever is the effective
stress path if it is towards the left of total stress path it will be positive a pore water pressure will
be positive.
If the effective stress path is towards the right of total stress path then the effective pore water
pressure will be negative. This also, please refer back to stress path plot for understanding what I
just told. Now in this case you can see that from P it moves and then turns right wards, now
depending upon the level of over consolidation or depending upon the over consolidation ratio
whether it is a lightly over consolidated or whether it is a heavily over consolidated.
There can be a situation where this particular envelope will result in zero pore water pressure,
excess powder pressure at the failure state. So, in undrained there is a point P for which Δ u
developed will be zero, because it is moving from positive to negative. So, there can be a
situation where positive goes to zero, so the final excess porter pressure is close to zero. So, that
is what it means, so there can be one particular point P.
Now this is a heavily over consolidated maybe towards a lightly over consolidated state we can
expect such a situation. The point is the intersection of unloading line with critical over
consolidation ratio line. Now critical over consolidation ratio line we have discussed which is the
566
line which divides a heavily over consolidated state and a lightly over consolidated state. Now
this particular unloading line, where it intersects?
Now this is very close to the critical state line as well. Now where it crosses the critical over
consolidation ratio line, that particular point if you start shearing then that will result in Δ u 0.
Please draw this and understand because your critical over consolidation ratio line will be
somewhere here, as we have discussed. Now the point where it coincides, this unloading line
meets that particular point is considered to be P, not here, somewhere here.
If the shearing is done from that P then the developed excess pore water pressure towards failure
will be close to zero, so that is what it means. Right side of critical over consolidation ratio Δ u is
positive due to compression during shearing. Right side of critical over consolidation ratio line
means, it is lightly over consolidated and normally consolidated, so Δ u will be positive during
shearing, and the same is negative towards the left of COCRL.
Now you should again keep in mind that in some textbook you will not find COCRL rather it is
referred to as critical state line itself. There is not much of difference but the actual definition is
in terms of critical over consolidation ratio line.
567
Now 2D analysis of normally consolidated soil subjected to CU and CD in v, q, p’ plot. As I told
we are discussing in both the framework. Now whatever we have discussed for e τ σ’ plot we
will just repeat it for q p’. Because ultimately for three dimensional integration we need v, q, p’
plot. Soil is initially isotropically normally consolidated to p’ c, it simply means that ICL, that is
a isotropic consolidation line is same as normally consolidation line.
And it is consolidated to point A’ where the pressure is p’ c; the effective mean stress is p’ c. So,
this is the point A, drained shearing is represented by stress path A’ B’. Now we know that A’ B’
from the module 3, we know it is at an inclination of 3. So, ESP = TSP because it is drained. So,
this is ESP = TSP for CD that is consolidated drained with a slope of 3 in q-p p’ plot.
So, this is what is represented, this is A’ B’ is the effective stress path or total stress path for CD.
So, here because of CD it is same and it is at an inclination of 3, how will it look like in v ln p’?
So, this is the representation which we want, so A’ will be on ICL or NCL. Now from here, we
know that v changes or v reduces due to compression. So, B’ is the failure point which is there
on the critical state line, so this is the point where it fails.
And that is denoted as A’’ and B’’, the homologous points of A’ and B’ on this plot is A’’ and
B’’. So, this is the path or this is the change in the state of the soil towards failure, so for A’’ to
B’’. Now consider initial u is zero, there is no initial pore water pressure. So, initial condition of
CU is same as CD, so there is no change in the point A’, so A = A’.
Slope of ESP of CU depends on A f, in the previous plot we did not have much of discussion
because it is direct reduction in σ’ when you have pore water pressure. But in this case we need
to determine the slope of effective stress path for a undrained test. And that also we know 3 / 1 -
3 A f is the slope. So, we know that it depends on pore water pressure parameter A f the slope.
And in the case of normally consolidated, we know that mostly it moves towards the left and
why? Because A is close to it is a positive value. So, ESP of CU will be towards left for normally
consolidated, this is the way in which it moves and this is C’. You might be thinking why we
568
have to draw with this lower slope, why not here, why not here? In fact both these points should
match.
Now we know since this is a undrained test, there will be no change in v, so we have to first
draw the curve here where this will be horizontal. So, this point and this point has to match, and
that is why it is moving in this direction. So, it A’ C’ is the effective stress path for undrained
shearing. And this is the point where it touches for consolidated undrained test; our total stress
path is valid only up to this particular point.
So, this effective stress path for CU and this is the total stress path for CU, so the same line,
because there is no initial pore water pressure. So, this C’ comes and this is horizontal and v
remains constant, that is the C’’ point.
In CD, v reduces due to compression during shearing for NC, so this is the reduction. A’ B’
moves down and meets critical state line, so this is the one which we have already explained. In
CU, v is constant during shearing for normally consolidated, so this is not changing, so this is the
point where it fails. In CU, Δ u excess is positive which we have already seen. Now if a series of
such CU and CD test are performed with different p’c, that is the initial effective mean stress.
569
Then the stress path will follow similar pattern, so this is the pattern or the trend in which the
TSP and ESP would move. So, this is the ESP for drained and this is the ESP for undrained, so
this you can see that for different combination of p’c1, p’c2 and p’c3. We can see that there are
different the same kind of response and this is the same kind of effective stress path for drain. So,
you will have a series of such stress paths.
Now let us see 2D analysis of OC soil subjected to CD in v, q, p’ plot. The OC soil exhibit both
peak and critical state, both response need to be accounted in CSSM. Soil is initially isotropically
normally consolidated, and then unload it to p’c, that is the point D’ is heavily over consolidated,
let us this is the point D’. So, what does it mean, it means that it is loaded, so this is the peak line,
so what it means is that soil is initially isotropically normally consolidated.
So, it is normally consolidated then it is unloaded to point p’c that is the initial effective
volumetric stress. So, point D’ is the heavily over consolidated, this is the point D’ and the peak
line is represented by U’ H’. So, we know that we need this particular point over here, why?
Because it has to reach there and then come back to critical state line. ESP = TSP for CD with a
slope of 3 in q p’ plot, now that is not going to change.
ESP first touches the peak line D’ E’ undergoes yielding, so this we have already discussed, so I
am not going to spend more time here. It reaches E’ and then ESP returns to critical state line
570
where it fails and that is at the same slope of 3. So, it comes back and fails at critical state. So,
this is ESP which is same as TSP for consolidated drained test, now this is the point D, D’’.
During shearing v first reduces, now I will tell you like from here it is isotropically consolidated,
it reaches somewhere here and then it is unloaded. So, there will be an unloading line on which
the point D’’ will fall. So, now we know that when it is sheared the v first reduces because there
will be some initial compression which is shown by this response. And once the initial
compression is over then in a drained test there will be increase in this specific volume.
So, this is the point E’ somewhere here, so it has already started yielding. So, then once it is
started dilating then it will be in the upward direction, v will start increasing. So, on further
shearing, v increases due to dilation and fails at CSL the point F’’. So, that F’’ is the point
towards which it fails, so v initially reduces then increases. So, whatever we have learned in the
module 2 it is all mapped onto a critical state framework.
2D analysis of OC soil subjected to CU, now we will see the undrained response. Again peak
and CS need to be accounted, first it is normally consolidated, unloaded the same starting point.
Peak line U’ H’, ESP and TSP are different for CU, it is no longer same, so this we need to keep
in mind. TSP same as consolidated drained, so this is will be the total stress path. And ESP of
571
CU depends on heavily over consolidated soil and the value of A f, and mostly A f will be
negative, so we know it will move towards right for heavily over consolidated soil.
ESP touches peak line, that is D’ G’, and then further shearing u tends to negative and meets
critical state line at I’, so this is the way in which it will move. So, this much detailing we have
not done in our earlier lecture. So, in the critical state framework we understand this clearly how
the shearing happens. So, this is the effective stress path for consolidated undrained, and this is
the total stress path.
So, u is positive for point G’, now this is the point which I was trying to make, now this is the
total stress path, this point G’ is towards the left. Now relative position of this keeps changing
depending upon the state of the soil and the soil type. And towards the end it comes towards the
right of the total stress, so this is because the pore water pressure is negative for I’.
Now again v is constant for CU, so D’ we know, now it is very easy to plot in what direction the
state of the soil would change, so it has to be horizontal. Peak point is given by G’, it has to be in
the same line, there is no because v remains constant in CU. Failure point on CSL is I’, so it is a
horizontal line and this is denoted by I’’.
572
Now 2D analysis of OC soil, it is again now we need to discuss about an important aspect. The
soil sample is loaded to a particular point on the NC and then we unload it, now how long you
can unload because the soil is subjected to a kind of swelling. So, there will be a limit towards
the left side of the critical state framework. The right side boundary we know, it is normally
consolidated line or isotropic consolidation line.
And there should be a boundary towards left otherwise what does it mean, towards left means it
is more tensile because it is minus. So, you are minusing the initial effective stress, so it keeps
reducing. And hence that reduction will be curtailed by some boundary, so that is what we are
going to learn now that is defining the left side boundary towards this area. So, soil is unloaded
from a given NC, so this is the unloading line.
During unloading there is a limit beyond which OC state cannot sustain, this is due to the tensile
stress that is imposed in the soil. Point A let us say it is a low tensile zone; here the possibility of
tensile stress is less because it is close towards the normally consolidated line. Point B is a
medium tensile zone and point C is a high tensile zone, this is a possibility of high tensile zone in
point C. Now soil cannot sustain tensile stress and failure would occur.
This we know soil is very, very weak in tension, because these are granular. Line OU’ with slope
3, limit the soil to non tensile failure. Now this is the point OU’, it is slope 3 similar to this total
573
stress path, so it is a slope 3 and this is the left most boundary. And that towards this direction
this will be a non tensile failure zone, and this particular zone will be a tensile failure zone. Left
of line OU’ that is towards this is the tension failure zone.
So, whatever has been marked here is tension failure zone, and the failure stresses would become
negative. Now please keep in mind this particular sentence, we will know this better in our
subsequent lectures where we will show that beyond this slope. Now this is at a slope of 3 in q p’
representation, this slope considered as 3 beyond which it will be a tension failure zone. Why it
becomes a tension failure zone?
Because the failure stresses tends to become negative beyond this slope of 3, as it increase the
slope beyond 3 the failure stresses will be negative. We will be showing this in our subsequent
lecture. So, this marks the boundary or the tension cut off or that is the region of left boundary
below which soil cannot exist. So, soil state is not possible in tension failure zone.
Now let us summarize this by saying critical state: the ultimate state. Critical state limits all
possible states of soil irrespective of stress history, drainage condition and loading condition, this
is what we have seen, let us put it all together in this slide. All ESPs finally reached critical state
line which is unique for a particular soil. So, let us start with a lightly over consolidated state of
the soil, till now we were discussing mostly about NC or HOC.
574
For this example let us consider lightly over consolidated state. Let us say A is the initial
condition let us say this is A, now you can see that it is towards the left of NCL and hence this
state is LOC. AB which represents UU test, that means ESP touches critical state at B at constant
v. Now let us say A is the starting point and we know we have conducted UU test, in UU test
there is no consolidation and it is an undrained shearing directly.
Now when you do undrained shearing we know that v is not going to change, it remains
constant. And hence it is very easy to draw the state of soil how it changes. So, this is the way in
which it changes, it is horizontally towards critical state line that is the point B. And the same is
mapped on to the critical state line in q p’ plot; this plot is needed only at failure. So, B is a
failure point that is mapped onto q p’ plot.
AC: let us say we are consolidating it first before shearing, so that consolidation phase means v
reduces and this is represented by AC. Now we do not have to map this here because C is not a
failure point. CD: indicates consolidated undrained shearing and v is constant, now from point C,
it is very easy, it is horizontal towards the critical state line and that is the point D. This CD is
not consolidated drained, it is the point CD, so it represents consolidated undrained shearing.
So, that is lying on the critical state line, so it is mapped to critical state line. CE: is consolidated
drained shearing, v reduces due to compression. Now from here C, it is sheared in a drained
manner, consolidated drain shearing but it is consolidated to same point C, so how it will move?
It will further reduce in v and it moves down and that is the point CE, where it indicates the
consolidated drained test. And this is on the critical state line, so it has to be mapped onto critical
state line.
ACD, represents that is A, C, D represents the entire CU test, so A, C and D represents the entire
cycle of the consolidated undrained test. And similarly ACE is the test for consolidated drained
one, so entire consolidated drain test. For normally consolidated, the trends are same except that
A and C will be on NCL. Now we have discussed for lightly over consolidated state, now if I say
575
that it is normally consolidated only thing is this will shift towards this point, and all other trends
remaining the same.
Now let us see for heavily over consolidated state, F is the initial condition same v as that of A.
Now we have the same specific volume but the point is the stress state or the stress history is
different, it is heavily over consolidated. From here UU test, you can see that this is the way and
it fails exactly at the point B because the initial specific volume is same. Now these are some of
the facts which we have not seen in our previous lectures, and which is very legitimately clear in
this particular lecture.
FG is the consolidation phase, so it is coming down, then this only consolidation. GH is

consolidated undrained shearing, so v is constant. So, that represents GH and H has to be
mapped down. Then GI is consolidated drained shearing, now in the case of HOC understand v
will be increasing because of dilation. And that is the point I where it meets the critical state, it is
mapped down I. So, FGH, FGH is the entire CU test and FGI is the entire CD test.
Now you can compare points A and F, soil state have same v and different stress history. So, just
because of different stress history means it does not mean that it will have different failure point
because it has same v. So, what is that initial state of v is playing a role, so for undrained
576
shearing failure stress p’f at B are same for both the stress history A and F. Normally
consolidated and LOC approach CSL from right side, so that is the wet side.
So, whatever be the NC and LOC will approach critical state from right towards left, that is from
the wet side, and HOC from left side that is the dry side. Now the summary is whatever be the
initial state stress history and loading condition, the final state approaches the ultimate state
which is the critical state, you can see it here. Whatever be the final state can all be mapped on
this critical state line in q p’, where the failure occurs.
We have discussed about various initial conditions for both OC and NC but then all can be
mapped onto critical states, so that is what it means critical state the ultimate state. So, let us try
to summarize today's lecture.
The 2D frame work of critical stage soil mechanics in terms of e τ σ’ and v, q, p’ is discussed.
Normally consolidated and lightly over consolidated soil approach critical state line from right
side which is the wet side. HOC approach critical state line from left side which is the dry side.
In drain test e changes and σ’ p remains constant, in heavily over consolidated soil reaches peak
state before reaching critical state, that is yielding has to happen before failure.
577
The concept of yielding becomes important, and that is the reason why we will be discussing
yielding of the soil in our subsequent lectures. In undrained test intersection of unloading line
with critical over consolidation ratio line correspond to Δ u = 0. Now this in some cases it is
considered as critical state line as well. Δ u is positive for initial state right side of critical or over
consolidation ratio line due to compression during shearing, we have discussed this, whatever
towards the right of COCRL will have positive view during shearing.
The same is negative when it is towards the left of COCRL due to the tendency to dilate. Please
do not read this as due to dilation, because dilation represents drain behaviour. So, whichever
soil which has got the tendency to dilate will have negative pore water pressure. Similar to
normal consolidation line, tension failure zone forms the left side boundary of the soil state to
excess. NCL becomes the right side boundary similarly we have a tension failure zone which
defines the left side boundary.
Irrespective of the initial state, stress history, loading condition, the final state approaches the
ultimate state that is the critical state. So, that is all about two dimensional representation of
critical state. Now we will discuss a few more aspects of critical state in the subsequent lectures
followed by soil yielding, that is all for now, thank you.
578
Lecture-47
Peak State
A warm welcome to all of you, in the past few lectures we were discussing about critical state
soil mechanics. We have already seen the basics, we have also seen the two dimensional
representation in the last lecture. While discussing the two dimensional representation, you might
have noticed that we discussed about the right most boundary for the soil to exist and that was
isotropic consolidation line or normally consolidated line which forms the right boundary.
In the last lecture we have also seen a tension cut off zone which appears on the left side beyond
which the soil cannot sustain. And bottom part we have the axis itself, where shear stress is equal
to 0, where q or the deviatoric stress is equal to 0 that forms the bottom boundary. But what is
the top boundary? One can conveniently say that it can be a critical state line; you remember
there is a line in q p’ or τ σ’ plot.
And that can be considered as the boundary provided why? Because the soil ultimately fails on
it, but in fact for those soils which dilates, we know that it has to first hit the peak and then come
back to the critical state line. So, the top proportion, if you want to really consider the boundary
it has to be the peak state. Because peak state is above the critical state even though the final
failure happens at critical state.
So, today's lecture we will see more specifically about peak state which happens to be another
boundary for the soil to exist. And finally what we are trying to do? Through critical state
framework; we are just trying to understand the bounds within which the soil can be loaded that
is where the soil can actually exist during shearing.
579
So, with this we will now start with what is known as peak state? Now let us consider 2 reference
void ratios e 1 and e 2. Let us say we are considering e σ’ plot, and the critical state line is marked
in this. Since we are considering the peak state, we do not have to really bother about normally
consolidation line. Because we know that peak state is relevant to over consolidated or dense
sand and which lies on the wet side of the critical state, mostly.
So, we will discuss about e σ’, now there are different ways of representation all majority of
them we have already discussed, it can be on log σ’ plot as well. But here specifically it is on a
linear scale. Now this is the reference void ratio e 1 and e 2, it is noted you can note that e 2 is less
than e 1 that means this is a more dense state. Now shear strain test is conducted at identical e
and different confining stress, and that is what we do.
If you want to determine the strength of the soil, we need to have at least 3 identical samples.
And these identical samples means, the initial state has to be comparable for all the 3 states. And
that means that the void ratio has to be same or the water content has to be same, the initial state.
And then we choose 3 different confining pressure, whether it is in direct shear test or whether it
is in triaxial test.
We need to conduct it at 3 different confining stress, that is what is written here, at least 3, so
three confining stress. So, this is what at e 2, these are the locations at which or these are the σ’ at
580
which the shear test is performed and the last point being on the critical state line. So, let this be
PQ and for e 1 the same void ratio but with different effective stress or the confining stress, so it
can be denoted as RS.
So, there is a PQ series and there is an RS series. Now different peak lines corresponding to e 1
and e 2. I hope you remember or rather you can refer back to module 2 wherein we have
discussed this particular aspect where for a peak you can have three different points and there
will be a peak line. And this peak line when it is fitted using Coulomb's equation will give the
cohesion and peak friction angle.
So, let us try to do the same thing here τ versus σ’, this is the critical state line. And for e2 that is
at a smaller void ratio we get the 4 points here, the last point being on the critical state line. So,
the peak line is PQ which meets the critical state line at point Q. Now if we extend this or rather
if we fit a Mohr Coulomb or a Coulomb's envelope we end up with a cohesion value and the
inclination will be peak friction angle, similarly for RS.
Now we can note that void ratio e 1 is greater than e 2 obviously the strength of e 1 will be less
than PQ, and that is very legibly shown in this particular figure. So, that gives C e 1, so cohesion
for void ratio series e1 is C e 1 and cohesion for void ratio series e 2 is C e 2. Peak state lines meet
critical state line at S and Q; we have also discussed this aspect. As we go on increasing the
effective stress it reaches a point where the effective stress suppresses all the dilation.
At that point the peak state line becomes same as the critical state line, this aspect we discussed.
But now there is a lot more clarity in those discussions, why? Because when we integrate the
explanation with respect to e σ’ plot, this gets quite comfortable and it is very easy to understand.
Because we know that as you go on increasing σ’, there exists a condition where it will meet the
critical state line, when you project that down that will give Q and S.
At this point onwards the dilation is suppressed due to higher σ’, this aspect we have clearly
discussed. Now from this point onwards it will be on the critical state line. Peak state lines are
terminated at P and R at low σ’. Now you can see here that it does not extend beyond this,
581
because it is very difficult to find out or to simulate a condition of low stress lower than this
particular point P or R, so it gets terminated, peak lines are terminated at P and R at low σ’.
Now beyond this it does not follow the Coulomb's equation τ = c’ + σ’ f tan φ’ at low σ’. Now I
do not know whether you will be able to really appreciate this particular point or not. Because we
never bothered about this particular aspect, when we conducted the test may be for finding out
peak friction angle and cohesion. So, cohesion would exist only if we are dealing in terms of
peak friction, so both goes hand in hand.
Now when we discuss this or when we have already determined this in the lab, we have got 3
points and we have just fitted. How low stress value we have considered, we have not bothered
about it. In fact the authenticity of the stress points towards the lower end is restricted because of
certain factors. One important factor is the tension cut off zone and the other one is what we are
going to discuss now.
So, this particular equation for the time being is more like a mathematical fitting parameters.
How sound or how much physically relevant these parameters are? Definitely φ’ has got it is
own physical significance. But then whether c’ is always justifiable or not, that is what we will
just see now. Peak line is equally valid for clays and sands, this also we know because when we
say clays we talk in terms of over consolidated or normally consolidated.
And if it is over consolidated we say that there will be dilation, there will be peak and there will
be c’ which is associated. And in the case of sand we say that for dense condition it will be
exhibiting a peak friction angle. But what about the condition close to 0 confinement, can it have
cohesion? Definitely not, so that is the reason that is another reason why we say that in general
the Coulomb's envelope or Coulomb's equation may not be valid for lower stress condition.
Sands do not have shear strength at 0 σ’, at 0 confinement sands do not have any shear strength.
So, in general cohesion C’ cannot be considered as shear strength available at 0 σ’. So, we
cannot generalize stating that σ’ is always the shear strength at 0 confinement, because that is not
valid for at least sandy kind of soils. So, where it will be applicable?
582
C’ is primarily a mathematical requirement of Coulomb's model and it is valid for soils with
cementation characteristics. Definitely one can always say that those soil which exhibits a kind
of cementation characteristics, it can hold on it is own even if there is no confinement. So, in
those cases and that can be also due to apparent cohesion. An apparent question may not be
available all the time, so that also we need to be careful about.
So, we are basically discussing about in depth the understanding of peak and the cohesion
behavior of the soil. Definitely if there is a kind of cementation, it is bound to exhibit cohesion,
and for that particular reason we can always extrapolate the results. So, that intercept can be
considered as cohesion.
We have seen that different e 1 and e 2 will result in different C and φ’ characteristics for peak
line. Now let us try to make it unified by the process of normalization, in the last lecture we have
already seen how to normalize a given plot. So, the normalizing parameter is σ’ c that is the stress
condition or the confinement which corresponds to the critical state line. So, σ’ c is the
normalizing parameter and in this case there will be σ’ c 1 and σ’ c 2.
So, σ’ c 1 for e 1 series and σ’ c 2 for e 2 series, this will be the normalizing parameter. And what
we will normalize? We will normalize τ and σ’, so the normalized parameters are τ / σ’ c, σ’ / σ’
583
c. We have done this in the previous lecture and we have seen that critical state line and normal
consolidated line comes to a particular point.
The line is represented by a point how this is so? Please refer back to those lectures. So, this is
the normalized plot. Here different peak state lines for different e, it reduces to a single line PL,
do not think that this is a critical state line, this is done with a specific purpose, that is why it is
drawn in dashed line, this dashed line is not critical state line. Critical state line is represented by
this particular point and this line is only meant to show the slope.
So, PL becomes the peak line which is normalized, now the effect of e 1 and e 2 goes away
whatever we have got as PQ and RS, it all merges to PL. All the data points will be on this line
because of the normalization. We have seen that this is 1 and NCL is a point on the x axis. So,
this parameter c’ p is a normalized parameter and φ’ p is the inclination of the normalized plot
rather φ’ p will be same as this plot as well, because it is only a division by σ’ c.
So, L is the critical state point where it merges on the critical state line. Now we can always
write the equation of the peak line PL which is the normalized plot as τ / σ’. Now I have added τ
p here just to make sure that it is the point on the peak line PL. So, τ p / σ’ c = c’ p + σ’ p / σ’ c
tan φ’ p. In which C’ p can be written as C e / σ’ c, C e is the cohesion corresponding to a given
void ratio, so it can be C e1 or C e2.
Now that will be getting reduced by or that will be getting factored by σ’ c, so C e 1 upon σ’1 C
or C e 2 upon σ’ C 2, whatever. So, that normalized value of cohesion is c’ p as you can note
from this figure. We can also write c’ p = tan φ’ cs - tan φ’ p, how this is possible? Now this
expression is possible only on the normalized plot, why?
Let us consider tan φ’ cs, this is the triangle, inclination tan φ’ cs, so tan φ’ cs is equal to let us take
this distance as y upon 1, because this is 1. And similarly if you consider this particular triangle,
we can write tan φ’ p, let us call this as y 1 and this will be y 2 / 1, so y 1 = tan φ’ cs y 2 = tan φ’ p.
So, in short we what we are writing is c’ p = y 1 - y 2, that means that is tan φ’ cs - tan φ’ p.
584
So, substituting this in this particular equation, we can write τ p / σ’ c = tan φ’ cs - tan φ’ p + the
same σ’ p / σ’ c tan φ’ p. So, now the parameters c’ p that is the normalized parameter, φ’ p φ’ cs is
independent of e or w which is the initial condition. Because now we have normalized it and
hence the effect of e1 e2 or it can be told in terms of water content for a saturated sample, the
effect goes away, so it is normalized now.
So, this parameter it is independent of the initial condition that is e. Hence these parameters can
be considered as fundamental parameters, only after normalization. But there is a catch here, can
it be really considered as a fundamental parameter under certain condition? If we consider
another parameter that is the confining stress, then you will see that the things will be different
and this we will explain in the slides to come.
For the time being if you are referring only to the initial state, that is the void ratio or w, we can
consider this to be still a fundamental parameter after normalization.
Now let us try to repeat this in v, q, p’ plot, because this is important. Finally we will be
explaining all these aspects on v, q, p’ plot. So, I am not going to spend much time, it is the same
explanation, we have a reference v for which four tests has been conducted PQ. And the same
has been shown on q p’ plot, where M c is the slope and G p is the intercept on q axis, p’ c
corresponds to critical state.
585
Here we can also add the tension cut off zone with an inclination of 3, let this slope be H p which
is given as tan φ’ p, where φ’ p is the inclination of PQ line. Equation of line PQ corresponding to
a particular v, please note it is for a specific volume q p, that is q on peak line p q = G p + H p into
p’ p, p’ p means p’ on the peak line PQ, so this is the equation and that is very state forward.
And if we normalize using p’ c which is the point on the critical state line, we get the normalized
plot where H p = tan φ’ p that remains same G pv is the normalized G parameter with respect to
specific volume v, this is 1 again. Equation of normalized peak line can be written, this is normal
consolidated line point q p / p’ c =G pv which is from here + H p into p’ p upon p’ c, a same
exercise as we have done before.
And this G pv = G p / p’ c, G p v can also be written as M c - H p, where M c is the slope of this line
or this particular line. And here this distance as we have done in the previous case it will be M c -
H p. So, q p / p’ c from here is M c - H p + H p * p’ p upon p’ c. As like before parameters G pv that
is normalized parameter M c, H p are fundamental parameters independent of e or w, it is
fundamental parameter.
Only because it is not dependent on initial state, it is independent of the state variable e or w. As
I told there is another hidden part here which may violate this fundamental nature that we will
see a bit later. The contribution of e or water gets included in p’ c in this particular parameter.
The effect of e and w is inclusive in the normalizing parameter p’ c that is why it gets
normalized.
586
Now let us see something what is known as power law equation of peak state. Now we have
represented the linear representation of peak line in a normalized parameter. We would also like
to see if we actually consider the real behaviour, how it would look like? When I say real
behaviour what I mean is we have told that peak line cannot be extended beyond a certain point
on the left side or towards the low confinement stress region.
So, if that is the case, is there a better model to represent peak state which is even more general
than the linear representation. Granular soils which are not cemented will have zero strength at
zero effective mean stress. Now this aspect is not considered in the linear representation for peak
state, so that is very important. So, when we represent the peak state using a linear fitting it is
bound to give an intercept, and this is not valid for soils which do not have cementation
characteristics. So, let us see this in detail.
It is apparent that the peak line should pass through origin for this particular case, where there is
zero strength then it has to pass through the origin where q p = p’, that q p means q on the peak
line, p’ p, p’ on the peak line that will be equal to 0. So, it has to move through the origin. We
also know that on the right hand side peak line would also meet critical state line, it cannot go on
and on, so it will meet at critical state line where the dilation gets suppressed.
587
And the peak behaviour is no more possible, a peak gets suppressed, when that is q c = M c p’ c
when where it meets the critical state line. Now this could be achieved only using a curved line,
if we want to meet this condition and this condition, it has to be a curved line. Now please
understand, we are slightly breaking certain rules which we felt it is actually the case when we
studied during our undergraduate.
So, as we move forward we have to add more and more details to what we have already learned.
And in the process we may have to break certain rules with which we have understood the shear
strength concept, and this is one. The speak behaviour is very, very critical and which we have
not discussed in detail during our undergraduate days. So, this could be achieved using a curved
line, because for soils we do not discuss about this curved failure envelope to represent peak
behaviour.
This is very much dealt when we discuss the behaviour for rocks, but the same is not considered
when we discuss for soils. So, this curved line a very simple equation can be of that of power
law. If we represent the peak state points using a curve like this which is a power curve, then this
will satisfy that it goes through the origin and it meets the critical state line, so both the points are
actually met.
And that can be represented as q p = M into p’ b, where M and b are the power equation
parameters. In log we can also write this as log q p = log M + b log p’, and that can be
represented as a straight line because then it becomes on a log it becomes a straight line, where
this log m b and in the normalized plot. We can write q / p’ c p / p’ c, we can write the power
curve in this particular manner m’c.
And this particular intercept will be equal to M c because the slope is M c that we have seen since
this is 1, this distance becomes M c. So, we can write the equation for this curve as q p / p’ c = M
c p’ p / p’ c raise to β. So, the same equation altered in the normalized plot, where M c and β are
soil parameters and depends only on the soil type.
588
Now let us make a comparison. Now we have discussed the various nuances associated with
peak state which we have not studied earlier. Now what is the implication of this, what is the
implication of the more generalized representation of peak state using power law curve? vice-a-
vice what we already know which is the Mohr Coulomb failure envelope or we can also call it as
Coulomb’s failure envelope.
Because we are representing in terms of τ and σ’, so how do we make this comparison? Let us
see, for this I am considering 2 cases of peak shear strength for three identical samples in
normalized plot. So, now we will consider only normalized plot because we have discussed in
detail about how e and w is going to affect, the initial state how it is going to influence.
Now the other aspect which is the role of confining stress that is what we are going to see in this
particular slide. Now for that what we are doing is, we are considering three identical samples
the way we use it for shear strength determination. And we are representing it in normalized plot
and there are 2 cases which we need to consider. Now the first case, case 1 on the normalized
plot, this is the power curve which has the more general representation of the peak state.
Let us consider these three identical samples are a, b and c, now why this is case 1? In case 1 the
three identical samples are close to critical state line that is towards the critical state line or the
right most portion of the peak curve. Now in normal practice what we do, we fit a best fit line
589
and we call it as Mohr Coulomb failure envelope or Coulomb's envelope and we get the cohesion
and the friction angle, that is φ’ p1, because it is case 1.
So, the peak friction angle is φ’ p 1 and the cohesion given is c’ p 1 and 1 represents that it is case
1. The power curve passes through the data, that is best fit and the critical state point. So, the
power curve passes through even origin through the data and the critical state point. Now we are
best fitting a linear envelope or Mohr Coulomb envelope to the data, so what we see?
In case 1, the data points are close to critical state and this is the best split curve, and we can see
that it gives a cohesion and friction. Now what is case 2? Now case 2, the data points are towards
origin, that is towards this or lower σ’. So, this is case 2, same power curve, same identical state
but we have chosen d, e, f that is the three normal stress or effective stress towards left side or
towards origin, now what we do?
We have to again fit the Mohr Coulomb failure envelope, so this is represented as MC2
envelope. And given by the friction angle φ’ p 2 and c’ p 2, so this is what we do for peak
envelope. Now we can see, we can compare the power curve which is general in both the cases
case 1 and case 2, whereas MC1 and MC2 there are marked differences between case 1 and 2.
So, MC1 and MC2 do not pass through the critical state point and origin.
So, for peak we know that it has to pass through origin, because to satisfy the lower end and it
has to pass through the critical state to meet to satisfy the upper end. Now both the Mohr
Coulomb failure envelope best fit line will not pass through these 2 points. So, you can see that
the Mohr Coulomb is a kind of violation for representing the peak behaviour. But still it is valid,
we will see how?
Now not passing through origin is not applicable for cohesionless materials and those soils with
only apparent cohesion. So, what is the implication of not passing through origin? If it is Mohr
Coulomb is not passing through origin means, it is not applicable for a cohesionless situation or
for the soil where there is no proper cementation it cannot hold on it is own, because, in apparent
cohesion we cannot consider it as a true cohesion.
590
So, those soils which cannot sustain on it is own, it may not be able to represent the lower stress
conditions, this is applicable only for true cementation in soils. Not passing through critical state
violates the fact that the soil keeps dilating even at high σ’. So, the contribution of suppression
by σ’ is not accounted, if the Mohr Coulomb is not passing through the critical state point.
So, these are some of the violations which we have to keep in mind when we use Mohr Coulomb
failure envelope for representing peak behaviour.
So, both MC1 and MC2 best fit and it adequately represent the data points; there is nothing
wrong in representing this particular line. Because it adequately represents this particular range
of data, it also represent this particular range of data, but what matters? The range of σ’
considered matters that is why MC1 is practically different from MC2. The parameters provided
by MC1 and MC2 are different, c’ p1 is greater than c’ p2 and φ’ p1 less than φ’ p2.
Now c’ p and φ’ p1 cannot be considered soil parameters if the range of σ’ is considered. Now we
are just going against what we have told in the previous slide where these normalized parameters
c’ p and φ’ p can be considered as a fundamental parameter. Provided, we are referring only to
the initial state, then it is correct, but if we are considering the range of σ’ as you can see here,
this is one particular range and this is another particular range.
591
If the range of determination of σ’ comes into picture then you can see that both c’ and φ’ peak
changes. So, it is dependent on the range of σ’ for measuring peak strength. So, MC envelope
parameters are widely used in practice because we never look at the power curve, what we are
interested in is the representation using Coulomb's envelope. All the numerical evaluation even
all the programs that we use it is all basically based on Mohr Coulomb failure envelope.
So, for uncemented soils, curved power law passing through origin and critical state is more
applicable or it is more general in nature than Mohr Coulomb failure envelope. Peak strength
measured for wide range of σ’ is close to power curve. Now if we include different range or
during the determination then that particular Mohr Coulomb envelope will be a more generalized
and close to what we represent using peak.
Now we can see another important aspect, Mohr Coulomb parameters are unconservative and
overestimate the strength for the range of σ’, please understand this, beyond which peak strength
is determined. Let us take the example here, now for these three data points it fits very well, so if
this is the range considered nothing wrong it goes hand in hand with the power curve. But the
moment it exceeds the range that is towards this side and towards this side we can see that Mohr
Coulomb failure envelope is above the power curve.
If this is the actual strength available the predicted strength based on Mohr Coulomb towards the
lower end is this, similarly in the upper end it is this. So, it is an overestimation, and whenever
we overestimate strength it becomes more unconservative. Whether that much strength is
available or not, that is a kind of debatable subject. But then according to the mathematical
framework what we need to understand is that the Mohr Coulomb beyond the range for which
the strength is determined.
Now it is same for this also, this is the range, so beyond this there is an overestimation of
strength by Mohr Coulomb failure envelope. We are basically working with higher strength, for
a designer this is not good. So, for the range for which the σ’ is determine, if the same
592
confinement is there in the field then whatever we are telling is correct. So, that is why we have
not discussed these aspects in detail during our UG learning.
There it is presumed that whatever peak strength values that we are determining, it corresponds
to the given confinement in the field. So, σ’ that we use for strength determination, it adequately
represents what is there in the field. Hence there is no dispute, otherwise there are a lot of
loopholes when we consider Mohr Coulomb failure envelope, which we need to keep in mind at
least we should be knowing and we should appreciate this better.
Another aspect which we need to discuss is the stress dilatancy and Mohr Coulomb envelope.
Let us consider four peak state points B 1, B 2, C 1, C 2 that is e corresponding to e1 we have A 1
and e2 which is A 2 which is the point on the critical state line. Corresponding to σ’ b we have B
1 and B 2. And corresponding to another reference effective stress σ’ c we have C 1 and C 2 and B
c is the point on the critical state line and CC is the point on the critical state line.
The same is shown all the critical state points are mapped that is B, C, BC, CC, A 1 and A 2.
Now, this four points represents many different aspects B 1 and C 1 same e 1, B 2 C 2 same e 2,
where e 2 is less than e 1. Now the points B 2 and B 1 corresponds to same initial effective stress,
that is σ’ b, but e1 is greater than e2. So, this and the over consolidation ratio will be different for
B 1 and B 2, the amount by which B 1 is over consolidated or B 2 is over consolidate it differs.
593
So, there this four points represents adequately different aspects of stress dilatancy. So, let us
consider peak which is represented by B 1 C 1, now this has same e1. So, this B 1 and C 1 will lie
on the same peak line and B 2 and C 2 will lie on the other peak line. And you can see that e1 is
greater hence the strength is lower. And the corresponding cohesion is c’ p1 and c’ p2. Now let us
consider there is a given line here which is drawn to show Ψ’ p, what is Ψ’ p?
You can refer back to module 2, it is the dilation angle which is over and above the critical state
angle, and summation of these 2 gives the peak friction angle. So, φ’p, if this is the peak line φ’ p
is the critical state plus the dilation. Now B 1 C 1 have same e1 and lie on Mohr Coulomb failure
envelope which gives c’ p1 and φ’ p, we have already seen that. B 2, C 2 have same e2 and it lie on
Mohr Coulomb failure envelope which gives c’ p2 and φ’ p.
B 1, B 2 have same σ’ b, now these 2 points we are referring it has same initial stress condition.
But B 2 is more OC than B 1, you can see that it goes down and has come so this much allowance
is there. So, B 2 is more OC than B 1, and e2 is less than e1, so these are the 2 important
conditions. Now if you refer to this B 2 is more OC than B 1 means if we draw a line through B 2,
you can see that the dilation angle is more for B 2 which is a good indication that, it is more over
consolidated.
If it is more over consolidated, it is bound to exhibit more dilation towards the lower stress
region. So, the dilation caused in B 2 will definitely be more than B 1 and hence the dilation angle
is more. And if you draw this line through origin you can see that Ψ’ p for B 2 is more than Ψ’ b
at B 1 and obviously the e2 is less than e1. B 1 and B 2 would reach CSL at BC, that is already
there, now B 2 have to dilate more than B 1.
Accordingly B 2 dilates more, now even though the initial stress condition is same; it does not
mean that the dilation is going to be same. We also need to consider how much over consolidated
the given sample is? Definitely B 2 is has to dilate more and hence the dilation angle will be
more, so B 2 have higher dilation angle Ψ’ p than B 1. Now let us consider B 1 and C 2, that is B 1
and C 2.
594
Now this point B 1 and C 2 has got same over consolidation ratio. Now if you see B 1 from here
and C 2 these are same over consolidation or the over consolidation ratio at B 1 and C 2 are same.
And that is very clear from here, because this particular line passes through B 1 and C 2, and the
dilation angle is Ψ’ p. So, both B 1 and C 2 has the same dilation angle, now this is possible only
if the over consolidation for B 1 and C 2 is same.
And that is why we say that the B 1 and C 2 has same OCR, but it has different e and different σ’.
Earlier case we discussed for the same σ’, now here we have different e and different σ’ but OCR
remains same. Now B 1 and C 2 has same Ψ’ p, which can be represented by τ / σ’ = tan φ’ cs + Ψ’
p. Because this τ / σ’ that ratio remains same for both B 1 and C 2 because the dilation angle Ψ’ p
is same for both.
What it indicates is that for peak points both the stress dilatancy also need to be considered and it
is important to specify e, that is the initial state parameter e or it can be in terms of w. The over
consolidation ratio as we have shown here, dilation Ψ’ p, τ and σ’ for analyzing and interpreting
the test results related to peak soil behaviour. So, interpreting peak is slightly more complex than
what we generally do for critical state.
So, we need to keep in mind these parameters, not to confuse, but to understand that these
parameters also significantly contribute towards the peak characteristics. So, that is why in
certain design we see that the cohesion characteristics is ignored, why because we do not know
whether it is going to mobilize that much of friction actually in the field. So, the safe way is to
nullify cohesion and only consider the frictional characteristics.
So, here what it means is that, it is not only the over consolidation or a single parameter that need
to be considered while interpreting the peak behaviour. We need to take into consideration, the
state parameters, OCR, this what range of σ’ we are talking about and so on.
595
So, let us summarize today's lecture. Peak state or peak line is one of the boundary state for the
soil. So, that is where we started off with NCL forms right boundary, tension cut off zone on the
left bottom we have the axis, and the top instead of critical state we have the peak line as the
boundary. Soil samples with identical e and varying confining stress σ’ gives a specific peak line
with φ’ p and c’ p.
Peak state lines are terminated at low σ’ and does not follow the Mohr Coulomb failure line, we
have shown this. c’ cannot be always considered as shear strength available at zero σ’. The
normalized parameters τ / σ’ c and σ’ / σ’ c takes away the effect of initial e and because it is
inclusive in σ’ c. In the normalized plot parameters c’p, φ’ p, G pv, M c, H p is considered as
fundamental soil parameters provided the range of σ’ is fixed.
These parameters are independent of e, v or w. Every soil state above critical state line in τ σ’
plot correspond to a specific e and OCR, that is Ψ’ p. Equation of peak line is derived, a power
lock curve satisfies the requirement of zero strength at zero effective mean stress and meeting the
critical state line at high σ’, that we have shown, because this is the most adequate representation
and the more generalized aspect.
596
The widely used Mohr Coulomb envelope parameters cannot be considered general for
uncemented soils. c’ p and φ’ p1 or rather φ’ p cannot be considered soil parameters if the range of
σ’ is considered. Power law curve is more general than Mohr Coulomb failure envelope for
defining peak behaviour. So, Mohr Coulomb parameters are unconservative and it overestimate
the strength for the range of σ’ beyond which the peak strength is determined.
So, so long as it remains within the range it is fine. So, it is important to specify the state
parameter e, OCR, dilation angle, τ and σ’ for analyzing and interpreting the test results related
to peak soil behaviour. So, we have discussed a very important aspect of peak behaviour which
you will see later in the state boundary surface. So, this information is valid when we consider it
there, so that is all for this lecture, thank you.
597
Lecture-48
Soil Yielding
Welcome back all of you, in the last lecture we have seen peak state. We have discussed various
aspects related to peak state and which is relevant for critical state soil mechanics in terms of the
boundary. Now we have discussed shear strength in module 2, stress path in module 3, and our
essential focus is on the mechanism of failure and to define a more unique failure concept. And
that is where we reached the fourth module where we talked about critical state soil mechanics.
And we have already understood the various aspects of critical state soil mechanics, the
introduction part. And we have also discussed the various boundaries the right most normally
consolidated or isotropic consolidated case at the top peak and to the left tension cut off zone we
have discussed. Now you will see that, adding one more very essential information to these will
enhance our understanding of soil in shearing, and that additional understanding is soil yielding.
There are certain aspects which remain hidden in module 2 as well as in module 3. Now a
complete explanation is possible when we define what is known as soil yielding and the limiting
surface or the limiting curve in two dimensional space given by the yielding. So, that is what we
will be seeing in today's lecture.
598
So, we will start with soil yielding. The plastic deformation results in permanent setting in soil,
all of us know this; this has been discussed in this particular course as well. Like after elastic the
soil yields and then it undergoes softening or hardening behaviour depending upon the stress
history and then it undergoes some plastic setting or permanent setting because of plastic
deformation.
For any stress state which is less than the maximum stress, soil exhibits stiff elastic behaviour,
this also we have seen and that is where the concept of stress history comes into picture. A soil is
already subjected to a maximum stress in the past. And any stress below that maximum stress
soil will behave in a stiff elastic manner. As the loading exceed the maximum stress soil will
again exhibit plastic deformation in the absence of failure.
Like when we start loading this is not a continuous phenomena, it will fail at some point of time
depending upon the nature of stress application. So, what it means is that it keeps yielding when
it crosses it is maximum past stress, and it continues till the point of failure. So, the yielding
happens if the soil has not failed. The past maximum stress is called yield stress, now this is what
is of importance in today's lecture.
And yielding is the point where plastic permanent deformation occurs in soil, well this is not
very difficult to understand. The point at which the yielding start is called the yield point, so we
599
are just understanding specific terminologies yield stress, plastic permanent deformation then
yield point. For one dimensional compression which is basically coming from the oedometer
response, the preconsolidation pressure is considered as yield stress, σ’ y.
All these are repetition of information which we have already seen in the past lectures. Yielding
happens in one dimensional compression, isotropic compression, undrained and drained
shearing. So, there are 2 aspects which we need to distinguish here, the first one is yielding and
the other one is failure. Yielding does not mean failure has occurred, that is the reason whether it
is a one dimensional compression or isotropic compression yielding would happen.
But that may not result in failure of the soil mass, for failure we basically talk in terms of
sharing. So, drained, undrained sharing also bring about failure, in all these cases soil will yield
and that we have seen, it is a very good example is oedometer test result you load, unload, then
again load at the pre-consolidation pressure it will again yield. So, that yielding means high
deformation that is on the NC curve again, so that is what it means.
So, yielding is a general phenomena, but it need not mean that the soil has failed. Now let us try
to understand this, this is a typical one dimensional or isotropic compression. So, which is talked
in terms of volumetric stress p’, so this is loading and from this point there is unloading, then
again reloading. And after this particular point we can see that the plastic deformation is quite
high, and v, the change in v from this point to this point is drastic again loading, unloading.
And once it crosses, now from here this point to this point what has happened? The maximum
stress to which the soil has been subjected to is changing from this point to this point. And then
for all the stress range which is less than this will be stiff elastic, and once it crosses due to again
reloading once it crosses then we will see that again the plastic deformation happens. Now this is
in terms of drained and undrained shearing where we talk about deviatoric stress versus shear
strain.
And then is the same response as that of the p’ versus v response. So, the yielding concept
remains same, stiff elastic response below maximum stress, we have already seen that. Now in
600
the first case, this is the maximum stress where we call it as the yield point σ’ y1 or σ’ y1 is the
yield stress, now you can see that this limp corresponds to elastic behaviour. Stiffness drops once
due to reloading σ’ increases beyond σ’ y1.
In the next loop when it is reloaded and when it exceeds σ’ y1, then the stiffness of the soil or the
elastic response drastically varies and it changes to plastic response. σ’ y defined as σ’ y2, now it
is the yield point σ’ y1 is redefined as σ’ y2, now this becomes the new yield point for the soil.
So, now let us see a bit on the concept of yield curve, because this yield curve concept is
important for defining the critical state framework. Let us consider identical soil samples with
same stress history and initial state. So, we have the same stress history and we have same initial
state. Samples are subjected to different loading conditions until it yields. Let us consider a
triaxial condition σ’1and σ’3.
Let us first consider a uniaxial loading wherein σ’1 is increased, and σ’3 is remaining constant, so
it is a uniaxial loading. Let it yield at point y 1, now this can be represented in q p’ plot, P is the
point that is the identical point with same stress history and initial state, it is also shown on σ’1
versus σ’3 plot. Now here it is very easy to mark because σ’3 is 0 that means this particular point.
601
And σ’1 is increasing means it is going along the y axis and at some σ’1 the soil would yield, so
that is the yield point y 1. Now if you compute this for q and p’ we will also see that it is
corresponding to the first one and then it is y 1. So, y 1 is the point at which this particular soil
yields. Now let us take another case, where σ’1= 0 and σ’3 is increasing, so what will happen σ’1
is 0.
So, here this point is 0 and σ’3 is increasing towards this and yield point is y 2, so let this be y 2
on σ’3 axis. The same thing you can see in q p’ plot let this be y 2. Another case is isotropic
compression increasing σ’3, now one may always wonder like isotropic compression is not going
to lead to any failure, but remember it can yield. So, isotropic compression increasing σ’3 let us
consider it to be yield point y 3.
So, this is isotropic compression, now please note that in isotropic compression there is no q,
there is no shear stress developed, so that is why it is along the x axis or p’ axis, and now let that
be y 3. Let there be one dimensional compression with zero lateral strain, now this is a typical
case of oedometer where there is no lateral strain because it is infinite rigidity offered by this
steel ring, so it is with zero lateral strain, yield point y 4, now let that be y 4.
Then we have undrained compression with u measurement, yield point y 5, so this is y 5. Now we
have discussed at least five cases, now there are different such cases where we can load the soil
by giving different σ’1, σ’3 combinations, and all of them will yield at some point of time. So,
that we can show it as different points, so these are different ways of stress path, applications
based on different combinations of σ’1, σ’3 and it yields.
Now if we join all these yield points, all the yield points correspond to a particular stress history,
we have to keep in mind it is the starting point is P and it corresponds to the same stress history
and initial state. Linking yield points give yield curve or yield locus, so let us join all these points
and that will give us the bound within which the soil mass remains elastic in nature, so this is the
yield curve.
602
So, that is what the basic concept of yielding, yield point, yield curve is all about. Now it is in
two dimensional spaces it is yield curve the same thing becomes yield surface when it is in three
dimensional representations. The yield curve forms the elastic bound for the soil for a given
stress history. Now this particular yield curve corresponds to a given stress history, now for that
whatever is the domain which is shown here, now that corresponds to elastic domain.
Soil with stress state within the yield curve behaves elastic, so both are elastic. So, within this
yield curve is a very important information that within the yield curve the soil state is completely
elastic, and for both the domains it has been shown here. Beyond yield curve, now once the let us
say the stress path reaches here and once it exceeds the yield curve then the plastic behaviour
starts. Soil exhibit elastoplastic behaviour, so beyond that it is plastic behaviour.
Yield curve, it is a generalized preconsolidation stress, what does that mean because in one
dimensional compression the extreme point corresponds to yield stress or preconsolidation stress.
Now here the yield curve is more like a generalized preconsolidation stress, so whichever is it
does not correspond to one dimensional consolidation alone but it encompasses different ways of
yielding for the same soil and with same stress history.
Now preconsolidation stress from one dimensional consolidation is one of the points on the yield
curve, so that is why it is called generalized preconsolidation.
603
Now let us say that the soil sample is isotropically consolidated to P, Q and R, this is v ln p’, this
is the ICL, NCL line and CSL is also shown, let P, Q, R be isotropically consolidated points.
From R the soil sample is unloaded to S, so this is unloaded to S. Now the same is shown P, Q, R
on, so this is on the x axis because it is isotropically consolidated. So, all the points are there on
the p’ axis.
Now yield curve in q p’ divides the domain between elastic and plastic soil behaviour, you may
see that some of these facts are getting repeated but that is important. Now we are slowly getting
into the concept of yield curve, and probably you will be able to appreciate why yield curve is
needed for understanding the shear behaviour of the soil better. Now this for point P, for point P
the maximum stress that it has been subjected to is σ’ or p’p.
So, this is the maximum point and hence the yield curve the extreme point is defined by the
maximum stress. So, this is the yield curve which corresponds to point P, why because this
should be the extreme point of the yield curve as well, for one dimensional compression. Now
you may be wondering that how did I choose a kind of elliptical curve? Now there are different
shapes and it need not be exactly elliptical.
604
Now you will see in the subsequent lectures later that the assumption has been made related to
the shape of yield curve, it is more like an idealization. It is close to but may not be exact, so here
we are just showing it as an example. So, here we are assuming it more or less close to an
elliptical behaviour because it is easy to understand. Now as we load from p to q what is
happening to yield curve?
As we load from p to q the yield curve increases or the yield curve gets expanded, so this is the
fact which we have to keep in mind. And when we load it to R, it again increases, now the yield
curve has changed from the black one to brown, and now the yield curve is now blue, because
the maximum stress has now becomes R. Size of the yield curve is dependent on the maximum
value of p’ to which the soil is consolidated.
Obviously there is an additional condition which is needed here provided the soil has not failed,
so beyond failure there is no concept of yielding. As p’ increases, size of the yield curve also
increases which is very well shown in this figure. Preconsolidation pressure is the extreme yield
stress on the yield curve, so that particular aspect also we need to keep in mind. In all the cases
the preconsolidation pressure is the extreme point of the yield curve.
The unloaded point S is heavily over consolidated, here this particular point, now from here the
unloading has taken place. When soil is reloaded from S to R, from here if it is reloaded we
know that from S to R is elastic; S is within the yield curve. So, from here to here, from S to R it
is completely elastic. When soil is loaded beyond R the size of the yield curve will further
increase. Now let us say that after unloading to S, reloading to R and further it is again increase
then what is going to happen is yield curve further increases in it is size.
605
Now let us try to add the drained response and try to understand in with respect to the yield curve
domain. Soil is first isotropically consolidated from p to q and then unloaded to R, let us see how
it is, v ln p’ plot, q p’ plot from P it is isotropically consolidated to point q and then it is unloaded
to point R. So, now this particular point R, do you have any kind of understanding like what will
be the state of point R?
R will be close to a lightly over consolidated soil, because we have defined critical over
consolidation ratio in the previous lecture. Above which or towards the right it is always lightly
over consolidated, and we have also seen that the critical state line is in between COCR and the
normally consolidated line. So, here point R we can confidently say that this is a lightly over
consolidated point.
So, this is the extreme point of maximum stress application and hence the yield curve also
corresponds to the maximum or the extreme point corresponds to p’ q and R is the lightly over
consolidated point. Now yield curve is defined based on the maximum yield stress at q, point R
is lightly over consolidated. Now soil is sheared from R under drained condition, now remember
in the previous slide we have not sheared it.
We have again followed the unloading, reloading line, so mostly in terms of one dimensional
compression in oedometer. But here we are specifically shearing it under drained condition. And
606
in q p’ plot, please refer back to stress path the slope of the effective stress path under drained
condition is 3, so that is what is written here, just for reminding you. So, from point R, now the
stress path will move at a slope of 3.
Now this is at inclination 3 with slope 3 and R S is the ESP. Now ESP from R to S it behaves
elastically. So, please remember our discussion on stress path in our previous lectures, we never
talked about yielding at that point of time. We know the starting point R, we know what is the
ending point and the inclination or the slope is known at 3, so if you just plot a line it will go and
meet the critical state line, it has to go and meet the critical state line.
Now in that case we never talked about yield, so yielding was more or less hidden in our
explanation, even in the module 2 also we never talked about yielding. So, this lecture on
yielding is an additional information in shear behaviour of soil. So, we need to understand that
for any sort of failure, the first thing is the effective stress path has to touch the yield curve,
please underline this fact because that is important in understanding the behaviour.
So, from R to S when the soil is sheared R to S exhibits elastic behaviour because now we know
what is the yield curve, so till that point it will behave elastically. Now that was not told in our
previous lecture. Now at this point S, soil yields and then it deforms plastically. Further shearing,
what will happen if you share further? Now the plastic deformation has started and further
shearing will expand the yield curve till the ESP touches the critical state line at T and fail.
So, let us see what happens on further shearing. On further shearing you can see that it is with
the same slope but it start propagating towards the critical state line. And once it crosses this
particular yield then the yield curve is also increasing. So, it is expansion of yield curve that is
happening. Now finally it touches the critical state line at point T for which the yield curve is
defined in red colour.
Now is there any case beyond this? No, because the soil has already failed, so T is the final point,
so T is marked here. For drained shearing from normally consolidated state ESP would start from
q. Now we have discussed here lightly over consolidated, so this point is LOC, now what will be
607
the situation? If the soil is normally consolidated, there is no difference the only aspect is Q is
the normally consolidated point this is NCL point, so this is NCL point.
So, instead of starting from R the stress path will start from Q, the drained stress path. So, that is
what is shown here which goes and fails at critical state line, and the failure yield curve is shown
in this figure, I have not completed the whole because that is not needed here. For NC the yield
point is at Q itself, now in the case of R which is lightly over consolidated point you can see that
it will yield in the process of shearing.
So, during shearing it yields, but for normally consolidated point that is the yield point itself or
the yield limit itself, the point Q itself is the yield point.
Now the same thing we will see for undrained response. Soil is isotropically consolidated to P
that is normally consolidated, and then unloaded to Q which is a heavily over consolidated point.
Let us say P and then it is unloaded to Q which is an HOC point, the same is shown in q p’ plot
and we know that this is the yield curve because that is the maximum pressure. Both the samples
P and Q are sheared under undrained condition.
Now normally consolidated is also sheared, heavily over consolidated is also sheared, now let us
see what happens. Point P is already at yield point, now this corresponds to yield stress, so this
608
corresponds to the extreme yield point on the yield curve, P is already yielded, there is no need
for further yielding. Now ESP of normally consolidated, we know that positive pore water
pressure develops and hence it will move left towards the effective stress path of normally
consolidated will move left wards for undrained shearing.
Please refer back if you have not followed this. We know that this is PR is the effective stress
path and it goes and touches at the critical state line. Now that is what is shown in v ln p’, this
part we have already discussed, I do not want to spend more time, what is important for us is that
the yield curve changes. Now ESP for OC, what happens for OC? It moves up till it touches the
yield curve and yield, we are going to discuss this again.
But here for the point Q, what it means? Now this is not an yield point, it has to yield during the
process of shearing. So, now the question is how we have to draw the effective stress path? Now
before discussing let me ask you, like this particular point Q whether it is plastic or elastic? We
all know from our previous discussion that it will be elastic in nature, now if it is elastic it is
effective stress path; we know that it is governed by the slope in q p’; it is 3 divided / 1 - 3 A.
So, if it is 1 - 3A and the value of A = 1 / 3 for elastic response, we know now the effective stress
path correspond to elastic because it has not yielded. So, A value is equal to 1 / 3 which is shown
here, if you substitute that in the slope of ESP which is 3 divided / 1 - 3A, if you substitute 1 / 3
here it becomes infinity. So, that is, it means that it moves vertically upwards, so this is what is
shown here Q S, now how long it will move vertical?
In the last lectures what we have seen this; we have seen that the stress path goes and touches the
peak point. Now here there is a kind of understanding that the peak response and the portion
defined by the yield curve is more or less the same. So, in the earlier lecture please refer back the
effective stress path goes and touches. And we have told at that point of time that it is something
relevant to yielding, the soil has to yield first, so that is what is shown here.
So, the soil has yielded at this point S, soil yields at S and deforms plastically. Further shearing
will cause the ESP move towards critical state line at T and fail. Now in this case it will be
609
negative pore water pressure because it is HOC and the final pore water pressure will be
negative. And hence the effective stress path will move towards T, S T that is shown in v ln p’
plot.
Now we have discussed more or less various aspects of yield curve, now how do we define this?
We have a clue we have to conduct the test, and see how where the points are yielding and then
join it but it is going to be a tough task always. So, if you want to determine because there are
various ways by which we need to generate the yielding condition. Now that is why the
mathematical framework is important even for yield curve.
So, let us see performing shear test with different stress path, that is initial isotropic
consolidation condition, it remains same but we understand different stress path and then join the
point of yielding. So, this is how normally it is done, the same thing can be done with initial
anisotropic consolidation but the initial state remains same for all the stress path. So, this is
another way by which but you can see that the shape of the yield curve will more or less remain
same.
Yield curve corresponds to a particular initial state, please remember when we say yield curve it
corresponds to a particular initial state or stress history. With change in initial state the size of the
yield curve would change. For example if it is normally consolidated to point P, we will have a
610
particular yield curve, now if it is further increased then the yield curve will increase. The shape
of the yield curve will be identical.
Now there is another way of obtaining the yield curve that is performing shear tests with
different stress ratio. Instead of different stress path different q / p’ can be obtained, so if that is
the case we will have again the yield curve. But you will see that whether it is determined using
different stress ratios or with stress path, the shape of the yield curve will be identical.
Now let us talk something about size and shape of the yield curve. Let us say soil from different
depth represent different stress history. For example in the field let us say this is a field
condition, you can see these are different depths from which the soil is collected. Now for this
particular depth at A, at B and C soil will be of identical stress history but A, B and C it
represents different stress history.
Now a particular yield curve corresponds to a particular stress history, now it is very clear to us.
Like for example at point A we have a particular yield curve and the maximum stress point is σ’
y A that is the yield curve at point A. As the stress history changes that means as the depth
increases the yield curve changes. The general shape of yield surface is same for different depth
B and C, this is how it goes.
611
Now size of the yield curve will depend on the depth or the stress history, now whatever we have
discussed in the previous slide we are just understanding it in terms of what is happening in the
field. Now it is very clear that this is the determining part of the size of the yield curve. Now if
we try to normalize q p’ plot with respect to σ’ y, what is going to happen? We can see that there
will be a normalized unique yield curve which is obtained.
Now experimental evidence indicate that ellipse is a good approximation for yield curve, and
that is very specifically defined in modified Cam Clay. Now Cam Clay is a critical state
framework mathematical model, we will be discussing in brief Cam Clay and modified Cam
Clay. Now it is apparent that there has to be an yield curve which is defined, now if it is a Cam
Clay you will see that the yield curve is defined in terms of logarithmic spiral and for modified
Cam Clay it is identified as an ellipse.
So, here I am just introducing that that ellipse is found to be a good approximation for the yield
curve otherwise it becomes extremely difficult for everyone to determine the yield point
separately. So, that is why it is quantified or it is idealized to be ellipse or any other shape. So,
slope, let us say the slope we know that in q p’ slope of CSL is M, but here I am generalizing it
to be k = M.
612
Now let us say we have point A, B and C, now A is inside this yield curve, so preconsolidation
pressure p’ y is the major axis of the ellipse. Since we have introduced ellipse here what it means
is that p’ y is the major axis of the ellipse, so this is p’ y, p’ y B for the yield curve B, and this is p’
yC for yield curve C. Point A, now we are just revising whatever we have told in the previous
slides point A is at elastic state.
At point B yielding occurs for the inner yield curve. As the stress goes beyond B, now once it
departs then the permanent deformation occurs. Stress path BC causes plastic response, so
preconsolidation pressure p’ y is redefined from p’ y B to p’ y C. Now all these aspects we have just
now discussed. This causes expansion of yield curve and the elastic region increases. So, now we
have redefined the yield curve, now all these position from here to here it becomes elastic.
If soil is unloaded before failure soil will respond elastically. Now this point x that is this
particular point x and y are intersection of yield curve with critical state line. Now since we have
introduced ellipse and it is the condition in modified Cam Clay, I would also like to add few of
the aspect here where x and y is the midpoint of major axis. Now this is only for this particular
specific case where in case of modified Cam Clay and ellipse.
So, p’ x = p’ yB / 2 and p’ y = p’ yC / 2, this information is handy as we move forward in our

lecture and we will discuss certain prediction in terms of modified Cam Clay as of now we just
need to understand this.
613
Now whatever we have learned previously we are just going to reiterate again in terms of
normally consolidated and LOC which is on the wet side. Now here what we are trying to
understand is we are just going to understand the concept of softening and hardening in terms of
yielding. So, yielding occurs for NC and LOC, yielding occurs at stress ratio k less than M at P,
that is yielding occurs before SP reach CS.
So, we need to understand that yielding will occur before the stress path reach the critical state.
So, here what is the relevance of k less than M? Now let us say this is the point P, so in the case
of normally consolidated or lightly over consolidated, we know that the value of k is less than M,
that is the stress ratio, stress ratio is q / p’, it is less than M, M is the slope of CSL, now that P is
shown on the NCL line. Further shearing after yielding causes plastic hardening, now we have
already learned this.
Now this is the point which is normally consoled, this is normally consolidated. Now it is already
yielded, there is no more yielding needed. So, once it is sheared it is directly proceeding towards
the critical state line towards failure. Now what it happens, we know that it is subjected to failure
by plastic hardening, and when it hardens what will happen to the slope k? So, definitely we
know it will increase the dimension of the yield curve, now due to plastic hardening that is going
to increase.
614
Now what is the net effect in the case of stress ratio, k also increases because we know the
failure is going to happen only when it reaches k = M. So, earlier k is less than M for an NC soil
now as for that point P and then it is going to increase and k finally when it reaches M it will fail.
At failure drained ESP that is PQ meets critical state line, so this is what from P to Q, it meets
the critical state line at top point of expanded yield curve.
Please remember top point of expanded yield curve is relevant only when you are considering
the elliptical yield curve, where point where the critical state line meets the expanded yield
curve, now this is the failure point PQ. Now this is the expanded yield curve, now this is at the
midpoint of this outer ellipse. At a perfect plastic critical state is reached at Q where k = M, the
stress ratio reaches M.
Here the stress ratio was less than M, and now when it reaches Q it has become k = M. For
undrained shearing ESP is PR, now that is shown and this we have already seen and that is the
yield curve for normally consolidated. Point R is at the top of the expanded yield curve, here
please note it is expanded yield curve for NC and LOC where CSL meet the yield curve, this is R
and this is the failure point.
Now let us say for yielding of HOC on dry side. For that the heavily over consolidated point
correspond to S which with it is initial state at S. So, this is point P, it has been loaded to point P
615
and then unloaded to S, now that is shown here as S which is the starting point. Now for drained
shearing yielding occurs at k greater than M at T. That is yielding has to occur before stress path
reach critical state.
Now please remember this statement is a very important while understanding the shear
behaviour. So, from S we know that there has to be a point T, now this is the point T where it has
to yield, this we have seen in our previous lecture as well. Now if you consider the stress ratio at
point T, the slope at T will be greater than that at M, so that is what is shown here k is greater
than M at point T which corresponds to the yield curve.
So, this is the maximum stress to which, so this becomes the reference yield curve for us. And
the stress path for drained it will be at slope 3, it will go and touches the yield curve and that is
the point where it yields. Now further shearing after yielding causes plastic softening, so NC it is
plastic hardening, now it is plastic softening. So, at from point T it is plastic softening and hence
the stress comes down and fails at critical state, you can see that from T it reverse back and then
meets the critical state line.
So, k greater than M as it is sheared it will meet the particular yield curve and then it gets back
where k = M, so M, so the k value the stress ratio is reducing from a higher value to M. Whereas
in the case of NC, k value has increased from an initial value to M. Decrease the dimension of
yield curve, please remember we have discussed more or less the same thing in most of the cases
but we are adding one by one in each of these slides.
In this particular slide what we are adding is about plastic softening and plastic hardening, even
that you know but then in the context of yielding how does it translate to? So, decrease the
dimension of yield curve in the case of NC it is increasing of yield curve, whereas in the case of
OC it is decreasing of yield curve. So, this is the point where the final yield curve would lie, so
that is what is shown here.
Now decrease the stress ratio k decreases to M from an initial value. At failure, drained ESP that
is STU it meets the critical state at the top point of contracted yield curve. Now this is the
616
contracted yield curve, remember this midpoint, top point it is relevant because we are discussing
this in terms of ellipse for our convenience, it can be other shapes as well. Now that is what is
shown in the case of critical state line.
Now we have now made it very, very precise, in the earlier case we defined it in terms of peak
line it is now replaced by yield curve. Now you can see that from here this point it is along the
unloading curve and then it moves towards U. Now if it is for undrained shearing from S, now
this is for drained shearing, we know that for undrained shearing the slope is governed by 3 / 1 -
3A, till it yields it has to be the slope for which A = 1 / 3.
Because A = 1 / 3 is for the elastic condition and hence when we put it has to be vertical. So, we
know that from S it has to be in the vertical direction. So, ESP SV it yields at point V, so now
how it will be? This is the reference yield curve from S, it moves towards V and it yields at this
point V, so up to here it is elastic behaviour. Then after yielding strain softens, now from here
also it has to drop down and yield curve decreases and meet CSL at W at which k = M.
So, this is the point where it meets critical state at point W. Now at this point k = M, at point V, k
is greater than M and that reduces to k = M, so that is shown in the v p’ plot as well. So, this is
the contracted yield curve for undrained, so this is for drained, and this is for the contracted yield
curve for undrained.
617
Now ESP touches the critical state at failure which we know. Continuous plastic deformation
shearing happens at this point, which point at the point where the ESP touches the critical state
line. Now this is what it means, it touches the critical state line at this point and this is the yield
curve. Now at this particular point, the plastic strain increment vector that is δεp of in plastic, this
p means it is volumetric, here p it means it is plastic.
Here it is deviatoric and this p means it is plastic, and this ratio is known as plastic strain
increment vector. Now that is equal to 0 and it is parallel to q axis, it is something like this, now I
will not get into the details of plastic strain increment vector, because it is purely based on the
plastic analysis. And we have not touched upon to that details in this course and we are not going
to do that as well.
But this information that it is parallel to q axis may be handy for you when you work in this
direction. So, that is the reason why I have just introduced this particular terminology. Perfectly
plastic critical state is reached with k = M at this particular point and this happens at the topmost
point of the current yield curve. Now this is the current yield curve, remember this is only for a
specific case where we discuss in terms of elliptical yield curve for modified Cam Clay.
At critical state which is the failure state indefinite plastic shearing can occur and that is the
reason why it is in this particular manner, without further expansion contraction of yield curve.
618
Once the yield curve meets or the effective stress path meets the critical state line and that
particular yield curve this plastic increment vector is equal to 0 and it is in upwards. And at this
particular condition what it means is that infinite plastic shearing can happen without any change
in the yield curve.
That is there is no volume change, there is no other change that is happening and there is no
plastic softening, plastic hardening that is happening, it is simply a plastic shearing case, an
infinite or indefinite shearing that happens.
So, let us now try to summarize the points that we have learnt. So, we have today understood a
very important aspect of soil yielding which need to be integrated with the shear strength as well
as the stress path portions which we have studied earlier which adds a complete meaning to shear
behaviour of soil and the failure mechanics of soil. The past maximum stress is called yield
stress, and yielding is the point where plastic permanent deformation occurs in soil, so up to
yielding it is elastic.
For failure to take place soil has to yield first, now this is a very important condition which is
needed in critical state soil mechanics. For stress state less than the maximum yield stress soil
exhibit stiff elastic behaviour. Yield stress changes in soil if the loading result in permanent
619
deformation without failure. So, that is what a very good example is one dimensional
compression.
For one dimensional compression preconsolidation pressure is the yield stress, σ’ y. Now this is
specific only to one dimensional compression because this is the extreme most point. Yield curve
corresponds to a given stress history, and initial condition or which is the specific volume. The
yield curve defines the elastic boundary for a given stress history that is within the yield curve is
elastic.
One dimensional compression, size of the yield curve is dependent on the maximum value of p’
to which the soil is compressed as yielding happens without failure that is either hardening or
softening the size of the yield curve increases or it may be decreasing as well. The increase in
size of the yield curve occurs till the ESP meets, please remember increasing happens because of
hardening and decrease happens because of softening.
Shape of the yield curve remains identical, even though we are conducting we are finding out the
yield curve at different stress history the shape of the yield curve is more or less identical. And
this is understood from the experimental evidence which is already reported in the literature.
Different shapes of the yield curve is idealized for different models that I have already told you
for example in the case of modified Cam Clay, it is elliptical.
620
So, it is a idealization of the shape of the yield curve. The logarithmic spiral, it is used in Cam
Clay and ellipse in modified Cam Clay are some of the shapes of the yield curve which has been
used by previous researchers. For normally consolidated lightly over consolidated yielding
occurs at stress ratio k less than M and on further shearing k increases to M at failure. For HOC
yielding occurs at stress ratio k greater than M and on further shearing k decreases to M at
failure.
For elliptic yield curve, plastic strain increment vector which is equal to 0, it is parallel to q axis
and it occur at the top most point or midpoint of major axis. At critical state which is the failure
state indefinite plastic shearing occur without further expansion, contraction of the yield curve.
Further expansion, contraction means further hardening or softening which results in the size of
the yield curve. So, that is all about soil yielding, we will see in the next lecture, thank you.
621
Lecture-49
Cam Clay
Welcome back all of you, in the last few lectures we were dealing with critical state soil mechanics.
If you have noticed we have been adding certain important information one after the other. And
lastly we discussed about soil yielding, now where are we heading to? We are heading towards
defining a state boundary surface for the soil. And when I say boundary surface, it means the
domain within which the soil would exist, soil would exist mechanically, mechanical means
because of it is mechanical response.
Otherwise soil is always there, now whether soil would exist as a load bearing medium when it is
subjected to shearing, that is what it means. So, when we say boundary surface, we are trying to
understand the limit within which the soil would exist during shearing as a load bearing member.
Now for that we need to define a limit within which the soil may not exist. And one possibility
was in terms of soil yielding.
So, yielding is a good indication, that, yes, soil is going to be in severe distress that is plastic
deformation starts setting in. And we have seen in the last lecture that as the yield curve
progressively increases it reaches a point where the yield curve meets the effective stress path and
the critical state. So, though that particular yield curve which passes through the point of
intersection of effective stress path with the critical state line, that particular yield curve happens
to be the final boundary.
During the expansion of yield curve that defines the final boundary. Now we have understood it
physically, now for any engineering application we need to define it mathematically. So, there
need to be certain models based on which the yield curve is defined. So, in today's and the
upcoming lecture we will see how yield curve is defined mathematically based on 2 popular
models which is based on critical state framework.
622
So, the first model is Cam Clay model which we will see in today's lecture. I would also like to
mention here that in this particular course you will not be studying all the aspects of such a model
or critical state framework. We are basically understanding the concepts in stress parameters in
terms of stress variations rather mostly applicable for stress path. We are not in a position to gain
more knowledge from the strain point of view.
I mean to say when I say a strain elastic strain and followed by plastic strain, a bit of it we will be
discussing in the next module, next lecture I mean. So, here what we are understanding would be
the basic understanding which is needed further for understanding the plastic response. So, with
that let us try to understand what is a Cam Clay model today and with that understanding we will
move on to the next model. And possibly you will be able to appreciate why we will not be able
to deal the entire aspects in this particular lecture.
So, the first model is the Cam Clay model. Now the name suggests that it is a kind of clay or
something like that, it is not so. I would strongly recommend you reading Schofield and Wroth’s
book wherein it is stated that cam is the name of the river which flows besides or close to
Cambridge University. Now clay is here clay in the sense it is not meaning a kind of particular
clay may be which is obtained from there, it is not like that.
623
Cam clay is a generic name for that particular model, it has nothing to do with the clay response,
it has something to do with the theoretical model which they have developed. And they wanted to
make it a more generic name in nature, so that is why such a name which is Cam Clay. Now as I
already told defining state boundary surface is important, we are heading towards that. Now there
are different ways by which you can define the state boundary.
Some of them we have already seen, we have seen a peak line that is also a state boundary that is
a part of state boundary. We have seen the yield kind of curve in the last lecture; we have seen the
tension cut off zone, so these are all fragments of state boundary. Now there can be a specific shape
and model based on which we define this, that is also state boundary. So, there are different ways
by which the state boundary surface is defined.
This state boundary surface gives the domain within which soil state can exist during loading. So,
this later part of the sentence is important, we are discussing the existence of soil state during
loading otherwise soil is very much there. We are not bothered about that existence; whether the
soil will be able to bear it is load without failure. The normally consolidated line, the peak line, we
will see this normally consolidated why it becomes a kind of fragment of SBS? That we will see
later.
Peak line we have understood clearly, tension cut off zone that also we have understood, we know
that nothing is going to be there towards the left end. Of course in normally consolidated in vp’
plot we have already seen that it is the right most boundary, so in that sense that is also a state
boundary surface. And yield curve, specific yield curve, when I say yield curve not all the yield
curves because yield curve can expand in it is size, is the final yield curve where I told already
which passes through the point of intersection of effective stress path meeting with the critical
state line.
These are all fragments of SBS. Now we will see, we will put all these facts together and finally
understand what is a kind of state boundary surface in 2D as well as in 3D variation or the
framework? Mathematical representation is needed for SBS then only it will be useful for
engineers. This is possible, now what is possible? Mathematical representation is possible by
624
fitting equations to experimental data developing purely a theoretical framework based on
mechanics or the combination of 2.
Now the biggest challenge is how good the framework represent the real soil behavior? Now this
is asking for more, like we are trying to mathematically quantify something which is beyond our
imagination and that soil behaviour. We have to admit the fact that in spite of all our knowledge
and development still there is a gap in understanding the actual soil behaviour. So, a particular
model which may work very well for a specific type of soil and situations may not work for another
type of soil in the same situation.
Or for a different situation it may not work for the same soil, so there are different possibilities.
So, we cannot generalize a particular model, but how close we are not looking for a 100%
matching, how close the given model is with respect to the actual or the real soil behaviour? That
is what the biggest challenge is. So, there is always a tradeoff between simplicity which is
essentially needed for computational ease and adequacy, how could it represent?
So, there is always a we need to balance this, if we want to clearly model every bit of aspect of the
soil that is going to be computationally highly intensive. And that may not be really advantageous
from a design perspective where there are equally good amount of uncertainties otherwise also.
So, we need to strike a balance between how simple our model is vis-a-vis how good it represents?
So, that is why simplicity and adequacy.
One of the basic constitutive models based on critical state framework is the Cam Clay model in
short CCM. So, we will now try to understand the critical state model, Cam Clay model which is
the primary development, this is the original Cam Clay model. And then it was modified a bit and
that is known as modified Cam Clay which we will see in the next lecture.
625
So, let us start with Cam Clay, understanding Cam Clay model all the understanding which we
have till now related to 2D representation of critical state and soil yielding becomes handy in
understanding the Cam Clay model. Without knowing that it becomes difficult to appreciate what
is Cam Clay model. So, that is why we have grown or we have dealt the lectures in steps, adding
one information one after the other.
So, I would strongly recommend that you have to go through the previous lectures before starting
understanding this Cam Clay model. This was put forth by Schofield and Wroth in 1968 which is
the original Cam Clay. Soil is assumed to be frictional, essentially it is assumed to be frictional.
Soil state boundary surface is taken as the yield surface or the yield curve in two dimension, it is
very easy for us to understand this particular concept.
It is a bound which we have seen and that can be used for defining the boundary. Plastic hardening
related to plastic volumetric strain. Now I would be a bit cautious in using these terminologies in
this particular course. The very reason is we are not discussing anything important or in depth
about the plasticity behaviour of soil. So, here plastic hardening whatever the expansion of the
yield curve is basically related to the plastic volumetric strain.
You remember the v versus p’ plot where the normally consolidated line it moves along the
normally consolidated line. So, every time it moves we have seen in the last lecture the size of the
626
yield curve would increase. So, basically the plastic hardening is related to plastic volumetric
strain. Now there is another kind of strain that is going to take place that is shear strain during
loading, so that part is also there, unfortunately we will not be able to discuss those in detail.
Different models in Cam Clay differs in terms of the equation used to represent the yield curve.
So, basically the difference between different Cam Clay models in the same Cam Clay family is
the kind of yield curve that is used. Otherwise the kind of critical state concepts that we use in all
these models are pretty same. In original Cam Clay, the yield curve is logarithmic spiral because
of the closeness in whatever we have seen the kind of boundary that we have seen logarithmic
spiral happens to be closer.
And that is why for Cam Clay model that shape is adopted as the yield curve. So, it looks something
like this, it is in q p’ plot the yield curve is shown in this manner, so this is the yield curve. Now
this is the logarithmic spiral yield curve, now for completeness we have shown for compression
as well as for extension. But in our subsequent lectures we will be basically concentrating on the
compression portion.
So, that is one half of the yield curve, in fact the yield curve is mirror image on both sides in
compression as well as in extension. So, we will be basically dealing with positive value of q. The
logarithmic spiral shape may not be convenient for numerical modelling; you can see this
particular quotient wherein there is a pointed edge here. Now for all numerical modelling there has
we need to ensure the continuity of the differential equation which is used.
So, basically here this particular location it creates problem in some of the numerical models and
in the stability of the algorithm. So, for that is the reason why the modification has been given, but
to understand the basics of Cam Clay we need to understand this particular model, original Cam
Clay also. So, it may predict unrealistic plastic shear strain increment, now again I would like to
use this particular word with caution.
627
Because we are not discussing this particular term the plastic shear strain increment, it is basically
relevant to plasticity behaviour of soil. So, there is an issue that it can predict because of this
particular model can predict unrealistic plastic shear strain increment.
So, let us understand what is the mathematical representation of Cam Clay model yield curve. Let
us consider a specific volume of vi, so this is q p’ plot and this is vp’ plot. Now let us consider the
reference vi, so this is the vi corresponding to which there is an initial consolidation point which is
Pi on the ICL, NCL. Now this can be either isotropically or it is one dimensional consolidated
point Pi for which the specific volume is vi.
Now let us try to understand it in q p’ plot, now this is the initial point Pi, logarithmic spiral curve
corresponding to this particular point. Now we have already seen in the yielding lecture that the
extreme point is given by this preconsolidation pressure. Now for this particular case it is Pi, so
the extreme point of the yield curve Yi will be Pi. So, we need to fit a logarithmic spiral Yi which
is the yield curve corresponding to Pi.
Now yield curve Yi correspond to a specific volume vi, now we need to understand that this
particular yield curve is specific to this particular specific volume vi. Now as the specific volume
changes which means P is going to change, so let us say if it is above or below the kind of Pi will
also change. Now when Pi changed that means that the yield curve also changes, so now for the
628
time being we need to understand that a particular yield curve is associated with a given specific
volume.
Or we can say that this particular yield curve represents a kind of undrained behaviour, what is
undrained behaviour? Where v is not changing, so it is both ways true, if I say that yield curve
corresponds to a given specific volume or this particular yield curve represents a kind of undrained
behaviour, both explanation is same. Point on critical state line is Ci, so let us say for this particular
vi the point on the critical state line is Ci.
Now when you transfer this how will it look like? Now this part is very important, now when we
transfer this particular point it looks like this. That is the point where the critical state line meets
the top portion of the yield curve, top portion means the extreme top portion, so that is Ci. So, this
relationship one should keep in mind. Let us consider another specific volume of vj which is lesser
than vi, so this is vj.
Now the yield curve corresponding to this will be y j and the points will be P j and Cj, so this is Pj
that is taken to the q p’ plot and this is Cj. So, that Cj will lie on the critical state line and the kind
of yield curve will be like this. Now yield curves are identical in shape with different size. The
shape of these 2 yield curves are very close to each other but the shape differs depending upon
whether it is vi or vj.
So, it is apparent that there will be a family of different yield curves. Family of yield curves defines
a closed boundary surface. Now this how does a closed boundary surface look like, I am not
showing it here we will discuss when we discuss about the state boundary surface. We need to
imagine that there will be different family of such yield curves leading to a complete boundary
surface in a given three dimensional space.
This closed boundary surface is called the state boundary surface. Again we will see about state
boundary surface bit later. Only when we understand this Cam Clay and the prediction clearly we
will understand the state boundary surface.
629
So, we have understood the basic framework and now the equation of state boundary surface or
yield surface, that is how do we represent this particular yield curve. The logarithmic spiral yield
curve is given by this expression which is
𝑞 𝑝′
+ ln ( ′ ) = 1 (𝑞 > 1)
𝑀𝑝′ 𝑝𝐶𝑖
And here we are taking the condition q greater than 0 that means on the compression. Now q we
know, M, M is the slope of critical state line. Which is again a critical state parameter, fundamental
parameter, p’ we know, ln p’, p’ci, what is p’ci? P’ci is the effective pressure corresponding to this
particular point ci, so it will be somewhere here. So, that is also known, so whatever is needed is
known from this particular equation equal to 1 and for the condition q greater than 0, so this is p’ci.
Now the main features of yield curve that it should pass through origin where q axis is the tangent
at that particular point. So, that is clearly visible here, you remember when we discussed about the
last few lectures, we discussed that it is better that the failure envelope passes through the origin
because every time we cannot ensure the existence of cohesion. So, that is why in this particular
framework we have already seen that it is purely taken as a frictional material.
So, it is passes through the origin with q axis as tangent, zero gradient at critical state. Now at this
particular location Ci and Cj what it means is that there is a zero gradient. If I draw a tangent at the
630
top most portion, it is horizontal, so it is zero gradient at critical states at this point Ci and Cj.
Vertex is at Pi and Pj and that happens to be the extreme point of the yield curve on the right side.
It also follows this important relationship that is

′
𝑝𝑝𝑖
ln ′ = 1
𝑝𝐶𝑖
That is the relationship between this point the extreme point and this point of intersection, that is
the mean stress. At this particular point and this particular point is related and what is the
′
𝑝𝑝𝑖 ′ ′
relationship? It is ln 𝑝′ = 1 or if you take the anti log it is 𝑝𝑝𝑖 = 2.718 𝑝𝐶𝑖 . So, this is one of the
𝐶𝑖
attributes of this particular yield curve.
𝑝′
Similarly if it is p’j it will ln 𝑝𝑝𝑗
′ = 1, so that relationship also holds. Now how this relationship is
𝐶𝑗
coming? We will see in the subsequent slide. The yield curve defines all the permissible
equilibrium state for soil with vi, so all the possible condition is governed or it is taken care of by
this yield curve. Now CCM follows the normality condition, the critical state model follows the
normality condition, and what is normality condition?
Again this has to come from the discussion of plasticity behaviour of soil. Now if somebody is
interested to know this more, I strongly advise you to go through the plasticity behaviour of soil
in this specific textbook; especially you can refer to the textbook by Potts, so what is meant by
this? It means that the plastic strain incremental vector is normal to yield curve, so in the last
lecture also we have seen that it is perpendicular, it means that. Now if I start discussing this our
designated portions is going to suffer, so I cannot get into the details of plastic behaviour.
631
𝑞 𝑝′
Now let us again start with I am referring to the same equation here (𝑀′ + ln (𝑝′ ) = 1 (𝑞 > 1))
𝑝 𝐶𝑖
𝑞 𝑝′
Let us start with this particular equation + ln (𝑝′ ) = 1 So, we will start from here this is the
𝑀𝑝′ 𝐶𝑖
yield curve and this is Ci Pi and this is marked. Now we have understood one part, now we are
slowly adding another part here, what is that part? That is the unloading line or the elastic
component line, where does it fall with respect to whatever we have discussed.
In the last slide we have not introduced this unloading line. Now you can see that P i is here, this
particular point Ci is nothing but the point of crossing where this unloading line is crossing the
critical state line, I would not say that it is touching. It is only notional in this particular plot v p’,
it is the point where it is crossing, so this particular point corresponds to C i, so what is the
relationship? That is what we are trying to understand.
So, when we understand Cam Clay model we need to keep this these aspects in mind, so then only
it will work. So, Ci is the point where the yield curve intersects the critical state line and that is
where what it is shown here. Now here you need to also see that this particular line the unloading
line represents elasticity, the yield curve also defines the domain of elasticity. So, that is the point
here the crossing point is Ci.
Now,
632
𝑞 𝑝′
= 1 − ln (𝑝′ ),
𝑀𝑝′ 𝐶𝑖
𝑞 ′
= 1 − ln 𝑝′ + ln 𝑝𝐶𝑖
𝑀𝑝′
So, here you can also designate this particular point as vf, why? Because that is the point where
the effective stress path would meet and this any point on the critical state line can be considered
as the failure point or VCS both are same. If we consider vf, we understand that it is the failure
point and that corresponds to VCS as well.
So, now this is the point that Ci is a point on the critical state line, so designating it as vf holds
good. It also mean it is p’f, now p’ f is same as p’cs. Now from this we can always write
𝑣𝑓 = 𝑣𝑘 − 𝜅 𝑙𝑛 𝑝𝑓′
This is the point and that point lies on this particular line as well here. Now how do we represent
this particular loading, unloading line again refer back to our critical state discussion wherein we
can write this particular line as v.
In general any point on this loading, unloading line can be written as 𝑣𝑓 = 𝑣𝑘 − 𝜅 𝑙𝑛 𝑝𝑓′ . We can
also write the same point because it is lying on the critical state line, now how do we represent
critical state line. Now here we know that it is always 𝑝𝑓′ because it is lying on the critical state
line. So, similarly for this particular point can also be designated as
𝑣𝑓 = 𝛤 − 𝜅 𝑙𝑛 𝑝𝑓′
Now we can see this both represents v f, we can equate it.
𝛤 − 𝜅 𝑙𝑛 𝑝𝑓′ = 𝑣𝑘 − 𝜅 𝑙𝑛 𝑝𝑓′
Now if you rearrange it,
𝛤 − 𝑣𝑘
ln 𝑝𝑓′ =
𝜆−𝑘
So, we got an expression for ln 𝑝𝑓′ . Now we know that 𝑝’𝐶𝑖 = 𝑝’𝑓 the same point.
633
So, now we have got an expression for 𝑙𝑛 𝑝’𝑓 , substituting that here we can write
𝑞
= 1 − ln 𝑝′ + ln 𝑝𝑓′
𝑀𝑝′
Now what are we doing? We know one expression which represents the yield curve; we are just
trying to represent the same equation in a more general form involving the critical state parameters.
𝑞 (𝛤 − 𝑣𝑘 )
′
= 1 − ln 𝑝′ +
𝑀𝑝 𝜆−𝑘
Now 𝑣𝑘 again we know v kappa is not a fixed parameter, it depends upon where the unloading has
been done, again we have discussed this point. So, we need to replace 𝑣𝑘 , so how do we replace v
kappa? Now we know that is
𝑞 (𝜆 − 𝑘) − (𝜆 − 𝑘)𝑙𝑛𝑝′ + (𝛤 − 𝑣𝑘 )
=
𝑀𝑝′ 𝜆−𝑘
Now we need to replace this 𝑣 = 𝑣𝑘 − 𝑘𝑙𝑛 𝑝′ again the equation for this particular loading-
unloading line.
From which we can write 𝑣𝑘 = 𝑣 + 𝑘𝑙𝑛 𝑝′ . So, we can substitute for 𝑣𝑘 .

𝑞 (𝜆 − 𝑘) − (𝜆 − 𝑘)𝑙𝑛𝑝′ + (𝛤 − 𝑣 − 𝑘𝑙𝑛 𝑝′)
=
So, we are substituting this to here. So, then if you expand it,
634
𝑞 (𝜆 − 𝑘) − 𝜆𝑙𝑛𝑝′ + 𝑘𝑙𝑛 𝑝′ + (𝛤 − 𝑣) − 𝑘𝑙𝑛 𝑝′
=
So, after simplifying we can write

𝑞 (𝛤 − 𝑣𝑘 ) 𝜆 ln 𝑝′
= 1 + +
𝑀𝑝′ 𝜆−𝑘 𝜆−𝑘
So, equation of state boundary surface in terms of all the critical state parameters can be rewritten
as
𝒒 𝝀 ′
𝜞 − 𝒗𝒌
+ ( ) 𝒍𝒏 𝒑 − =𝟏
𝑴′𝒑 𝝀−𝒌 𝝀−𝒌
So, this is the equation for the yield curve which is represented by original Cam Clay model.
And you can see that lambda, kappa, M these are all the critical state fundamental parameters of
soil, so what we have done? We have defined the yield curve in terms of fundamental soil
parameters which is coming from the critical state framework. So, this is one part of the story, and
in this particular course we will be discussing only this part of the story, what is the other part of
the story?
The other part of the story deals with the plastics straining, where we need to define how the
plastics straining happens and how it propagates. So, that particular aspect we are not discussing
635
in this course. So, here 𝑀, 𝜆, 𝜅 𝑎𝑛𝑑 𝛤 are soil specific parameters. The yield surface correspond to
constant specific volume, we have already discussed that. This would correspond to undrained
stress path, why?
Because it is corresponding to a constant specific volume, so it is as good as telling it corresponds

to a undrained stress path. Now equation for undrained stress path, we can also derive substitute
′
for 𝑣 = 𝛤 − 𝜆 ln 𝑝𝐶𝑆 which is the point of intersection of undrained stress path or yield curve with
critical state line. Now if this is the particular point, now if that is considered as 𝑝’𝑓 𝑜𝑟 𝑝’𝑐𝑠 .
′
And for this particular v, we substitute this equation 𝑣 = 𝛤 − 𝜆 ln 𝑝𝐶𝑆 , what we will get? So, we
get
𝑞 𝜆 𝑝′
+ ( ) ln ( ′ )=1
𝑀𝑝′ 𝜆−𝑘 𝑝𝐶𝑆
Now considering this to be yield stress (𝑝𝑦′ ), we will start from the same equation
𝑞 𝜆 ′
𝛤 − 𝑣𝑘
+ ( ) 𝑙𝑛 𝑝 − =1
Because we need to find out the relationship between 𝑝’𝑦 𝑎𝑛𝑑 𝑝’𝐶𝑖 . So, for point Pi, so this
particular point Pi q = 0, at this particular point q = 0, also if you see here this is the same point if
you consider here 𝑝’𝑦 , this is considered to be vy corresponding to that particular yield.
636
We can write
𝑣𝑦 = 𝑁 − 𝜆 ln 𝑝𝑦′ (𝑁𝐶𝐿)
what is N? N is the point on the normally consolidated line corresponding to unit pressure which
we have already seen. So, specifically for vy we can write so
𝑣 = 𝑁 − 𝜆 ln 𝑝′
Now for this particular point we can write 𝑁 − 𝜆 ln 𝑝𝑦′ , so that is what is written here from the
normally consolidated line. Substituting this in the state boundary surface equation or the yield
curve equation q is 0,
𝜆 𝛤 − 𝑁 − 𝜆 ln 𝑝𝑦′
0+( ) 𝑙𝑛 𝑝′ − =1
𝜆−𝑘 𝜆−𝑘
Now rearranging, the denominators are same, so we can write
𝜆 ln 𝑝𝑦′ − 𝛤 + 𝑁 − 𝜆 ln 𝑝𝑦′
=1
𝜆−𝑘
We can write
𝑁−𝛤 =𝜆−𝑘
So this relationship comes from here.
Now for the critical state point, this particular point we can write
′
𝑣𝐶𝑆 = 𝑣𝑘 − 𝑘 ln 𝑝𝐶𝑆
point from the loading, unloading line. From the critical state line the same point can be represented
as
′
𝑣𝐶𝑆 = 𝛤 − 𝜆 ln 𝑝𝐶𝑆
Now substituting these two,
𝛤 − 𝑣𝑘 = (𝜆 − 𝑘) 𝑙𝑛 𝑝𝐶𝑆
637
So,
𝑣𝑦 = 𝑁 − 𝜆 ln 𝑝𝑦′
equation from the normally consolidated line, vy can also be written from this particular loading,
unloading line.
𝑣𝑦 = 𝑁 − 𝑘 ln 𝑝𝑦′ (𝑢𝑛𝑙𝑜𝑎𝑑𝑖𝑛𝑔 𝑙𝑖𝑛𝑒)
Substituting these 2 we can write

′
𝑁 − 𝑣𝑘 = (𝜆 − 𝑘) ln 𝑝𝐶𝑆
Now from the previous slide considering the critical state point we can write
′
𝛤 − 𝑣𝑘 = (𝜆 − 𝑘) ln 𝑝𝐶𝑆
Now (𝑁 − 𝑣𝑘 ) − (𝛤 − 𝑣𝑘 ) will give

𝑝′
𝑁 − 𝛤 = (𝜆 − 𝑘) 𝑙𝑛 ′ .
𝑝𝐶𝑆
But we already have derived this particular relationship where 𝑁 − 𝛤 = 𝜆 − 𝑘

substituting that we can write
𝑝′
𝑙𝑛 ′ =1
𝑝𝐶𝑆
638
this is where we started off with that this particular relation holds good. So, that we have just
now proved that this particular point and this particular point on the p’ axis is related by this
𝑝′
expression 𝑙𝑛 ′ , that is the attribute of this particular yield curve and
𝑝𝐶𝑆
𝑝′
′ = 2.718
𝑝𝐶𝑆
So, that is all some very basic information related to critical state framework in Cam Clay model.
So, the one part is what we have discussed now, let us try to summarize this particular portion. We
need to understand that this type of understanding we need to develop in steps. So, what we are
exactly doing is that adding one by one. So, here for this particular course we need to stop here
our level of understanding.
Now the next level of understanding for you will be in terms of plasticity behaviour of soil, where
both the complete constitutive relationship becomes important. And for which the stress strain in
both elastic and plastic framework need to be understood. So, let us summarize today's lecture.
The Cam Clay model CCM is a simple and basic elastoplastic constitutive model based on critical
state framework.
639
Now again I would like to use a caution while telling elastoplastic constitutive model, we have not
completed the entire constitutive model. We can say it is constitutive model when there is a proper
relationship between stress and strain. So, we have discussed the partial framework right now. It
defines the state boundary, now here our entire effort is in understanding the state boundary
surface.
And that is fully justified in this type of discussion, and that you will see that whatever we have
learned just now we will understand that in terms of three dimensional state boundary surface. So,
that part is completely dealt in this particular course in which the soil exists during shearing. CCM
is one of the mathematical representations of the state boundary surface; there are different ways
by which it is done now this Cam Clay model is one of such representations.
In original Cam Clay the yield curve is represented by logarithmic spiral, yield curve correspond
to a particular specific volume when the specific volume change the size of the yield curve also
changes but the shape essentially remains the same. So, yield curves are identical in shape but with
different size, size changes with specific volume. Yield curve passes through origin and have zero
gradient where it intersects with the critical state line and that is at the top most portion.
The equation of yield curve in Cam Clay model is

𝑞 𝜆 ′
𝛤 − 𝑣𝑘
+ ( ) 𝑙𝑛 𝑝 − =1
And it also holds this particular relationship,
𝑝′ 𝑝′
𝑙𝑛 ′ = 1, ′ = 2.718
𝑝𝐶𝑆 𝑝𝐶𝑆
So, that is all for this particular lecture, we will see the modification of this original Cam Clay in
terms of modified Cam Clay framework in the next lecture, thank you.
640
Lecture-50
Modified Cam Clay
Welcome back all of you, in the last lecture we have started off with critical state model and the
first one that we discussed was Cam Clay model. We know that this is an elastoplastic model,
which utilize the concept of critical state. We have also seen the yield curve and we noted that it
is a logarithmic spiral. Now there is some other developments in the critical state model and that
led to modified Cam Clay.
So, in the last lecture we told that there are certain aspects related to plasticity behaviour of soil
which also need to be incorporated. In the last lecture I purposely did not introduce those, because
we first need to understand what a critical state model would look like. We have got a feel of it,
we have defined the yield curve we have defined it mathematically. So, now we are conversant
with the whole process.
So, in today's lecture we will see the modified Cam Clay model along with some basic aspects of
plastic behaviour of the soil, I mean to say mechanical plastic response of the soil.
641
So, today's lecture we will start with modified Cam Clay model. So, this is what we will be dealing
with in today's lecture. In fact Cam Clay and modified Cam Clay model, it was more or less
developed during the same period. The initial one was proposed in Schofield and Wroth and later
the modified Cam Clay model was proposed by Roscoe and Burland in 1968. And we know that
these models were essentially developed for defining the triaxial loading or the triaxial test.
These are popular elastoplastic model, that can incorporate hardening, softening that happens
during shearing. I mean to say it is not like elastic response and then failure, after yielding there is
some other response that we have already noticed; it can be either hardening or softening. And
when that gets incorporated we say it is an elastoplastic response. Critical state is the bounding
surface, now what is so specific about Cam Clay and modified Cam Clay is it is entirely based on
the critical state concept.
And hence the critical state forms the boundary of the soil, it is the bounding surface and the soil
shears under the consideration of ideal plasticity. The parameters of modified Cam Clay model are
obtained from experimental data and that is the beauty of Cam Clay model, rather we can get the
parameters needed in a simplified manner. For example 𝜆 it comes from consolidation test, either
it comes from isotropic or one dimensional consolidation.
The failure response in terms of M, it comes from triaxial testing. So, there is a way by which you
can determine the parameters in a more simplified manner. The major difference between Cam
Clay model and the modified Cam Clay model is in the shape of yield curve, I think we have
discussed this before also. So, the major difference is in terms of the shape of the yield curve, the
earlier one we have seen that it is a logarithmic spiral.
So, in Cam Clay model yield curve is logarithmic spiral whereas in MCCM it is ellipse. Now
ellipse defines a more symmetrical and regular shape, it is not about symmetry, it is about the
corners like in the logarithmic spiral we have seen that it comes and meets at a particular point.
Now this may sometimes create issues in numerical instability as you solve the problem, because
there is an issue of continuity of differential equation.
642
So, what happens is in the case of modified Cam Clay it becomes an ellipse, so it is more smooth
shape we can say. It is not about symmetry, it is about more smoothness in the shape. Now the
thing is the assumed shape may not be truly representative of the observed yield curve, if we
actually want to determine the yield curve experimentally, we have seen that, it is a quite an
extensive procedure.
You need to find out different ways by which the soil gets yielded and then you join all those
points, it is not that easy. So, that is why the definition of the shape becomes important. Now what
it says is that the shape which has been assumed, which is an ellipse may not be truly representative
of the actual soil behaviour. So, it is an idealized condition, the actual behaviour may be something
else some other shape.
But it is very close to it, a logarithmic spiral or ellipse can be considered close to the realistic yield
curve. And it is essentially assumed because of it is simplicity in shape in defining the yield curve.
And it is close to the isotropic consolidation in the lab which can be easily captured in this shape.
It is found to be applicable for clays and non clay soils as well, even though there are deviations
for some soils it may not behave properly.
But in most of the soils, in the absence of any other model this is found to be a fair approximation.
Now the beauty of ellipse is the ratio of major to minor axis can be conveniently considered as the
shape parameter. Now when I say ellipse, we also need to know how it changes depending upon
the condition. Now these changes can be incorporated in terms of the shape parameter. And in this
case the major and minor axis becomes a very important shape parameter, we will see how?
643
Now, let us start again with this statement. The essential elements needed for any elastoplastic
models like CCM or MCCM are elasticity, yield criteria, plastic potential and hardening rule. It is
very easy to understand what is elasticity because that is what we were exposed to most of the time
during our undergraduate. And in some of the courses that we deal with in the post graduate as
well, so it is pretty easy.
And the parameters are quite simple, if we make an extra assumption that it is linear, elastic,
isotropic and homogeneous. So, this simplifies a lot of things and you are left with only 2 elastic
parameters. You can define the elastic response of the soil using 2 elastic parameters, one is
modulus and the other one is Poisson's ratio. And this modulus can be ends or Young’s or bulk
modulus that we have seen earlier, so that part is done.
Now for material like soils which exhibit plastic behaviour during shearing, we also need to know
where it would yield. Now in this particular course we have extensively discussed on how soil is
going to yield, at what point it would yield and we have discussed about yield curves. The various
aspects related to yield curve, now when I say we have extensively discussed, it is still at an
introductory level, please keep this in mind.
If you want to know more you have to read a lot further related to plasticity behaviour in geo
mechanics. So, we have just understood how to learn and where to start with? So, we have discuss
644
to that particular aspect, what is yielding and how do we define yield? And later, in the last case
in Cam Clay model, we have also discussed about how it is going to yield and what is the
boundary? That we have fixed.
So, we have clearly defined at some point the elasticity behaviour changes. Now yielding is not
failure when it comes to elastoplastic behaviour. Elastic, perfectly plastic like models like Mohr
Coulomb, we have seen that yielding itself is a failure condition. Now if you want to incorporate
hardening, softening condition, this is not possible. So, that is where the elastoplastic models are
important. Now yielding is the point where it undergo transition from elastic to plastic behaviour,
but that is not failure.
The yield curve expands, and how it will expand? That part we do not know, we know that as the
effective stress path go past beyond the yield curve, the yield curve start expanding. And when it
meets the critical state line, I mean to say the effective stress path meets the critical straight line.
That particular yield curve which is at the point of intersection of the effective stress path with the
critical straight line that particular yield curve denotes the failure one, beyond which there is no
further expansion.
Now how do we explain this many sequence beyond yielding? For that we need to know about the
plastic behaviour of the soil as well. This particular aspect, we just told in Cam Clay model but we
did not discuss this. Now putting everything in one shot is going to be difficult for you to
understand, so we did not discuss this intentionally in the last lecture. But at the same time we
need to know some basics of what happens to the soil during plastic behaviour beyond yielding?
So, this is the starting point where you can capitalize further and do further learning. So, this is
important, please understand, whatever we are going to discuss now in terms of the third one that
is the plastic potential and hardening rule which essentially state the plastic behaviour of the soil.
This part is the beginning and this shows you what to learn and from where you have to start related
to the plastic behaviour of soil?
645
And now you need to build on this for actually implementing this model into practice, so let us
start. Now yield curve is ellipse in the case of MCCM, we will be discussing everything with
respect to MCCM but that is remaining common for any other plastic behaviour as well. So, first
of all elastic behaviour of the soil, then the yielding, then the plastic response, then the failure, so
there are four stages in which we need to conceive a soil behaviour when it is loaded.
Any material will exhibit some amount of elastic behaviour, now it may be at a very small stress
range or it may be at a higher stress range depending upon the stiffness of the material. Now there
is a point at which the yielding happens and that is defined by the yield curve. Up to that stress
strain response can be defined by elastic behaviour or elastic models. Then you need to fix a point
where it will yield? That is yield criteria.
So, that also we have seen to some extent. Now beyond yielding it can undergo hardening or
softening behaviour. Now there are different models to account for this behaviour and that is
included in the elastoplastic models, various types of elastoplastic models. In that one particular
model is Cam Clay model or modified Cam Clay model. Now how this is going to proceed beyond
yield curve? That is defined by plastic potential or hardening rule.
That states the stress strain response under plastic condition. So, once the plasticity sets in how
will you define the stress strain behaviour of the soil, how it will proceed? So, that is given by this,
and finally the state boundary which is defined by the critical state defines the failure. So, these
are the four steps in which you need to understand the soil when it is subjected to shearing, so yield
curve is ellipse.
Now for initial isotropic compression, ellipse is centered on p’ axis and it passes through origin.
Now this particular definition is with respect to MCCM, how the ellipse would look like in q p’
space? So, let us first define in q p’ space, so what it tells is now yielding as I told it need not mean
failure. It can happen in isotropic compression, it can happen in one dimensional compression or
it can happen in a triaxial loading.
646
So, in initial isotropic compression, ellipse is centered on p’ axis. So, the center of the ellipse is on
p’ axis, so it is on this axis. And it passes through origin, the one side of the ellipse it passes through
origin, let us see how? So, this is how we define the ellipse, so you can see that the center is on the
p’ axis and it passes through the origin. So, this is an elliptical yield curve in MCCM, let us call it
as Yi.
And the yield curve is defined by it is p’y, that is the yield point. So, this also we know, we have
already seen this. Now this on the ICL, NCL, p’y is transferred to this particular point. Now we
need to see that this is another kind of very important aspect that the midpoint of the major axis is
here, so this is p’y/2, this particular point. And that is the point through which the critical straight
line goes, and how does that look in vp’ plot, so this point.
That is the point where the unloading-reloading line passes or it crisscrosses the critical state. It
does not mean that this meeting, but this is the location, so up to here what it means is that up to
here this particular point it is elastic response and then it goes beyond. So, here in particular the
elastic response of the soil and this is the point, so this is very specific to MCCM model that is the
yield curve, please note these are very important aspect which you need to keep in mind.
That is when the yield curve touches the critical state line that corresponds to this loading,
unloading line which is unloaded from p’y, now p’yis the extreme point of this ellipse. Now this is
true for any bigger ellipse as the effective stress path moves ahead. Definitely this particular point
will be M into p’y/2 provided if that is the failure point. Now the stress ratio that is q upon p’, if
𝑞
you call it as 𝜂 = 𝑝 ’.
𝑞
Anyway we are going to discuss this later, q by p’ if it is 𝜂; 𝑝 ’ is the stress ratio. Now this 𝜂 will
be equal to M, when the yield is corresponding to the failure condition, till that it is 𝜂. Now this
𝑀𝑝’𝑦
let us consider it to be a failure condition, in that case q is given by because this is p’ y, and
2
the slope of the critical straight line is M. So, at this particular point where it touches at the extreme
𝑀𝑝’𝑦
top point of the ellipse it is 2
.
647
𝑝’𝑦
And this distance is half of the major axis of the ellipse which is . Now it is very easy since the
2
elliptical yield curve is defined, we just need to define the equation of the ellipse, and that comes
from our basic geometry. Now equation of ellipse with center at (h, k), now this is the point h, km
this is the center. You can write it as
(𝑥 − ℎ)2 (𝑦 − 𝑘)2
+ =1
𝑎2 𝑏2
where a and b are half of major and minor axis.
So, a is major axis /2, and b is minor axis /2, because as per the equation of ellipse, and h and k
are the center. Now here h is for p’, that is p’y/2, and k it is on the p’ axis, so it will be 0. So, this
if you substitute,
2
𝑝𝑦′
(𝑝′ − 2 ) (𝑞 − 0)2
+ =1
𝑝𝑦′ 2 𝑀𝑝𝑦′ 2
(2) ( 2 )
Now what we have done is, we have incorporated some parameters which will define the yield,
and that is in terms yield curve, that is in terms of p’y and M. So, p’y and M will take care of the
shape of the ellipse. If you further simplify you can write
2 2
𝑝𝑦′
′
𝑞2 𝑝𝑦′
(𝑝 − ) + 2 = ( )
2 𝑀 2
648
If you expand the previous one. So, then we will get
2 2
′2
2𝑝′ 𝑝𝑦′ 𝑝𝑦′ 𝑞2 𝑝𝑦′
𝑝 − +( ) + 2 =( )
2 2 𝑀 2
So, we are left with the ellipse equation which is equal to
′2
𝑞2
𝑝 − 𝑝′ 𝑝𝑦′
+ 2=0
𝑀
So, this comes from this particular expression, so this is the equation for elliptical yield curve. And
that defines the boundary where the soil is going to yield. So, if you further rearrange this particular
expression, we can write
𝑀2 𝑝′2 + 𝑞 2
− 𝑝′ 𝑝𝑦′ = 0
𝑀2
Now if we divide it by p’ square, so we can write

𝑀2 + 𝜂2 𝑝𝑦′
− ′ =0
𝑀2 𝑝
𝑞
And the stress ratio 𝜂 = 𝑝′. Further we can write the another form of elliptical equation as
𝑀2 + 𝜂2 𝑝𝑦′
= ′
𝑀2 𝑝
Why I am discussing all different forms is that you will find these type of forms when you refer
to the standard textbooks. So, the equation can be rewritten as
𝑝′ 𝑀2
=
𝑝𝑦′ 𝑀2 + 𝜂2
Equation gives a set of elliptical yield curves passing through origin. And what is the kind of the
size of the elliptical yield curve is defined by p’ and M. A change in size of the ellipse happens
during yielding, because how do you consider the increased size of the ellipse.
And that increase in size is completely governed by p’y, whereas when you define p’ in terms of
p’y, and 𝑞 = 𝜂𝑝’. Now 𝜂𝑝’ is important, because the yield curve goes on increasing and that can
go on till failure. Now when it is at failure, M gets activated, so that is the final bound but the
concept of yielding keeps on happening.
649
And that will depend upon whether it is hardening or softening. Now for defining plasticity
completely, it is important to determine the plastic deformation. Now whatever we have done for
Cam Clay model that is defining of the yield curve, we have done it for MCCM as well. Now this
much we are comfortable with, now we will try to add a bit on what happens beyond yielding at
till failure.
So, that is what we will see in this, and that is to determine the plastic deformation. Now for
determining plastic deformation, we need to assume something known as flow rule, how does? It
is more like a plastic flow. What happens to the soil behaviour during shearing, what kind of flow
takes place once the yielding happens? So, that is dictated by some flow rule which we need to
assume, how much it is close to the realistic behaviour of the soil that is a different aspect.
Because for engineers we need to define it mathematically and for that we need to make certain
assumptions, so flow rule is one such kind of an assumption. Strain in soils, now where does this
comes from? Why do we need to assume? Why do we need to account for flow rule? It is because
strain in soils beyond yielding is divided into elastic and plastic strain. Now elastic strain is
determined based on the concept of linear elastic theory.
650
Plastic strain is determined, now the next aspect is how do we determine the plastic strain? Now
this can be determined based on flow rule what is known as plastic potential and then what is
known as hardening rule. So, we need to have some basics on these aspects and that is what we
will be doing in the further slides. Flow rule represents the plastic strain increment beyond
yielding, so once it yields how the plastic strain increment happens which is defined by flow rule.
Defining yield curve gives yield function, and that is an ellipse in MCCM and such an yield
function is a logarithmic spiral in Cam Clay model. Yield function is equated to 0 when soil yields
and when the plastic deformation happens. So, once it is made equal to 0 that means that it is close
to it has yielded and it is close to failure. Now yield function less than 0 is what it corresponds to
elastic behaviour that is within this yield curve.
Whatever is the state of the soil within the yield curve, we know that that is elastic condition. And
hence the yield curve function or the function of the yield curve less than 0 is supposed to be elastic
behaviour. And yield function can never be positive, that is it is up to 0 where it yields, now beyond
that it is not valid. So, greater than 0 is not possible, greater than 0 means either that particular
yield curve is no longer valid.
It goes to the next yield curve and next yield point or if you are talking in terms of failure what it
means is that, there is no soil state beyond that. So, that is why there is no question of yield function
greater than 0, it cannot be positive for a given p’y. So, that is why it is important, for that particular
extreme yield point or pre-consolidation pressure the yield function cannot be greater than 0.
But it can expand, and when it expands the earlier yield curve is no longer valid, now it is a
redefined yield curve. And if it is at a failure state if the effective stress path is touching the failure
line or the critical state line, then there is no question of soil beyond that, so that forms the
boundary.
651
So, now let us see some basics of flow rule, plastic potential and hardening rule. Yielding can
occur in isotropic one dimensional or triaxial compression. So, this notion should be very clear to
you that yielding is not meaning failure. So, it can happen in one dimensional compression,
isotropic compression where there is no shear stress developed, there also yielding can happen and
it can also happen during shearing, so yielding is different from failure.
Now this is a general representation of yield point, so this is v versus ln p’. So, this is the point at
which it yields and there is a substantial change in the v. This we know, this aspect we have already
seen. Now in the case of qp’ the yield curve, so you can see that this is the way in which the
isotropic compression yielding happens, so this is the point. The one dimensional compression
also can yield and triaxial compression in qp’ that is shearing can also yield.
Now stress path beyond yield surface. So, it means that the stress path can still move forward, that
also we have already seen in our previous lectures. When it moves beyond yield surface, means
the existing yield surface is no more valid. So, once it crosses then it is for the new yield curve,
then it keeps on extending till it fails. The existing yield surface expands or contracts depending
upon whether it is a softening or a hardening behaviour.
Expansion of yield occurs by hardening or now this hardening behaviour is expressed in terms of
isotropic or kinematic hardening. Now what we understand is that, after yielding if it is expansion
652
of yield curve that is because of hardening. Now there are 2 ways in which the hardening is defined,
that means expansion of the yield curve happens. Now that hardening is defined in terms of either
isotropic or kinematic hardening.
Now this is our yield curve, the initial yield curve in general, so it need not be a critical state or
modified Cam Clay or Cam Clay, it is a general yield curve. Now in the case of isotropic hardening,
what do you mean by isotropic hardening? It is simply an increase in the size of the yield curve,
the one which we have seen till now. And that is what we will be using because the Cam Clay
model and the modified Cam Clay model uses the concept of isotropic hardening.
Now what do you mean by isotropic hardening? It is simply the increase in size of the yield curve
as the effective stress path crosses beyond the existing yield curve. So, yield surface expands,
expands isotropically. Origin of the yield surface is stationary; this point is not going to change,
Cam Clay models that is both Cam Clay and modified Cam Clay models they are following
isotropic hardening.
Now what is meant by kinematic hardening? This is the initial yield curve, and in kinematic
hardening what happens is the whole of the yield curve gets shifted. You can see that it no longer
follows the initial origin, it is getting translated from it is earlier position. So, that is what is meant
by kinematic hardening. That is yield surface translates from this particular location to that
particular location.
It has the same shape and area, that is initial shape and it is area is not going to change, it simply
it is relocated. It translates in the direction of stress increase or how these stress path moves. No
longer at the same origin, so that is essentially the difference between isotropic hardening and
kinematic hardening. Now what is discussed in this particular course and what is not discussed in
this particular course?
We understand what is hardening, we understand what is isotropic hardening and kinematic

hardening. But we are not going to translate this into an exclusive mathematics for this particular
model. So, that is not what is dealt in this particular lecture or in this particular course, and that
653
will be the extension of your learning from taking this as the basics. So, that is about hardening
like what happens to the yield curve beyond it yields. So, now let us see, we also need to define
the plastic strain, now how do we account for plastic strain?
And for that we need to understand what is meant by plastic potential? Now beyond yielding there
will be plastic as well as elastic strain that is why it is called elastoplastic response. So, elastic
strain will also be there, and there will be quite a bit of plastic strain as well. Now we need define
or we need to account or we need to determine these strains, then only our model is complete. And
that is what we have not doing in this particular course, we are yet to define the strain completely.
But we will discuss like what we need to have further, and that is both elastic and plastic strain.
So, let us take a typical 𝜎 versus 𝜖 response, this is elastic response, then it yields and there is a
bilinear response which is shown here, 𝜎’𝑦 is the yield point. And this is a typical hardening
behaviour, and this is d𝜖, that is beyond the yield what is the total strain that is happening.
Now if I expand this and show here, this is the total strain that is happening beyond yield, this
much portion is elastic strain and that is represented by dϵe. And the plastic strain is represented
by dϵp, so total gives the dϵ which is the total strain, dϵe is the recoverable elastic strain, dϵp is
irrecoverable plastic strain. And the total elasto-plastic strain dϵ = dϵe + dϵp Now to define this
further I mean to say the strain, we need to make now further assumptions.
654
In modified Cam Clay or rather in Cam Clay models co-axiality is assumed. Now this is another
assumption that we need to have in order to determine the strain. That means this particular aspect,
the plastic strain that is dϵp it will have components in q and p’. That means the plastic strain due
to volumetric behaviour or volumetric changes and plastic strain due to deviatoric changes and
that is represented.
And that plastic strain is decomposed into the component due to q and the component due to p’.
𝑝
And this is designated as 𝑑𝜖𝑞𝑝 and 𝑑𝜖𝑝′ . Now this is due to the plastic strain due to q, and this is
the plastic strain due to p’. So, this is the typical yield curve. Now this is what it represents 𝑑𝜖𝑞𝑝
𝑝
and 𝑑𝜖𝑝′ . So, this is corresponding to p’ and this is corresponding to q, now who has told that this
particular increment is in this particular direction?
That means it is in the direction of p’, this is in the direction of q, who has told this? Now this is
represented by how the strain increment vector happens, that is the plastic strain increment vector
which is represented by 𝑑𝜖̂𝑝 with a crown. So, this is the plastic strain increment vector, now this
plastic strain increment vector or it is knowledge is quite essential for defining the plastic strain.
So, plastic potential followed by plastic strain increment vector, these are 2 essential elements. But
the question remains who has told that this is in this direction? Now in co-axiality, that is the
𝑝
assumption of co-axiality, we assume that 𝑑𝜖𝑞𝑝 and 𝑑𝜖𝑝′ is in the direction of q and p’ respectively.
Now that is the importance of co-axiality assumption. That means whatever is the strain that
happens in the direction of stress, so that is what is mean by co-axiality. Coordinate axes of stress
and strain are superposed in co-axiality assumption.
You can see that this particular axis and this particular axis which denotes the plastic strain is
superposed with q and p’ axis respectively, so this assumption is what is known as co-axiality. So,
when you start learning plastic behaviour of soil or if you want to learn more mathematically, you
will come across these terminologies. So, it will be very easy for you to understand further when
you know these terminologies.
655
So, the assumption of co-axiality is valid for MCCM model, so for us we are discussing this with
respect to critical state models. So, let us further go ahead, so now we have defined yield curve,
we have defined the co-axiality assumption. And then we have seen the hardening rule that is
isotropic or kinematic, now these are some of the elements. So, now we need to move further.
Plastic potential is a function that is assumed to exist in stress space. Now we need to define plastic
strain increment vector. Now for that we also need to define what is known as plastic potential,
now what is plastic potential? It is a function that is assumed to exist in stress space. Now what is
the duty of this function? It is a scalar function which when differentiated with respect to that
particular stress gives plastic strain.
Now you understand how do we get plastic strain, it is very essential to define plastic potential.
Because it is a scalar function and when you differentiate that particular scalar function with
respect to the corresponding stress or with respect to a specified stress, it gives plastic strain
corresponding to that particular stress condition. So, define plastic potential and get plastic strain.
𝑝
So, let us say this is a plastic potential contour, this we have already defined, so 𝑑𝜖𝑞𝑝 and 𝑑𝜖𝑝′ and
this is the plastic strain increment vector.
656
Plastic potential contour describe the direction of plastic strain increment vector, so very essential
to define the plastic potential, so that this is obtained and this is also defined. The change in yield
surface is related to plastic volumetric strain, and that is called hardening rule. That we have seen
the expansion of yield curve is in terms of p’y and that comes from the hardening behaviour, and
that comes from the plastic volumetric strain.
Defining specific mathematical function for plastic potential is required for determining plastic
strain during shearing. So, we need to define or we need to allocate certain function for this plastic
potential, then only we will be able to determine plastic strain. Now this opens up another concept
which is very essential in plastic behaviour of soil.
And that is what is known as associated and known associated flow rule. How do you define plastic
potential? So, there it has something to do with this particular concept which is known as
associated and non associated flow rule. When plastic potential function is considered same as
yield function, it is called associated flow rule, very easy. Like what we are considering is we have
already defined the yield function which is ellipse or logarithmic spiral or any other shape.
What we define is that the plastic potential function is same as the yield function, when you make
such an assumption this is known as associated flow rule. The plastic flow rule is associated with
the yield curve. Now it makes our life very simple, like we need to work with only one function
657
and remaining things are known like how to evolve the plastic strain. Of course, I am not discussing
that in this particular course, but definition of plastic potential is important.
So, we have already done that in the case of associated flow rule. For isotropic material, the axis
of principle stresses and principle strain are coincident, that is co-axiality, we have already
𝜕𝑓
discussed. For associated flow rule, yield function F is same as plastic potential g. And 𝜕𝜎 is in the
direction normal to yield curve, so that is also very important. So, that is how it looks like q versus
p’ plastic potential which is same as the yield curve in the associated flow rule.
So, this is yield curve as well as plastic potential. And we have the plastic strain increment vector
and that is normal to the yield curve. So, this is 90 degrees, so that is normal to yield curve. Now
plastic strain incremental vector is outward normal to yield curve. So, this is outward normal to
yield curve, so such an assumption is called normality rule, what it states? The plastic strain
increment vector, this direction is perpendicular to the tangent at this particular point.
So, such an increment that is at 90 degrees it is outward normal to yield curve, this is called
normality rule. So, that is also need to be satisfied and hence yield curve and plastic potential both
are same, as you can see. And this is the stress path, and the plastic strain increment vector is given
in this manner. Now when you define this mathematically that leads to the determination of plastic
strain.
So, normality rule may sometimes overestimate the plastic volumetric strain, may not be realistic
that is what it means. But these are this makes our life simple, that means the mathematical
complexities reduced by such assumption. But always keep in mind there is a possibility that it
may overestimate plastic volumetric strain. Now Cam Clay models both Cam Clay and modified
Cam Clay obey associated flow rule and normality rule.
So, both associated flow rule and normality rule is applicable for these models and hence our
discussion is mostly associated with that.
658
What is meant by non-associated flow rule? When we consider yield function is different from
plastic potential or rather plastic potential g is different from F. So, then we say it is a non-
associated flow rule. So, yield curve and the plastic potential they are different. For simplicity,
associated flow rule is mostly preferred. For more realistic behaviour a non-associated flow rule
is preferred.
Now further reading on this, now I cannot go beyond this because for this particular course this is
mostly an introductory aspect of plasticity behaviour. There has to be an exclusive course on plastic
behaviour of soil and for that the readers or the participants are advised to go through this particular
reference book and any other reading material on plasticity behaviour of soils.
659
Now some important aspect related to plastic strain increment and with respect to the critical state.
Let us say q p’, now stress path to failure is defined. Now let us say this is the starting point of
shearing, this is the critical state line and we can define the yield curve. The yield curve is small;
ESP has not reached the critical state line. Now at this particular point if we try to mark the plastic
potential or plastic strain increment vector, how it will look like?
Now the plastic strain increment vector is not parallel to q axis, that means the plastic volumetric
strain increment is not 0, I will explain what actually is this? Now you can see that this is the plastic
strain increment vector, the direction, at this particular point where the effective stress path is
touching the yield curve. Now this has not failed, now at this particular point the direction of this
that is the plastic strain increment vector, this is not parallel to q axis.
Now if it is not parallel to q axis, it means that plastic volumetric strain increment is not 0. Now
what is the implication of this statement plastic volumetric strain increment? We will see what is
the implication of this statement? This is the plastic strain increment vector. Now let us say the
ESP reaches critical state line, this is the condition and that is defined by the yield curve or the
elliptical yield curve.
Now at this particular point what happens is you cannot further increase the yield curve. Now we
have seen in the previous slide, that increment of the yield curve is dictated by the plastic
660
volumetric strain increment that is this one. Now once it has reached the critical state line soil has
failed, there is no more question of existence of the soil beyond that. That means the plastic
volumetric strain increment cannot happen beyond this.
From here it is possible, the plastic volumetric strain increment happens as the effective stress path
moves further. But, once it reaches this there is no more, it has failed, there is no more soil beyond
that particular point. So, the plastic volumetric strain increment becomes 0 at this particular point.
𝑝
And when that becomes 0 that is 𝑑𝜖𝑝′ that becomes 0 means it has to be parallel.
That means this plastic strain increment vector it has to be parallel, that is yield curve means the
point of intersection that is what I told and 0 gradient at this point on the yield curve. What is
meant by 0 gradient? At this particular point it is 0 gradient that is at the point where it, so here the
slope is 0 and the plastic strain increment vector is parallel to q axis. Now this is the situation
𝑝
where only 𝑑𝜖𝑞𝑝 exist, 𝑑𝜖𝑝′ it does not exist.
That means plastic volumetric strain increment is 0, only plastic shear increment happens
indefinitely, now this is indefinite. Now we kept on telling this right from the day one when we
started discussing critical state, at critical state the soil shears indefinitely. Now you will clearly
appreciate what is meant by the statement in this particular slide? Like this plastic strain increment
vector in the q direction means it is shearing indefinitely is infinite.
𝑝
Whereas , 𝑑𝜖𝑝′ which shows the plastic volumetric strain increment, that becomes 0 and there is
no further change in stress because soil does not exist beyond this point. So, 2 conditions are
satisfied when the effective stress path reaches the critical state line and for the yield curve which
𝑝
is defined by this. So, here, 𝑑𝜖𝑝′ = 0 and that is exactly what is meant by this condition.
𝑝
Now for this particular plastic strain increment vector, there is , 𝑑𝜖𝑝′ , that is possible, that is why
it is expanding. But at this point it stops, it becomes 0, and that is the condition where this plastic
strain increment vector is parallel to q. No further volumetric change, so there are 2 things which
661
are satisfied here. And no hardening possible, so I will reach ultimate state of failure and no change
in stress.
So, no volumetric change and no change in stress but shears indefinitely. So, this particular figure
shows exactly what is happening at critical state. Now if you want to understand this, we need to
understand the basics of plasticity behaviour what we just discussed. So, that is all about MCCM
and some basic elements of plastic behaviour which is very important. And now let us try to
summarize what we have learned in this lecture.
MCCM is a popular elasto-plastic model that can incorporate hardening, softening during shearing.
In MCCM the shape of the yield curve is ellipse, but we have talked about hardening only, it is
basically an isotropic hardening model. So, in MCCM the shape of the yield curve is ellipse and it
is logarithmic spiral in CCM. It may not be applicable for all soil type, as I told when we do a
mathematical modelling or mathematical definition we cannot make it generalized.
So, there will be some cases where it may not apply. For ellipse the ratio of major to minor axis is
considered as the shape parameter, how? It accounts for further expansion. The essential elements
needed for elasto-plastic models are elasticity, yield criteria, plastic potential, hardening rule, all
of them the basic definition and what it actually means? We have discussed. Flow rule represents
plastic strain increment beyond yielding.
662
Cam clay models follow isotropic hardening so we have these are basically hardening models. In
MCCM or Cam Clay models it follows coaxiality which is important for defining the plastic strain
increment vector. Plastic potential is a function defined for determining plastic strain. In critical
state models plastic potential function is taken equal to yield function which is known as associated
flow rule.
So, plastic strain increment vector is outward normal to yield curve, which is normality rule. For
more realistic soil behaviour a non associated flow rule is preferred. The slope of yield curve is 0;
please understand under what condition, where it meets the point of intersection of ESP with
critical state line. And that is why at that particular point the plastic strain increment vector is in
the upward direction.
Plastic strain increment vector is parallel to q axis at this particular point. At this point plastic
volumetric strain increment is 0, indefinite shearing happens at constant volume and stress. The
application of CCM and MCCM requires the definition of complete stress strain constitutive
relationship. Now this course does not deal with the determination of elastic and plastic strain. We
have defined the basic elements of plasticity behaviour but the complete constitutive relationship
beyond the yielding we have not defined.
663
That means the plastic strain we are not actually determining in this particular course. So, that is
all for the critical state models. Now we will see a bit about prediction of critical state behaviour
in the next lecture, so that is all for now, thank you.
664
Lecture-51
Prediction of Soil Behaviour from MCCM 1
Welcome back, all of you. We were discussing about critical state models and we have finished
Cam Clay model and modified Cam Clay model. And most of the numerical modelling adopts
mostly modified cam clay. So, in the next part of the lecture will be how to predict the failure state
of soil, based on the critical state model specific to modified Cam Clay and the procedure remains
the same.
So, we will have 2 lectures in this, the first lecture will be a general discussion on how to predict.
And the next lecture will be specific to most of the matter remains the same, but then we will
discuss with respect to consolidated drain, consolidated undrained for NC and HOC soil. So, this
is what we will be doing the next 2 lectures, so it is all about prediction of failure state from
modified Cam Clay model.
665
So, just to refresh, yield curve in modified Cam Clay model is ellipse. We will not go into the
details of it, this has been discussed. We have q p’ axis, the first yield curve is marked, the yield
stress marked as 𝑝𝑦′ , the point where the critical state meets this yield curve, that is marked as M
′
𝑝𝑦 ′
𝑝𝑦
and we know that this is . The equation is given
2 2
𝑞2
𝑝′2 − 𝑝′ 𝑝𝑦′ + =0
𝑀2
Soil is isotropically consolidated to 𝑝𝑦′ , this one, so this point. Same is shown here p’ y from where
the unloading can happen. Soil is unloaded to p0’ which is a lightly over consolidated state. So,
somewhere here, so that it is slightly over consolidated, so p0’, this is the unloading line, so you
can see that it unloads to p0’. Now I is the intersection of unloading line with CSL, this is the point.
And that corresponds to this particular point, I is the point of intersection of yield curve with CSL,
now all these information we have already discussed. And we also have the condition, so without
knowing MCCM, this lecture will not be good for you to understand. So, what are the properties
′
𝑝𝑦
according to MCCM, we know that 𝑝𝐼′ = , because these informations are needed for formulating
2
the prediction
V at p0’ is v0, so this is the starting point because we are starting with an lightly over
consolidated state, so the v corresponding to p0’ is v0, so this one. We can write this particular
point of p0’as
𝑣𝑜 = 𝑣𝑘 − 𝜅 ln 𝑝𝑜′
from the equation of unloading reloading line. Now this point vI, we can represent in terms of
unloading, reloading line as well as in terms of critical state line.
So, in terms of unloading, reloading line we can write

𝑝𝑦′
𝑣𝐼 = 𝑣𝑘 − 𝜅 ln
2
666
We have
𝑣𝑂 = 𝑣𝑘 − 𝜅 𝑙𝑛p′o
𝑝𝑦′
𝑣𝑡 = 𝑣𝑜 + 𝜅 ln p′o − 𝜅 𝑙𝑛
2
from where it comes from we are substituting it for 𝑣𝑘 . And vI can be written as
p′o
𝑣𝑡 = 𝑣𝑜 + 𝜅 𝑙𝑛 ′
𝑝𝑦
2
Now based on critical state line we can also write
𝑝𝑦′
𝑣𝑡 = 𝛤 − 𝜆 𝑙𝑛
2
the same point but from critical state line. We can substitute it for vI, so when you substitute for vI
we can write
𝑝𝑦′ p′o
𝛤 − 𝜆 𝑙𝑛 = 𝑣𝑜 + 𝜅 𝑙𝑛 ′
2 𝑝𝑦
2
So, we can write gamma as

𝑝𝑦′
𝛤 = 𝑣𝑜 + (𝜆 − 𝜅) ln + 𝜅 𝑙𝑛 𝑝𝑜′
2
Now the knowledge of p 0’, p’ y and v 0 is required for MCCM prediction.
667
Now if the initial point is p’y, so whatever we have discussed is for the lightly over consolidated
point. Let us consider this particular point itself that is the normally consolidated point, so this is
the point, this is normally consolidated. And v is vy, so the initial condition is vy, so if that is the
case then vI can be written as
𝑝𝑦′
𝑣𝐼 = 𝑣𝑜 + 𝜅 ln p′o − 𝜅 𝑙𝑛
2
Now here p0’ it is written in general, now our starting point here is p0’ = py’.
So, here vI can be written as instead of v0 the starting point is vy because it is a normally
consolidated point. So, normally consolidated point it is vy here,
p′o
𝑣𝐼 = 𝑣𝑜 + 𝜅 𝑙𝑛
𝑝𝑦′
2
So, we can write
𝑣𝑡 = 𝑣𝑜 + 𝜅 𝑙𝑛2
Also
𝑝𝑦′
𝑣𝑡 = 𝛤 − 𝜆 𝑙𝑛
2
668
If we substitute we can from the same procedure for LOC, we are doing it only the starting point
is different. So,
𝑝𝑦′
𝛤 − 𝜆 𝑙𝑛 = 𝑣𝑦 + 𝜅 𝑙𝑛2
2
𝑝𝑦′
𝛤 = 𝑣𝑦 + 𝜆 𝑙𝑛 + 0.693𝜅
2
Now gamma can be predicted based on the given conditions. If we know vy, py’ and 𝜅 which are
the critical state parameters gamma can be determined. Now once gamma is known then the
particular critical state line is also defined.
Now let us come to specific of drained triaxial test. Like what are the failure stresses from MCCM.
Now we know what are the failure stresses? we are just trying to understand how we will determine
the failure stresses based on the given MCCM model. Now where is the role of MCCM coming
into picture, because the yield curve matters? So, with respect to any effective stress path yield
curve matters, so that is defined in addition to what we have done in module 2, module 3 onwards.
So, here we are adding another component while defining the failure state of soil and that is yield
curve. Again soil is isotropically consolidated to py’; unloaded to p0’, let us consider a case of LOC
in general. So, we have py’, we have defined the yield curve corresponding to py’, and critical state
line, and then p0’, up to here there is no need for discussion. Now whatever is needed we have
discussed till our last lecture, now it is only summing up.
669
So, please follow the previous lectures very properly, so that we will be a bit fast in this lecture
and hence you will be able to pick up further. Slope of ESP = TSP because it is a drained triaxial
’
test. Now ESP from p0’ yield at 𝑝𝑜𝑦 , when ESP touches the yield curve, so it is this particular point
where it yields and the slope is 3, this information comes from our third module. Now this is the
’
yield stress that is 𝑝𝑜𝑦 , where it touches the yield curve.
Now remember the same discussion we have done in module 3. We do not have any idea about
where it is going to yield; we just talked about the final state or the failure state where the ESP
touches the critical state line. Now we are adding one more information that is the yield stress,
’
now where the ESP touches the yield curve, that is 𝑝𝑜𝑦 . Deviator stress at yield is 𝑞𝑦 , so that 𝑞𝑦 is
given. Soil fails when ESP touches CSL at F, so this is the point F. Failure stresses are 𝑝𝑓′ and 𝑞𝑓 .
So, all these can be determined, like we can determine 𝑞𝑦 , 𝑝𝑓′ , 𝑝𝑜𝑦
′
is written because the initial
point is 𝑝𝑜′ and the failure stress 𝑞𝑓 , and 𝑝𝑓′ , all can be determined. Now looking at the geometry
of this figure; that is considered this triangle 𝑝𝑜′ and 𝑝𝑓′ . We can always write
𝑞𝑓 = 3(𝑝𝑓′ − 𝑝𝑜′ )
So it is very conveniently we can write this. And we know
𝑞𝑓 = 𝑀𝑝𝑓′
670
So, then this can be substituted and written
𝑀𝑝𝑓′ = 3(𝑝𝑓′ − 𝑝𝑜′ )
So, then rearranging, we can write
3𝑝𝑜′ = (3 − 𝑀)𝑝𝑓′
From which the failure stress can be determined that is
3𝑝𝑜′
𝑝𝑓′ =
3−𝑀
You can see this particular equation where 𝑝𝑓′ is the required failure stress now that is a function
of 𝑝𝑜′ which is the initial condition and M which is a critical state parameter.
So, once we know 𝑝𝑓′ , we can determine q, 𝑞𝑓 . Now for one additional condition which we need
to discuss here is for M = 3, 𝑝𝑓′ = 𝑞𝑓 = ∞. If you check here what will happen to 𝑝𝑓′ when you put
M = 3, so this will become infinity, so that is 𝑝𝑓′ = 𝑞𝑓 = ∞.. That means if at the particular point
3, if you remember our discussion earlier like we have discussed about tension cut off zone from
the origin.
When we discussed about this earlier in our previous lecture, we told that this same thing will be
discussed later how you are drawing a tension cut off zone at an inclination of 3 from the origin.
Now what will happen if M is greater than 3? If M is greater than 3 you will have 𝑝𝑓′ to be negative.
671
So, for a line passing through origin and M greater than 3 both 𝑝𝑓′ and 𝑞𝑓 will be negative, and that
is the reason and the soil cannot sustain tension.
So, hence M greater than 3 is an impossible state, and that impossible state is at an inclination of
3 we have drawn. Now anything more than this is not possible because it will be negative and soil
cannot sustain tension, and that is the reason we have defined it as a tension cut off zone. another
state boundary on the left side. So, that is the boundary beyond which the soil will not exist.
Now I hope the explanation is very clear and it is justified how we have drawn an inclination of 3
at origin and marked it as a tension cutoff or an impossible state of the soil. So, that is otherwise
known as tension cut off zone.
Now for the same case undrained triaxial test for lightly over consolidated state. Soil is again the
same situation 𝑝𝑦′ unloaded to 𝑝𝑜′ lightly over consolidated state, so 𝑝𝑜′ . The same is marked in the
figure, so the initial condition is fixed. Now this is an undrained test, the very important fact that
𝛥v or the specific volume there is no change in undrained test. When this is the condition we know
that the initial specific volume and final specific volume they are same.
Now this makes a lot of difference in our mathematical computation of failure stress. So, here the
failure happens when it meets the critical state line, and that F is the failure point. So, 𝑣0 = 𝑣𝑓 ,
672
3 1
now slope of 𝐸𝑆𝑃 = 1−3𝐴 and we know that till yielding it is elastic 𝐴 = 3 in fact this particular
aspect we have discussed several number of times. Now what is happening is some bit of
information is getting added all the time.
So, it becomes almost very clear when we repeat these facts but make sure like we are pointing
towards the same idea. Whatever we have discussed the same ideas we are discussing. But in this
particular case we are adding yielding coming into picture, but all these facts remain same. So,
now for if it is elastic behaviour, A has to be 1/ 3, when you substitute it, it will be infinity, so it
will be vertically upwards, so ESP is vertically upwards.
′
Now once it yields this is an undrained condition, so the yield stress is 𝑝𝑜𝑦 and 𝑞𝑦 . Now after this
it yields and then what happens, ESP shifts towards leftwards due to positive pore water pressure,
this is an undrained test. So, here the effective stress path moves leftwards, so this is due to positive
pore water pressure, fails when ESP meets critical straight line is F. So, we are not going to spend
much time here because all these facts are known. Now failure stresses are 𝑝𝑓′ and 𝑞𝑓 which need
to be determined. So, 𝑞𝑓 and 𝑝𝑓′ , so this is the failure stress.
That is for undrained test for lightly over consolidated. Now let us see undrained traction tests for
heavily over consolidated. So, soil is isotropically consolidated to 𝑝𝑦′ and then unloaded to 𝑝𝑜′ , the
673
same procedure but the unloading is much higher. So, 𝑝𝑦′ , yield curve, critical state line, now you
can see that it is slightly towards the left end because it is heavily over consolidated.
The same is marked on v ln p’ plot, you can see that this much is the over consolidation, that is
3
happening 𝑣𝑜 = 𝑣𝑓 , Δv = 0 in undrained test, so 𝑣𝑜 = 𝑣𝑓 slope of 𝐸𝑆𝑃 = 1−3𝐴 ESP till yielding
is elastic, the same information yield when it touches the yield curve. So, up to here it is all clear,
′
so this is 𝑞𝑦 and 𝑝𝑜𝑦 , the yield stresses. After yielding ESP shifts rightwards due to negative pore
water pressure, again it is a known fact fails when ESP meets critical state line at F. So, that is F
it is undrained test, so it meets at F, and 𝑝’𝑓 and 𝑞𝑓 are the failure stresses.
Now equation of MCCM yield curve is this. Now we know that this effective stress is meeting the
yield curve. So, if you substitute this particular point,
′2
𝑞2
𝑝 − 𝑝′ 𝑝𝑦′ + 2=0
𝑀
𝑞𝑦2
𝑝𝑜′2 − 𝑝𝑜′ 𝑝𝑦′ + 2 = 0 (where ESP meets the yield curve)
𝑀
In this particular case 𝑞𝑦 is defined by the yield curve, can also be considered as q peak. Now keep
𝑞𝑝𝑒𝑎𝑘 is the state boundary for heavily over consolidated soil. You might be remembering this
very well, like in our previous lectures we have discussed with respect to peak point of for a heavily
over consolidated soil. So, what happens, the soil the stress strain response increases, it reaches
the maximum that is the peak point and then strain softens.
674
Now what we have drawn is, we have drawn a peak line and then it touches. So, the maximum
point to which the soil can reach as a heavily over consolidated soil can reach is the peak point.
So, here in this particular case we are defining it by means of a yield curve. So, here if you consider
yield curve to be same as the state boundary surface then you can always approximate yield point
to be that of peak point, in fact that is depending upon the critical state model.
Now whether the soil actually yields at that peak point, this is not very clear, it can yield slightly
before or maybe slightly after or maybe at peak, this is not very clear. But here we are assuming
the yield point to be same as the peak point. Because both serves as the maximum point, state
boundary point, so that is why. So, 𝑞𝑝𝑒𝑎𝑘 is the state boundary and hence 𝑞𝑦 that is the yield point
this particular point can be considered equal to 𝑞𝑝𝑒𝑎𝑘 .
If that is the case we can write

𝑞𝑦2 𝑝𝑦′
= −1
𝑝𝑜′ 2 𝑀2 𝑝𝑜′
So, then 𝑞𝑦 that is the yield deviatoric stress can be written as
𝑝′
𝑞𝑦 = 𝑀𝑝𝑜′ √𝑝𝑦′ − 1.
𝑜
′
𝑝𝑦
So, this is a quick way of computing 𝑞𝑦 , and we know that = 𝑂𝐶𝑅 with respect to the initial
𝑝𝑜′
point. So, you can replace

𝑞𝑦 = 𝑀𝑝𝑜′ √𝑂𝐶𝑅 − 1 𝐹𝑜𝑟 𝑂𝐶𝑅 > 2
and this is mostly valid for heavily over consolidated soil where OCR is greater than 2, it is not
mostly valid, it is valid for heavily over consolidated soil.
675
Now let us discuss about unconsolidated undrained triaxial test. We know
𝑣𝑜 = 𝑣𝑓 = 𝛤 − 𝜆𝑙𝑛 𝑝𝑓′
because it is an undrained test. Now please refer back if you are finding it difficult, now we are
not going to repeat it. So, 𝑝𝑓′ can be written as
𝛤−𝑣𝑜
𝑝𝑓′ = exp and 𝑞𝑓 = 𝑀𝑝𝑓′
𝜆
Undrained shear strength

𝑞𝑓 𝑀 𝛤 − 𝑣𝑜
𝑆𝑢 = = 𝑒𝑥𝑝
2 2 𝜆
For a given soil M, 𝛤 , 𝜆 are constants. So, what 𝑆𝑢 is merely a function of v 0 that is the initial
specific volume or we can write in terms of initial water content, wo is the initial water content.
676
Now let us say unconsolidated undrained test for NC sample. If it is 𝑣𝑜 it is the same equation we
are considering,
𝑣𝑜 = 𝑣𝑓 = 𝛤 − 𝜆𝑙𝑛 𝑝𝑓′
we have
𝛤 − 𝑣𝑜
𝑝𝑓′ = exp
𝜆
But we also have since it is an NC sample we can also write
𝑣𝑜 = 𝑁𝑜 − 𝜆 ln 𝑝𝑜′
Now if you substitute for 𝑣𝑜 here,
𝛤 − 𝑁𝑜 + 𝜆 ln 𝑝𝑜′
𝑝𝑓′ = exp
𝜆
for 𝑝𝑓′ can be written as
Now 𝑆𝑢 is also equal to
𝑀 𝛤 − 𝑣𝑜
𝑆𝑢 = 𝑒𝑥𝑝
2 𝜆
So 𝑆𝑢 can be written as
𝑀 𝛤 − 𝑁𝑜 + 𝜆 ln 𝑝𝑜′
2 𝜆
677
𝑀 𝛤 − 𝑁𝑜
𝑆𝑢 = [( ) + ln 𝑝𝑜′ ]
2 𝜆
Knowing 𝑝𝑜′ , 𝑆𝑢 can be determined because all others are the part of critical state parameters.
For a given soil, these are all constants. 𝑝𝑜′ increases with depth, why?
Because 𝑝𝑜′ is a mean vertical stress, so mean stress condition and with depth mean stress condition
increases. So, since 𝑝𝑜′ increases with depth the 𝑆𝑢 for normally consolidated soil increases with
depth. So, if you want to prove that 𝑆𝑢 of NC will always increase with depth, this is the proof and
𝑝𝑜′ increases with depth. And according to this equation 𝑆𝑢 is merely a function of 𝑝𝑜′ because all
others are constants.
Now the same exercise for OC sample

2 𝜆
𝑣𝑜 = 𝑣𝑘 − 𝜅 ln 𝑝𝑜′
Because this is an unloading-reloading case. Substitute for 𝑣𝑜
𝑀 𝛤 − 𝑣𝑘 − 𝜅 ln 𝑝𝑜′
2 𝜆
Again like previous case we can write
𝑀 𝛤 − 𝑁𝑜 𝜅
𝑆𝑢 = [( ) + ln 𝑝𝑜′ ]
2 𝜆 𝜆
678
Also we have
2 𝜆
we can substitute for 𝑣𝑜 that is equal to
𝑀 𝛤 − 1 − 𝑒𝑜
2 𝜆
what is 𝑒𝑜 ? 𝑒𝑜 can be substituted as 𝐺𝑤𝑜 because Sre = G w. So, based on that
𝑀 𝛤 − 1 − 𝐺𝑤𝑜
2 𝜆
So, let us try to understand why we have done the previous expression is to understand what is the
influence. How much is the influence of initial water content on undrained shear strength? Just to
prove that,
𝑞𝑓 𝑀 𝛤 − 𝑣𝑜 𝑀 𝛤 − (1 + 𝑒𝑜 )
𝑆𝑢 = = 𝑒𝑥𝑝 𝑆𝑢 = 𝑒𝑥𝑝
2 2 𝜆 2 𝜆
Let us consider 2 samples A and B of the same soil with 1% difference in initial water content wo.
So, we are just trying to see this 1% difference in initial water content will translate to how much
difference in S u. So, we are considering the critical state parameters or in the critical state
parameters 𝜆 = 0.15 and specific gravity as 2.7.
679
𝑀 𝛤 − (1 + 𝑒𝑜𝐴 )
𝑆𝑢𝐴 = 𝑒𝑥𝑝
2 𝜆
And 𝑆𝑢𝐵 is 𝑒𝑜𝐵 only, the initial void ratio there is a difference.
𝑀 𝛤 − (1 + 𝑒𝑜𝐵 )
𝑆𝑢𝐵 = 𝑒𝑥𝑝
2 𝜆
𝑆
Now if you divide 𝑆𝑢𝐴 , we can write exp,
𝑢𝐵
𝑆𝑢𝐴 𝛤 − (1 + 𝑒𝑜𝐴 ) − (𝛤 − (1 + 𝑒𝑜𝐵 ))

= 𝑒𝑥𝑝
𝑆𝑢𝐵 𝜆
So, here you have
𝑆𝑢𝐴 𝑒𝑜𝐵− 𝑒𝑜𝐴
= exp
𝑆𝑢𝐵 𝜆
So, substituting for 𝑒𝑜 = 𝐺𝑤𝑜 we have
𝑆𝑢𝐴 𝐺(𝑤𝑜𝐵 − 𝑤𝑜𝐴 )
= exp
𝑆𝑢𝐵 𝜆
Substitute for G that is 2.7, the difference in water content is 1%,
𝑆𝑢𝐴 2.7 − 0.01
=
𝑆𝑢𝐵 0.15
𝑆𝑢𝐴
= 1.19 ≈ 1.2
𝑆𝑢𝐵
So, 1% difference in wo results in 20% difference in 𝑆𝑢 , so that is the conclusion. So, a very small
difference in wo will be a glaring difference in 𝑆𝑢 . So, natural water content in the samples should
be preserved as far as possible.
When we are doing sampling from the field, it is very important that is why it is told that it has to
be packed properly. And hence we can preserve what is the natural moisture content or field
moisture content as it is otherwise it is going to influence the strength which is determined. Now
here it is influence on undrained strength is quite high. So, if we do not take care of initial water
content, then 𝑆𝑢 would considerably vary and we will not be getting the condition that is existing
in the field.
680
So, that is all for the initial part of prediction. So, we have tried to understand the various equations
that can be used for finding out the failure stresses.
So, let us try to summarize today's lecture. The relationship for predicting failure state in MCCM
is derived. Failure state in drain triaxial test is dependent on initial stress state and M parameter.
Equation for peak deviatoric stress or deviator yield stress is expressed in terms of OCR for heavily
over consolidated soil. So, we have also found out that 𝑞𝑦 and 𝑝′𝑦 also can be determined for any
given type of situation.
Undrained shear strength 𝑆𝑢 is expressed in terms of critical state parameters for NC and HOC.
Undrained shear strength S u of NC increases with depth that we have shown. 𝑆𝑢 is highly sensitive
to initial water content, so that is important to preserve natural moisture content. So, that is all for
today's lecture, we will be continuing the same in the next lecture with more details added to it.
All the discussions remains the same but we will also try to see the prediction of 𝐶𝑢 𝑎𝑛𝑑 𝐶𝑑 for
NC and OC material or soils and with respect to the volume change or the pore water pressure also
added to it. So, we will see this in the next lecture, that is all for now, thank you.
681
Lecture-52
Prediction of Soil Behaviour From MCCM 2
Welcome back all of you, today's lecture is the continuation of last lecture which is basically
prediction of failure conditions from MCCM. In fact in the last few lectures what we have seen is,
we have got the concept of strength, we have got the concept of stress path then we integrated
yielding and yield curve with these concepts. And then saw how a soil state from it is initial state
reaches to the critical state.
So, in the process it gets yielded and then it reaches the critical state. So, both the yielding
conditions and the failure conditions can be obtained from the given critical state models. In some
sense by knowing the critical state parameters which we have defined all the while in the last few
lectures, and by knowing the geometry by substituting the required parameters into the equation
of ellipse which is the yield curve and corresponding to the failure conditions.
So, this much we have seen, now today's lecture we will be adding a delta information to what we
have already seen. That we will try to integrate how this looks like, the whole of the failure looks
like from the stress strain response and the volume change or pore water pressure, depending upon
whether it is a drained or undrained test. So, that is what we will be seeing in today's lecture.
682
So, it is about prediction of soil failure from MCCM. So, first let us see prediction of normally
consolidated and LOC for CD test. It is more or less the same, much of these aspects we have
covered in the last few lectures. So, it should not be a problem, so it will be more like a revision
come understanding. So, I think it should be easy for you to consolidate what all you have
understood till now with this lecture.
Let us start with, because in the critical state model what we will be dealing with will be MCCM
that is the ellipse. So, the equation of the ellipse
′2
𝑞2
𝑝 − 𝑝′ 𝑝𝑦′ + 2=0
𝑀
Now for NC or LOC because both of them exhibit more or less similar behaviour and we have
seen that OCR less than 2 can be considered as LOC. And when it is greater than 2 in our previous
lectures we have defined it as 4.
So, this is not a very strict definition of OCR mostly we will come to know by it is response and
the type of the soil. So, OCR what is the actual definition for LOC and HOC, that would differ and
it is not a very strict rule of thumb. So, soil is isotropically consolidated to 𝑝’𝑦𝐴 , so this is 𝑞𝑝 ’ plot.
So, this is the maximum yield point and that comes from the isotropic consolidation.
683
Soil is now unloaded to 𝑝𝑜 ’, that is point O, why? To create a LOC state, so this is 𝑝𝑜 ’. And the
yield curve corresponding to this will be this, so that is the maximum point is 𝑝’𝑦𝐴 . Now critical
state line has been marked and according to MCCM it passes through the center, so it is p. Now
we have v ln p’ plot, so this is the isotropic consolidation point 𝑝’𝑦𝐴 . Let us say it is unloaded from
𝑝’𝑦𝐴 to generate 𝑝𝑜 ’, so this is point p and here you can see it is 𝑝𝑜 ’, so it is unloaded to 𝑝𝑜 ’.
So, this is very essential, now to find out to understand this sequential steps. Now in our previous
lectures these were not very strictly followed, I will tell you one more aspect which we did not
follow very strictly. But now things are very clear when we integrate the yield curve along with
the stress path, so now 𝑝𝑜 ’ is on the unloading curve. As I told we will try to also integrate q versus
𝜖 behaviour and volume change.
Because now it is a CD test, so it is volume change Δ𝛴𝑣 which is cumulative volume change. So,
these many details we are going to understand for a typical NC, LOC in a CD test. So, now we are
ready with the initial point which is 𝑝𝑜 ’ . Now we know from our previous discussion also
𝑝’𝑦𝐴 determines the size of the yield curve. And we have also seen that these are the state
parameters based on which size of the yield curve gets determined, this one and q.
Now 𝑝’𝑦𝐴 determines the size of ellipse, so the major axis is given by p’ y A. Now this particular
state 𝑝’𝑦𝐴 is NC and 𝑝𝑜 ’ is LOC, it is very clear. Now let us do a drained shearing from O that is
the initial point is fixed then it is drained shearing. Now we do not have to explain much, because
now we know how ESP would look like in a CD test which will intersect the elliptical yield curve
at B.
So, this is the condition like the ESP meets the initial yield curve at B, so that B is given. So, it is
slightly difficult, because if you try to map it then the amount of space available is quite less. So,
you can see that from p 0’ it should move towards B, ESP from O that is O intersects the elliptical
yield curve at B where it yields. Now you need to note one important thing from O to B it is elastic
behaviour and hence it has to move along this green line, that is the loading, unloading line.
684
I come to this again, because we are going to have a bigger picture of this in the next slide. So, I
will just tell that the path in v ln p’ is denoted by this small red line, which is difficult to see but it
will be clear in the next slide. Now when you map this to q 𝜖, you can see that up to B it exhibits
a linear behaviour because it yields only at point B, so that is the point. So, O B this is linear
behaviour, same, the volume change also it will not be much because it is in the elastic range.
So, the B point has been mapped in all the figures. Now when it is further sheared from here since
it is NC and LOC, we know that strain hardening happens. So, failure occurs when ESP touches
the critical state line at F that is B F, so this is the further extension of the path. So, it is very easy
that we have already seen before and then it completes, so it fails at F, so from B to F it moves.
Now what happens?
You can see that this is a hardening response after yielding and hence the volume change also will
be more. In fact you are moving towards this particular point and then it comes down, so it is no
longer on the loading, unloading line and hence the volume change is going to be high, so from B
to F there is significant volume change.
Now the same figure we will see in an enlarged version to understand a few more things which is
very important. Now stress strain response is curved up to B it was elastic and beyond B it is
curved. Beyond B, yield surface expands and then the previous yield curve is no longer valid, now
685
this is very important aspect. The moment the stress path crosses the yield curve I mean to say the
initial yield curve, then the previous yield curve which was there that is the initial yield curve now
it is no longer valid, what is the significance of the statement let us see.
So, it is again q p’and v ln p’, I have done this critical state line it is a smaller line you can see, this
is mainly done to accommodate a bigger figure. The same figure that we discussed in the last slide,
we are just going to repeat it. So, it is again 𝑝𝑦 ’, the initial point is 𝑝𝑜 ’, critical state line, and this
is the effective stress path we have already seen point B, B F. Now this is an intermittent yield
curve, this green one, before failure.
You can see that this particular yield curve it is not touching the point of intersection of ESP with
critical state line, this we have discussed in our previous lectures that which yield curve defines
the failure one, that yield curve beyond which there is no further expansion of yield curve. So, that
yield curve which is passing through the point of intersection of effective stress path with the
critical straight line, and that is this particular point F.
So, this intermittent point is given as C that is point of intersection of ESP with the green yield
curve which is an intermittent yield curve and then the final yield curve is given as red one. So,
what it means is that once it crosses point B, this yield curve is now no longer valid it goes with
green and then comes to red where it stops. Now there is no further expansion of yield curve, and
this we have already talked with respect to plastic potential and plastic strain vector.
𝑝’𝑦𝐴 , we are just mapping it into v ln p’ and from here the unloading happens, and then the starting
point is p 0’. And when it is loaded that is from the point O, I can also write here it is point O. So,
from point O, we know that up to here it is elastic response up to B. So, then if it is elastic response
it cannot deviate from any but it has to move along the green line, why? Because we know that
unloading, reloading line is an elastic response, so we can easily take it to be along O B, so that is
what is marked.
And this is the same thing we have seen in the previous slide. Now you imagine a situation where
we do not know the concept of yielding and the yield curve is not marked, and that is what we
686
have done in our previous lectures. We did not consider yield curve into picture, when we
discussed the first few lectures of critical state, you can see that we have not discussed the yield
curve.
Now we have drawn the stress path to failure, we have also marked in v ln p’. If you remember
from this particular point we would have just drawn a curve like this to failure to critical state line.
That is because we have not introduced the concept of yielding, now once this particular yield
curve comes into picture things have very clear. That means up to where exactly it will follow the
unloading, reloading curve and from where it will deviate towards the critical state line.
Unless we have the yield curve in picture we cannot do this, and we have followed a line like this
which goes to critical state. You can refer back and see for yourself, and this information is
complete only when we draw the yield curve along with the stress path, so that is all I just want to
′
add here. Now consider intermittent yield curve with 𝑝𝑦′ = 𝑝𝑦𝐶 , just because this is named as C,
′
the extreme yield point is named as 𝑝𝑦𝐶 , so that is given here.
′
Now let us draw a loading, unloading line from 𝑝𝑦𝐶 , just for our understanding. Now QR this
′
particular line, QR is the unloading line corresponding to 𝑝𝑦𝐶 ,. Now shearing from B to C causes
deviation from reloading line due to shearing and also by following the concept of yield curve.
Now further shearing at point B, now it will deviate from elasticity and it will undergo hardening
behaviour.
Now from B, it will deviate from point B initially it was along the unloading, reloading line which
is the elastic response, but shearing at B it has to deviate. So, at this particular point B only it is
deviating, initially we have drawn it as a curve, now it is coming along the unloading curve and
then it is coming towards the critical straight line. So, this particular point from where it will
deviate it comes from the concept of yield curve.
So, point C let us say that, that particular point is there on the unloading line QR because it is
touching here. Now we are redefining the elastic limit, let us consider that the point is HC, so this
particular point is the point C. Now the moment it deviates from point B or the moment it passes
687
point B, we are defining a new yield curve. Now let us say that the stress path has reached up to
point C, and it is at point C right now.
So, that particular point defines a new yield limit, and that is what is shown here at the point C that
is QR is the new elastic response for this particular soil corresponding to point C. So, when it
deviates it has come to point C, now it is at this particular point. Now let us say that you are
unloading from point C, that is you have sheared, now let us consider that you are unloading from
point C, so what will happen?
It will follow this particular unloading, reloading line because now it will become more like an
elastic behaviour if you are unloading from point C. But if you continue to shear it will fail at F,
that is normal response that we know and it is like this. So, you can see here now from the starting
point it is following a curve like this. And initially without knowing this point we have done like
this, you can see for yourself that how we have drawn the failure line from O to critical state line.
So, on further shearing it fails at F, now from C if it is 1 dimensionally loaded it will proceed along
RQ, now this information is very important. Let us say from C for assumption, I am not telling it
is already under shearing procedure. But then to understand how this path will follow depending
upon the condition of loading that you apply that is what I mean to say. Now if from point C, let
us say that it is one dimensionally loaded, not sheared, if that is the case then it will move along
this line.
Because this is the one dimensional unloading, reloading curve. So, it instead of moving C F, it
will move along this, provided it is one dimensionally loaded. But in this case we are shearing
from C, so it will continue and it will fail at F, shearing can also be done from 𝑝’𝑦𝐴 . Now let us
say that we have followed a lightly over consolidated point at O, but one can always load without
unloading you can load from point 𝑝’𝑦𝐴 itself.
Now the beauty of this point is that it has already yielded, it is already at it is yield limit. So, it is
already at yield point, so elastoplastic shearing occurs from 𝑝’𝑦𝐴 . So, if you start loading from
here it will follow the path and it will move to the critical state line, it will go like this and it will
688
fail on the critical state line because there is no yielding needed. So, whatever we have understood
in our previous lectures it remains same.
But here one additional aspect is the information what the yield curve provides to us in terms of
predicting the failure, so that we have to keep in mind.
Now prediction of normally consolidated LOC for CU test, now initially it was CD test, now it is
CU test, most of the discussions remain same only thing is the volume change is replaced by excess
pore pressure generation. And this is again a prominent point we know, there is no volume change
so 𝛥𝑣 = 0 𝑣𝑜 = 𝑣𝑓 so this is a very important clue for solving the problem.
Later lectures while we try to solve the problem that 𝑣𝑜 = 𝑣𝑓 assumption is important for solving
the problem. So, again it is the same response, the soil is isotropically consolidated 𝑝’𝑦𝐴 soil is
unloaded to 𝑝𝑜′ , that is point O. Now that is mapped on to v ln p’, I do not think we need any further
discussion on this. Soil is sheared and drained from point O, ESP is denoted as OBCF, so till B the
response is elastic again.
3
I will not discuss this because the slope is (1−3𝐴) ; A is equal to 1/ 3. So, it takes the path
upwards till it yields at point B. That is the point, same point O and B is in the same line, so O
689
and B, so here also O and B is a point, same point. Then it fails to critical state line that is at F, so
it is the curved response and this is an intermittent point C, that intermittent point is marked here.
So, from O to F it is a horizontal line, all these things we have already seen before.
Yield stresses are determined from the intersection of ESP with the yield surface. Now you can
see that this ESP meets the yield surface at this point. So, one can always substitute the values in
the elliptical equation and obtain the value of 𝑞𝑦 and 𝑝’𝑦 . Similarly for F also we will get the value
of 𝑝’𝑓 and 𝑞𝑓 . And that mostly you can get the results from the geometry of the figure itself.
Now B, you can see that this is a linear response, here also in the pore water generation also it is a
smaller response. Because it does not have much of because it is an elastic loading, the pore water
pressure generation is also more or less linear. Now F, from once it yields you can see that there
is a non-linearity coming up and there is a sharp increase in the pore water pressure. And that is
why there is a drastic reduction and it moves towards F.
After yielding u increases non-linearly, so you can see here this B F is a kind of non-linear
response. Difference between homologous points between ESP and TSP gives u excess, so if you
want to predict excess pore water pressure that is also possible. Because we know undrained test
what is the total stress path. This is the effective stress path and this is the total stress path it is at
an a slope of 3.
So, any difference between these homologous points, homologous point means the same point on
ESP and TSP, that difference is going to give u excess. So, u access at yield u excess at failure,
everything can be predicted merely looking at the figure and determining the total and effective
stress corresponding to that. For example if you want to determine what is the u excess at this
particular point that is yield point 𝑝’𝑦 .
We need to determine 𝑝’𝑦 , so we know 𝑝’𝑦 here it is equal to 𝑝𝑜 ’. So, this point is known, one can
also determine this particular point. So, this difference will give you u excess at yield, so it is very
convenient to predict what is the kind of pore water pressure that gets generated, so that is what it
means. So, based on the MCCM that is the elliptical yield curve and the effective stress path
690
information jointly will be able to predict the yield point and the failure point. But obviously for
all these we need to know what is the critical state parameters.
So, now let us see prediction of HOC for CD test. Soil is isotropically consolidated 𝑝’𝑦𝐴 and it is
unloaded to 𝑝𝑜 ’, the same procedure. But then now the unloading is a bit more, so that HOC state
gets emerged and OCR is greater than 2, so that we already know. So, again the repetition of the
same, now only thing is here it is volume change and we need to consider both compression and
expansion.
Because we expect dilation to happen in heavily over consolidated soil. 𝑝’𝑦𝐴 , so this is the yield
curve associated with 𝑝’𝑦𝐴 , the extreme isotropic consolidation point. The same is marked on v ln
p’, unloaded; you can see that it is very near to the q axis or v axis. So, OCR will be definitely
greater than 2. Critical state line, so yields at point B, now it is a CD test, it is at an inclination of
3, so it is not meeting and failing here, we have already seen this, it has to yield first.
Now this point B is similar to the peak condition. Now from O to B it is elastic because it has to
yield first. So, here also it has to move along the unloading, reloading line, so this line, green line
it has to move because it is an elastic response. Now in our previous lecture, we have associated
this 2 peak because the maximum point of q that you can get for OC is the peak point. Now here
691
we are limiting it by yield, earlier we have limited this by a peak line; you can refer back and see
in our earlier lectures we have done that.
Now here this peak and the yield point is analogous to each other, we are considering this to be
more or less same. And what it will look like in q epsilon, so up to B it should be a in elastic
response. Now in the case of MCCM we are forcing certain conditions here, when it reaches peak
it is yielding, so that is what it means. So, yield point B is same as what the peak point, so here
this is one imposing and we do not know like whether the soil is going to behave in that manner.
Whether the soil is going to yield at the peak point, this is not clear. But according to the
explanation of critical state model specifically MCCM what we are dealing now what it states is
that the soil will yield at it is peak. So, this is a kind of idealization one has to keep in mind. And
hence up to peak the response is also linear; this is also an additional condition that gets imposed
because we are considering a critical state framework.
So, you need to understand that many things gets idealized which may not be same in the real soil
behaviour. So, from here, now another important aspect, like what is the cumulative volume
change. Now up to here if it is yielding, it is more or less the peak behaviour is the end of their
initial compression. These things we never bothered when we plotted the results in the previous
lectures, we have done all these in our previous lectures right from module 2 onwards. But we
never bothered how we are plotting it. So, here there are certain clear indications which we need
to follow.
Now at this particular maximum point is more or less the end of compression and where it will
start dilating. So, 𝑞𝑦 is the yield point and q y is the yield stress which is same as q peak. So, this
𝑝’𝑦 is the yield mean stress, returns and fails at critical state line with shrinking yield curve,
shrinking yield curve also we have seen because this is a strain softening behaviour, so it comes
and fails at this particular point.
So, from here it is dilating and that is why there is a sudden drop and that is the point F, the failure
point. I am not getting into the details again because we know this we have discussed this, now
692
that is the failure point 𝑝’𝑓 . Now how does it look like in q epsilon plot? You can see that there is
a strain softening, so this strain softening we have discussed before but now at what point it will
start strain softening, that dilation happens this is adding more meaning to it.
So, by defining the yield curve you can see that we know along this it will move that is the elastic
response and then it will move upwards or dilates and fails at F. Now these information becomes
quite handy in the presence of yield curve, that is the yield curve supplements these knowledge
along with our discussion. So, from here at this particular point B strain softening means there is
a kind of dilation that is happening up to F where it fails.
So, yield and failure stresses can be determined from MCCM by substituting at these appropriate
points. So, one can also get the failure and the yield stresses. Now OB is compression and BF is
dilation, now this becomes quite strict in the MCCM model, resulting in strain softening. Now
according to MCCM the actual elastoplastic behaviour starts at peak, now this is defined because
it is explained in that manner.
In reality elastoplastic behaviour may start even before reaching peak which is not captured in
MCCM, now that is what we need to understand. All the time the model may not represent the
actual behaviour but it is a fair good approximation. By saying that the elastoplastic behaviour
more or less sets at the peak. Because that is where the dilation would have started that is the
yielding would have started.
Now exactly the peak may not be correct but it will be around that peak, so that is what I just need
to point out those differences when you try to idealize the shearing behaviour of soil using MCCM.
693
Now for CU test for HOC. So, we have defined all the four combinations that is NC LOC, CU CD,
HOC CU CD. So, now we will discuss CU test. Soil is isotropically consolidated to 𝑝’𝑦𝐴 , again
the same condition let us skip this, it is the same condition. So, u excess, now instead of cumulative
volume change, now we have excess, excess pore water pressure. Same yield curve, everything
remains same, loading, unloading curve, 𝑝𝑜′ the point O, the initial point critical state line.
Now yields at B similar to peak condition, now only difference is the slope of the curve that is
3
, we know that it is vertical. So, OB 𝑝𝑜′ from O to B is the yielding, so here B and 𝑝𝑜 ’ is at
1 − 3𝐴
the same point. Now how does it look like in q 𝜖 plot and u excess plot, it is positive instead of
compression. So, here the positive pore water pressure everything remains same, 𝑞𝑦 is the point
and 𝑝’𝑦 .
So, 𝑝𝑜 ’ = 𝑝𝑦′ according to this particular plot. Returns and fails at critical state line with shrinking
yield curve, this also we have already discussed, so this is the failure point F. So, it fails at F and
𝑞𝑓 𝑎𝑛𝑑 𝑝’𝑓 is the failure stresses, and strain softening behaviour in q 𝜖 plot. And you can see that
it more or less becomes a negative pore water pressure at failure for OC sample.
Because of the tendency to dilate in a drain test is same as negative pore water pressure failure in
OC. Now OB represents positive u, here it is positive pore water pressure and at BF it becomes
negative pore water pressure. And this is very clear from the total stress path with an inclination
694
of 3 which is drawn. Yield point it is on the left side of TSP, so at this particular point it is positive
u, and at failure point here it will be negative u.
Because it is crossing the TSP and now it is on the right side. So, 𝑝’𝑓 is greater than 𝑝𝑓 , now
𝑝𝑓 corresponding to failure line, this is the point F we can see that this is the particular point if you
draw it down that is 𝑝𝑓 . Now you can see that 𝑝’𝑓 is greater than 𝑝𝑓
So, effect of that is all about heavily over consolidated and the prediction of 𝑞𝑓 , 𝑝’𝑓 , 𝑞𝐼 , 𝑝𝑦 will
happen based on the geometry of the figure and by substituting the relevant equations in
combination with the yield curve. So, we will come to that when we solve the problem. So, effect
of ESP on failure. So, we just need to understand 2 important points here, slope of ESP like in
what manner it is loaded.
Slope of ESP means in what manner it is loaded, what is the load combination that we follow.
Now slope of ESP would influence how it will fail, in which manner it is going to fail? So, that is
what we want to demonstrate here. So, it is the initial yield curve 𝑝’𝑦𝐴 , let us say that loading is
done in such a manner that slope is less than M, now what is M? M is the slope of CSL.
Now let us say you consider a point which is slightly over consolidated at O. And the loading is
done such that ESP 1, here it is ESP 1 is with a slope less than M, you can see this, what would
695
happen in this case? ESP 1 is not going to intersect critical state line, so the loading can happen
without failing; probably it is very close to isotropic loading. So, if ESP is progressing with a slope
less than M, it is not going to touch the critical state line, in that case it is not going to fail.
The loading is not going to bring about failure; this is a very important information which comes
from here. ESP 1 will not produce shear failure in soil. Now let us consider another case again the
starting point is same and ESP 2 which corresponds to NC or LOC soil which is mostly applicable
for excavation. So, this is the way in which the ESP 2 is progressing. There is a loading such that
ESP 2 takes this particular slope, now this is a typical case of excavation.
For ESP 2, yielding has to take place first before failure, we have seen that in all our previous
discussions for NC and LOC it yields first and it fails. But then it was very easy, it never cross the
critical state line, when did it cross the critical state line? First the effective stress path crosses
critical state line yields and then comes back means it is a typical HOC behaviour. But now our
point is LOC, in all our previous discussion LOC and HOC it yielded first and then went ahead
towards the critical state line.
But now if the loading such as that in excavation if it is done in such a manner that it follows ESP
2, then it will cross the critical straight line as you can see here, it crosses it goes and yields and
then will come back to failure. So, that is what just wanted to highlight here, an LOC can give a
behaviour close to HOC during shearing, provided the loading combination is in such a manner
that it produces an effective stress path similar to ESP 2.
So, initial point is NC or LOC, but shearing behaviour is similar to HOC, why? Because it crosses
critical state yield and then gets back to failure.
696
Now that is all related to the prediction and another aspect that we can predict based on critical
state framework or very specifically MCCM is the soil stiffness. So, the next discussion is
determination of soil stiffness. Soil stiffness is characterized by elastic modulus E’, shear modulus
G and bulk modulus K’. K’ and E’ is shown to show that this is relevant only for effective stress
condition when it comes to soil behaviour.
We also have u and 𝐾𝑢 which comes from the undrained test, but for very specifically the
mechanical properties we will just consider E’ and K’, and G does not need any prime because it
is a shear modulus. Now all of these are elastic parameters. Now in the critical state framework
can we use MCCM or critical state framework for predicting the values of these stiffness. So, we
need to confine only to the elastic response of the soil.
Now probably you would get a hint here, like we will be referring only to the unloading, reloading
line, not the ICL. Because ICL exhibits elastoplastic behaviour, we are concerned only about the
elastic stiffness. So, for prediction we need to now account for the unloading, reloading line of
critical state framework. Soil stiffness can be determined using critical state model, triaxial testing
and isotropic consolidation results.
697
We need for determining the critical state parameter, isotropic consolidation and or the triaxial test
results, yes. So, how to do this? Let us see, so unloading-reloading line represent elastic behaviour.
Now I think none of few would be having any doubt related to this, that it is unloading-reloading
line that represents the elastic behaviour. Now when we talk about stiffness, it is always better that
we discuss it in terms of void ratio, because that is what we are use to.
But the same, everything remains same only thing is the axis changes by 1 + e in the case of v. So,
either I discuss in terms of e or in terms of v the responses are same, but the slope would change
it is 𝜆 and CC, CC in e plot, 𝜆 in v plot. So, there is only difference in the slope values, but all the
concept remains same, it is v or e versus ln p’. So, here instead of writing v = v𝜅 - 𝜅ln p’ I am
writing
𝑒 = 𝑒𝑘 − 𝜅 ln 𝑝′
This 𝑒𝜅 is same as corresponding to one unit pressure that is 1 kilo pascal and instead of 𝑣𝜅 , it is
𝑒𝜅 . So, 𝑒𝜅 is similar to 𝑒𝑘 𝑖𝑠 𝑠𝑖𝑚𝑖𝑙𝑎𝑟 𝑡𝑜 𝑣𝑘 ln 𝑣 − ln 𝑝′ . Now if you differentiate this we will get
𝜅𝑑𝑝′
𝑑𝑒 = − . And the elastic volumetric strain increment that is given as
𝑝′
𝑑𝑒
𝑑𝑒𝑝𝑒′ = −
1 + 𝑒𝑜
Please note that here the p’ corresponds to volumetric stress, and e corresponds to elastic.
𝑝
The same thing we have seen 𝑑𝑒𝑝′ means it corresponds to plastic. So, here it is elastic volumetric
strain increment, you can write it as -de that is change in void ratio upon original volume that is 1
+ eo. So, initial volume is 1 + eo, change in volume to it is original volume that is 1 + eo that will
give the elastic volumetric strain increment. So, we know what is de from this expression, if you
substitute this you can get
𝜅 𝑑𝑝′
𝑑𝑒𝑝𝑒′ =
1 + 𝑒𝑜 𝑝′
698
But we also know
𝑑𝑝′
𝑑𝜖𝑝𝑒′ = 𝐾′: 𝑏𝑢𝑙𝑘 𝑚𝑜𝑑𝑢𝑙𝑢𝑠
𝐾′
Now substituting this,
𝑑𝑝′ 𝜅 𝑑𝑝′
=
𝐾′ 1 + 𝑒𝑜 𝑝′
One can write the expression for bulk modulus
′
(1 + 𝑒𝑜 )𝑝′
𝐾 =
𝜅
You can see that the bulk modulus can be determined by knowing the value of 𝜅 and 𝑒𝑜 and the
value of p’. Now we already know the relationship,
𝐸 ′ = 2𝐾 ′ (1 − 2µ) µ: 𝑃𝑜𝑖𝑠𝑠𝑜𝑛′ 𝑠 𝑟𝑎𝑡𝑖𝑜
Substituting for K’ we will get the value for E’ that is
3𝑝′ (1 + 𝑒𝑜 )(1 − 2µ)
𝐸′ =
𝜅
And G we know it is
𝐸′
𝐺=
2(1 + µ)
substituting in the value of E’
3𝑝′ (1 + 𝑒𝑜 )(1 − 2µ) 1.5 𝑝′ (1 + 𝑒𝑜 )(1 − 2µ)
𝐺= =
2(1 + µ) 𝜅(1 + µ)
699
So, elastic constants we can see whatever we have discussed here, it is all proportional to p’. You
can see in the expression K’, E’ and for G it is a function of p’. So, elastic constants are proportional
to p’, this indicates the non-linear elastic behaviour. So, it is not that we simply substitute and get
the result; it should be in an incremental manner. The computation of stiffness should be done
incrementally.
For OC samples the elastic stiffness can be determined from isotropic consolidation, because we
know it would follow the loading-unloading line. And why should we have the elastic modulus for
normally consolidated because that is going to exhibit elastoplastic behaviour. When we say elastic
response, we know that it is mostly relevant for OC samples. So, it was noted that estimating elastic
constants from in situ stress states it gives low values.
That means whatever we determine based on the in situ stress state will give low values of elastic
constants. So, according to Randolph, in 1979 he suggested that G should be determined from the
maximum value of K’. Now what is meant by maximum value of K’? Maximum value of K’
means, it should correspond to maximum p’. And what is the maximum p’? That is the yield stress
for that particular yield curve that is the extreme point of the yield curve which is 𝑝𝑦′ .
700
′
So, 𝐾𝑚𝑎𝑥 should be
′
(1 + 𝑒𝑦 )𝑝𝑦′
𝐾𝑚𝑎𝑥 =
𝜅
So, that is according to the recommendation given by Randolph and others. So, this 𝑝’𝑦 it is
′
referring to and the corresponding void ratio is 𝑒𝑦 , and G is taken as 0.5 of 𝐾𝑚𝑎𝑥 according to the
suggestion. So, this is reported in this particular paper which is driven piles in clay-the effects of
installations and subsequent consolidation, published in Geotechnique. So, one can refer to this for
further reading.
A variation of soil stiffness with strain level, we can note that for a nonlinear material like soil
stiffness is a function of strain, it keeps changing. So, at a low strain we know it is a linear response
and then it becomes a nonlinear response. Now increase in shear strain decreases G and E’. K’
decreases with increase in volumetric strain, because each of this is associated with shear response
and volumetric change.
So, there is a net decrease in soil stiffness with increasing strain. The response is divided into three
zones based on the level of shear strain, which is given as how G, K’, E’ varies with shear strain
or volumetric strain or deviatoric strain, so how does it vary? So, it varies something like this, is
an s shaped curve. There are some indications of the values which is 0.001%, 1%. So, the strain
701
which is less than 0.001 is considered as small strain 0.001 to 1% is intermediate strain, and greater
than 1 is large strain.
So, you can see that even a very small percentage of strain itself is considered as a large strain, in
general discussion it is considered like that. At small strain, stiffness is constant and soil behaves
like linear elastic. So, here this is a linear elastic case where the response is stiff and we have very
less volume change, so or whatever is the amount of strain that is happening is quite less and hence
it remains more or less constant.
At intermediate strain, stiffness reduces soil behaves like elastoplastic and there is a significant
reduction that happens. At large strain, stiffness again becomes a constant value as the soil
approaches the critical state. So, as it reaches critical state then you can see that it reaches more or
less the constant value again. For soil in the field shear strain distribution will be non uniform with
higher value close to loading.
So, when some loading happens in the field what we can see that, the strain will be high close to
the loading location. And hence whatever is the stiffness, in that particular zone that drastically
reduces as per the given discussion. So, as the strain level increases, the stiffness starts reducing,
so closer to the loaded points you can see that the stiffness will be less as compared to other part
of the soil. And that is how the failure zones or the failure points starts generating at from a given
zone. So, soil stiffness will be low in the loaded zone and hence the failures are initiated.
702
So, that is all about a prediction based on MCCM. Let us summarize today's lecture prediction of
failure of NC, LOC and HOC for CD and CU test discussed with respect to MCCM. For LOC,
when ESP move beyond the yield surface that is in q p’ plot, it results in deviating from the
unloading line. So, we have shown that relationship with respect to yielding and the yield curve.
So, from unloading reloading line it deviates towards the critical state line.
And that information we get because we have integrated yield curve along with ESP. After yielding
u in CU test increases nonlinearly. The slope of ESP is less than M, such an ESP will not meet
critical state line and it will not cause failure. LOC soil can shear similar to HOC for cases such as
excavation. Determination of soil stiffness is determined based on MCCM. According to MCCM
elastic constants are proportional to p’.
Computation of stiffness should be done incrementally even though there is an equation, it does
not mean that it gives a particular constant value for the range of p’, it will be different. Variation
of soil stiffness with strain level discussed. At small strain and large strain the stiffness approaches
a constant value. At intermediate strain stiffness reduces drastically. So, that is all for this lecture,
in the next lecture we will see prediction of strains from the critical state concept. So, that is all for
today, thank you.
703
Lecture-53
Strain From MCCM
Welcome back all of you, in the last lecture we have completed prediction of stress states based
on modified Cam Clay model. In that lecture we have seen how handy the information of yield
curve is for interpreting, how the stress path moves beyond yielding and what will be it is
implication in the volume change behavior? That is v ln p’ plot; it has added more clarity we have
seen.
We have also discussed about determination of soil stiffness from MCCM. Now stiffness means,
it basically relates to the elastic behaviour. So, we have completed more or less these stress states,
the stiffness. Now there is one more component which is left out which is strain. Now the question
is, can we determine strain by knowing MCCM or by adopting MCCM? Now strain is an integral
part when it comes to the defining the constitutive relationship.
That is a stress strain relationship, and when you consider a plastic model like MCCM we need to
have both the ranges that is elastic as well as the plastic range after yielding, both are important.
Now unfortunately we will not be touching upon the constitutive modelling using MCCM. But we
will try to understand whether we can determine strain by considering MCCM model, and that is
what we will see in today's lecture.
And when I say strain, there are 2 strains because we are considering everything in q p’ plot, so
deviator stress and mean stress. Hence we also have deviatoric strain and volumetric strain. We
have also seen the yield curve, the plastic potential, plastic strain increment vector where it is
defined in terms of 𝜖𝑞 and 𝜖𝑝′ . So, when I say determination of strain, it pertains to 𝜖𝑞 and 𝜖𝑝′ ,
that is deviatoric and volumetric strain.
704
Now this strain has got 2 components, one is the elastic part, the other one is the plastic part. Now
if you are considering elastic part, it is quite easy. We just need to know only a single information
based on which for different stress condition the response we can predict, because it is a linear
response, but the moment it yields it becomes nonlinear. Now we cannot just simply substitute the
value and get the result, rather since it is a nonlinear behaviour it has to be in an incremental form.
So, small, small increments, both for stiffness as well as for strain determination, so that is very
important. So, whatever we are going to study in today's lecture is like incremental strain
determination.
So, let us move on to the subject that is determination of strain from MCCM. So, as I told there
are 2 parts, one is the volumetric strain and the other one is the deviatoric strain. So, the total
change in volumetric strain can be represented as 𝛥𝜖𝑝′
𝑝
𝛥𝜖𝑝′ = 𝛥𝜖𝑝′
𝑒
+ 𝛥𝜖𝑝′
𝑒 𝑝
𝛥𝜖𝑝′ is the elastic part, and 𝛥𝜖𝑝′ is the plastic part.
So, this is the elastic component, and this is the plastic component, summation of 2 gives the total
change in volumetric strain. Soil sample is isotropically consolidated to 𝑝’𝑦 that is our starting
sentence, it is unloaded to 𝑝𝑜′ a typical LOC. So, q p’ critical state line, in the last lecture also when
705
we discussed about stiffness, we told that either void ratio or specific volume either of the one will
do.
So, here also for convenience I am considering void ratio because that is what we are familiar
about, what specific volume we do not use it that frequently, what we use is void ratio. So, since
we are discussing about stiffness in the last lecture and strain in today's lecture I would like to go
ahead with the void ratio but everything other than that remains same. Now this is the yield curve
which is related to 𝑝𝑦′ , so the maximum yield stress is 𝑝𝑦′ and it is unloaded to 𝑝𝑜′ .
Drained shearing, because when we say volumetric change it is drained response. So, the drained
shearing will cause yielding at A, so this is the point A, so this is the point where it yields. Small
increment of stress AB, now what goes beyond is the expansion of yield curve and that is denoted
by C. So, a small increment, we are talking about a small increment of stress, so that we are
determining a small increment of strain as I told in the beginning.
So, it is all in incremental form beyond yielding, so that is the point B. So, let us take it downwards,
so this is 𝑝𝑦′ which is isotropically consolidated, then there is unloading to 𝑝𝑜′ . So, this point is the
unloaded point then the point where it yields is given by A, so up to yield 𝑝𝑜′ to A, we know that
it will be on the reloading line. So, then the expanded yield curve the moment it crosses the point
A, then it becomes expanded yield curve C so that is again plotted as C in e ln p’ plot, we also
have an unloading line at C.
For this particular condition this is not important but you will see why we have drawn this
unloading line. So, this point B ideally can be represented on the unloading line of C. Now you
may question like why it is not there on the unloading line of 𝑝𝑦′ 𝑝𝑜′ . Now you need to see here
there is a plastic hardening that has taken place when the ESP moves from A to B. So, a plastic
hardening has already taken place means the compression has happened.
That compression is given by this, so this part 𝑝𝑦′ to C it has already happened; now it is on the C
unloading line. So, that unloading line is marked here, so that is why B will lie on this particular
unloading line. So, this is represented 𝑝𝑜′ A is from 𝑝𝑜′ A then as it moves from A to B that
706
hardening is shown here, that is AB. So, the void ratio corresponding to A is given as 𝑒𝐴 , void ratio
corresponding to B is given as 𝑒𝐵 .
Now 𝑒𝐴 to 𝑒𝐵 is a plastic response that is plastic volumetric strain, and 𝑝𝑜′ to A is an elastic strain.
So, change in void ratio for this stress increment, that is when the stress changes from A to B is
given as ∆𝑒 = |𝑒𝐴 − 𝑒𝑏 | Now I can ideally we have to write 𝑒𝐴 - 𝑒𝐵 because 𝑒𝐴 is higher than 𝑒𝐵 .
But it hardly matters you can write 𝑒𝐵 - 𝑒𝐴 also provided you are putting modulus, because we are
not interested in the negative sign. So, e modulus of 𝑒𝐴 − 𝑒𝑏 in general is the change in void ratio
when there is an expansion of yield curve from A to B. So, now we know that it is MCCM, hence
we know all the relationships related to critical state. So, we can conveniently write
𝑒𝐴 = 𝑒𝜅 − 𝜅 ln 𝑝𝐴′
so it is on this unloading-reloading line and hence the void ratio at A can be obtained by this
expression by knowing the critical state parameter κ. For D, that is for this point which is the
maximum yield stress corresponding to the starting point or the starting envelope, that can be
obtained as 𝑒𝐷 . It again it is on the same unloading-reloading line, so it is
𝑒𝐷 = 𝑒𝜅 − 𝜅 ln 𝑝𝑦′
because D corresponding to the maximum that is 𝑝𝑦′ . So, this particular point what we are talking
about, that is what is the void ratio at this? This particular point, so that is given by 𝑒𝐷 . Now for
𝑒𝜅 we can substitute, because this lie on the same unloading-reloading line.
𝑒𝐴 = 𝑒𝐷 + 𝜅 ln 𝑝𝑦′ − 𝜅 ln 𝑝𝐴′
𝑝𝑦′
𝑒𝐴 = 𝑒𝐷 + 𝜅 ln
𝑝𝐴′
Similarly you can also get the expression for 𝑒𝐵 .
707
Now same way we have
𝑒𝐵 = 𝑒𝜅 − 𝜅 ln 𝑝𝑎′ ,
Please remember even though I have written e κ here this would change, why? Now the point B is
on another unloading-reloading line and this point we have specified when we discussed about the
critical state parameters. Like 𝑒𝜅 or 𝑣𝜅 keeps changing depending upon the point where it is
unloaded. So, accordingly this 𝑒𝜅 is different from the previous 𝑒𝜅 .
But since this 𝑒𝜅 is not going to come into the equation, I have not used a separate terminology.
So, but then you need to understand we are now discussing about the line CB, so accordingly 𝑒𝐵 =
𝑒𝜅 − 𝜅 ln 𝑝𝑎′ , just like what we have got for 𝑒𝐴 . And this point 𝑒𝑐 can be written as
′
𝑒𝐶 = 𝑒𝜅 − 𝜅 ln 𝑝𝑦𝐶
So, 𝑒𝜅 can be substituted.
′
𝑒𝐵 = 𝑒𝐶 + 𝜅 ln 𝑝𝑦𝐶 − 𝜅 ln 𝑝𝐵′
.So, just like you have written for 𝑒𝐴 , we can write
′
𝑝𝑦𝐶
𝑒𝐵 = 𝑒𝐶 + 𝜅 ln ′
𝑝𝐵
Now for 𝑒𝐴 we have 𝑒𝐷 , and for 𝑒𝐵 we have 𝑒𝐶 , and we also know that this point D and this point
is related, how it is related? Because it is falling on the ICL line or NCL line, accordingly point C
that is 𝑒𝐶 can be written as 𝑁𝑜 . 𝑁𝑜 means it is corresponding to unit pressure
′
𝑒𝐶 = 𝑁𝑜 − 𝜆 ln 𝑝𝑦𝐶
708
𝜆 is the slope of NCL. And 𝑒𝐷 can be written as
𝑒𝐷 = 𝑁𝑜 − 𝜆 ln 𝑝𝑦′
If that is the case you can substitute for 𝑁𝑜 , so
′
𝑒𝐶 = 𝑒𝐷 + 𝜆 ln 𝑝𝑦′ − 𝜆 ln 𝑝𝑦𝐶
′
𝑒𝐶 = 𝑒𝐷 − 𝜆(ln 𝑝𝑦𝐶 − ln 𝑝𝑦′ )
′
𝑝𝑦𝐶
𝑒𝐶 = 𝑒𝐷 − 𝜆 ln
𝑝𝑦′
So, this is the expression for 𝑒𝐶 in terms of 𝑒𝐷 , so, you can substitute for 𝑒𝐶 in 𝑒𝐴 - 𝑒𝐵 .
So, now our ultimate aim is to get the total volumetric strain Δe. So, we have now obtained the
expression for 𝑒𝐴 , we have obtained the expression for 𝑒𝐵 , we have also obtained the expression
for 𝑒𝐶 in terms of 𝑒𝐷 , so one will get eliminated. So, we know that
′ ′
𝑝𝑦𝐶 𝑝𝑦𝐶
𝛥𝑒 = |𝑒𝐴 − 𝑒𝐵 | = 𝑒𝐷 + 𝜅 ln ′ − 𝑘𝑙𝑛 ′
𝑝𝑦 𝑝𝐵
By substituting for equations 𝑒𝐴 𝑎𝑛𝑑 𝑒𝐵 . If that is the case
′ ′
𝑝𝑦′ 𝜆𝑝𝑦𝐶 𝑝𝑦𝐶
𝛥𝑒 = 𝑒𝐷 + 𝜅 ln ′ − 𝑒𝐷 + 𝑙𝑛 ′ − 𝜅𝑙𝑛 ′
𝑝𝐴 𝑝𝑦 𝑝𝐵
Now if we expand we can write
′ ′ ′
𝛥𝑒 = 𝜅𝑙𝑛𝑝𝑦𝐶 − 𝜅 ln 𝑝𝐴′ + 𝜆𝑙𝑛𝑝𝑦𝐶 − 𝜆𝑙𝑛𝑝𝑦′ − 𝜅 ln 𝑝𝑦𝐶 + 𝜅 ln 𝑝𝐵′
Now we can rearrange the terms we get
709
′
𝛥𝑒 = (𝜆 − 𝜅)(ln 𝑝𝑦𝐶 − ln 𝑝𝑦′ ) + 𝜅(ln 𝑝𝐵′ − ln 𝑝𝐴′ )
So, Δe can be written as

′
𝑝𝑦𝐶 𝑝𝐵′
𝛥𝑒 = (𝜆 − 𝜅) ln + 𝜅 ln ′
𝑝𝑌′ 𝑝𝐴
So, what is actually these points, let us see, 𝜆 − 𝜅 into ln, that is the expanded yield curve, we are
trying to find out this incremental strain. So, total volumetric strain which it has undergone when
it moves from when the ESP expands from A to B. So, for that what is the maximum yield stress
′
corresponding to expanded yield curve, that is 𝑝𝑦𝐶 .
′
So, ln 𝑝𝑦𝐶 that is a maximum, divided by the maximum yield stress of the initial yield curve, so
′
𝑝𝑦𝐶 𝑝′
that is the ratio. So, ln ′ into 𝜆 − 𝜅 plus the elastic portion that is κ into ln 𝑝𝐵′ , so that is the
𝑝𝑌 𝐴
expanded point that is B upon A. So, that will give us the total change in volumetric change in that
is represented by void ratio, so Δ e we have obtained.
So, we can very well define the total change in volumetric strain as 𝛥𝜖𝑝′ , that is the total change
in volumetric strain that is Δ. So, Δ is important incremental, please keep this in mind. So, 𝛥𝜖𝑝′ is
equal to we know that it is change in void ratio upon initial void ratio. So, that is initial volume
not initial void ratio initial volume, so change in volume is represented by
𝛥𝑒
𝛥𝜖𝑝′ =
1 + 𝑒𝑜
Now we have obtained the expression for Δe. So, we can write the total change in volumetric
strain
′
1 𝑝𝑦𝐶 𝑝𝐵′
𝛥𝑒𝑝′ = [(𝜆 − 𝜅) ln ′ + 𝜅 ln ′ ]
1 + 𝑒𝑜 𝑝𝑦 𝑝𝐴
710
So, that is about the total change in volumetric strain, we will just highlight. Now what we want?
We are not interested in total volumetric strain rather we want the components both the elastic
volumetric strain and the plastic volumetric strain. So, we will see how to get that? Before that I
will show the expanded view of this particular circle which is denoted as I. So, e ln p’ we are just
focusing on the part I.
So, AB which is this is represented by this line, and there is unloading that is happening. So, if you
consider this 𝑒𝐴 𝑒𝐵 is the void ratio and the points are 𝑒𝐹 , now what is this unloading? Like if you
consider with reference to point A, we can say that this is the void ratio that it got regained, when
the load is released what happens is this much of the strain or this much of the void ratio it regained,
the soil regain, so that is nothing but the elastic component.
So, 𝑝’𝐴 , 𝑝’𝐵 the pressure FE this particular distance, that is FE is the elastic component of 𝛥𝜖𝑝′ and
that is represented by 𝛥𝜖𝑝𝑒′ . Now soil unloaded from B will move along the unloading line with
respect to the maximum yield point 𝑝’𝑦𝐶 and that is what I have already explained. Now you are
unloading it from B means that corresponds to the unloading-reloading line at corresponding to
point C.
There is a maximum yield stress is 𝑝’𝑦𝐶 . So, 𝛥𝜖𝑝𝑒′ can be written as
711
𝛥𝑒 |𝑒𝐸 − 𝑒𝐵 |
𝛥𝜖𝑝𝑒 = =
1 + 𝑒𝑜 1 + 𝑒𝑜
So, this will give the elastic component, because this much is the elastic volume change or
regaining. So, we know that 𝑒𝐵 = 𝑒𝜅 − 𝜅 𝑙𝑛 𝑝’𝐵 , the way we have done it before and 𝑒 𝐸 =
𝑒𝜅 − 𝜅 𝑙𝑛 𝑝’𝐸 . So, if you rearrange this and you substitute for 𝑒𝐸 − 𝑒𝐵 what we are going to get
is this
𝜅 𝑝𝐵′
𝛥𝜖𝑝𝑒′ = ln =
1 + 𝑒𝑜 𝑝𝐸′
That is 𝜖𝑝𝑒′ is the required elastic volumetric strain increment. So, volumetric strain increment the
elastic component is 𝜖𝑝𝑒′ We can also write it
𝜅 𝑝𝐵′
𝛥𝜖𝑝𝑒′ = ln ′
1 + 𝑒𝑜 𝑝𝐴
because 𝑝𝐸′ and 𝑝𝐴′ both are same. So, you will get now this is the reference point when the ESP
moves from A to B what is the elastic component of volumetric strain? That is 𝛥𝜖𝑝𝑒′ = κ is the
critical state parameter that represents the unloading-reloading line divided by 1 + 𝑒𝑜 , 1 + 𝑒𝑜 is
nothing but the initial volume, ln which is the final point 𝑝𝐵′ which is the initial point 𝑝𝐴′ on the
yield curve, I mean to say from yield curve to yield curve. We also know
𝛥𝑝′
𝛥𝜖𝑝𝑒′ =
𝐾′
because this is an elastic parameter, it is a linear response, p’ which is the stress upon the bulk
modulus K’.
712
Now we have set the volumetric elastic strain, we need to also determine volumetric plastic strain.
We have determined the total volumetric strain; we have determined the elastic volumetric strain.
So, obviously the plastic volumetric strain is total minus the elastic component, and that is what is
written here,
𝑝
𝛥𝜖𝑝′ = 𝛥𝜖𝑝′ − 𝛥𝜖𝑝𝑒′
Symbol Δ is important as I told in the beginning incremental stress. So, if you substitute this in the
equation,
′
𝑝 (𝜆 − 𝜅) 𝑝𝑦𝐶 𝜅 𝑝𝐵′ 𝜅 𝑝𝐵′
𝛥𝜖𝑝′ = ln ′ + ln − ln
1 + 𝑒𝑜 𝑝𝑦 1 + 𝑒𝑜 𝑝𝐴′ 1 + 𝑒𝑜 𝑝𝐴′
There is a total volumetric strain minus the elastic component of volumetric strain will give the
plastic strain and that if you expand the terms will get cancelled off and finally we will be left with
′
𝑝 𝜆 − 𝜅 𝑝𝑦𝐶
𝛥𝜖𝑝′ = ln
1 + 𝑒𝑜 𝑝′ 𝑦
So, the plastic component can be represented by the critical state parameters. Now for undrained
condition we know that the total volumetric strain will be 0.
𝑝
𝛥𝜖𝑝′ = 𝛥𝜖𝑝𝑒′ + 𝛥𝜖𝑝′ = 0
So, this component 𝛥𝜖𝑝′ = 0, in that case we get the expression

𝑝
𝛥𝜖𝑝𝑒′ = −𝛥𝜖𝑝′
713
So, we have now completed the volumetric strain part, now we have left with shear strain. So, we
need to determine the shear strain. Now for determining shear strain, we know that about the plastic
potential and how it expands. So, yield curve Y is represented as
2 𝑞2
𝑌 = 𝑝′ − 𝑝′ 𝑝𝑦′ + =0
𝑀2
Because this is MCCM and the yield curve corresponds to ellipse, this is the equation of the ellipse.
We have already seen that it follows associated flow rule and normality rule which is applicable
for MCCM, we have discussed this in detail please refer back if you want to revise it. Now the
resultant plastic strain increment which is represented as 𝛥 𝜖̂𝑝 = 𝛥𝜖𝑝 , we are just taking for sake
of convenience that is the incremental form. For a given stress increment, we know that because
of normality rule it is normal to the plastic potential function.
And we also know because of associated flow rule in MCCM plastic potential function is same as
yield curve. So, we do not have 2 equations, the equation for yield curve itself is good enough for
defining plastic potential function. So, here in q p’ this is what we have already seen, this is the
ESP, this is the yield curve the same as the plastic potential function, this is the direction which is
normal to the curve at this particular point of contact is 𝛥 𝜖𝑃 which is plastic strain increment
vector.
714
We have used this particular symbol initially which is considered same as 𝛥 𝜖𝑃 . Now here you
𝑝
have the component of 𝛥𝜖𝑝′ which we have already determined in the previous slide. We are left
with 𝛥𝜖𝑞𝑝 which we not know which is the plastic component of deviatoric strain, and this we have
already determined. So, our next task is to find out what is 𝛥𝜖𝑞𝑝 .
Now normal to the yield curve can be obtained by differentiating Y. Now basic calculus says that
we can get the normal to the yield curve by differentiating the equation of yield curve. So, we get
tangent and from tangent we can get the normal, so
2𝑞
𝑑𝑌 = 2𝑝′ 𝑑𝑝′ − 𝑝𝑦′ 𝑑𝑝′ + 𝑑𝑞 = 0
𝑀2
So, if you rearrange it the slope
𝑝𝑦′
𝑆𝑙𝑜𝑝𝑒
𝑑𝑞
= 2 − 𝑝′
𝑑𝑝′ 𝑞
𝑀2
Now normal to the yield curve that is if you multiply the slope of normal and tangent it is going to
be -1 and that is what we are written here.
1 𝑑𝑝′
𝑁𝑜𝑟𝑚𝑎𝑙 𝑡𝑜 𝑡ℎ𝑒 𝑦𝑖𝑒𝑙𝑑 𝑐𝑢𝑟𝑣𝑒 = − =−
𝑑𝑞 𝑑𝑞
𝑑𝑝′
715
1
So, the reciprocal of minus of 𝑑𝑞 is the normal to the yield curve and that is represented as
𝑑𝑝′
𝑑𝑝′
− . So, normal in terms of plastic strain we know from this particular figure it is Normal in
𝑑𝑞
𝑝
𝑑𝜖𝑞
terms of plastic strain = 𝑑𝜖𝑝
𝑝′
So, this slopes can be equated, so we can always write

𝑑𝜖𝑞𝑝 𝑑𝑝′
=−
𝑑𝜖𝑝𝑝′ 𝑑𝑞
𝑑𝑝′
Now − can be written as
𝑑𝑞
𝑞
𝑑𝜖𝑞𝑝 𝑀2
=−
𝑝𝑦′ 𝑑𝜖𝑝𝑝′
′
2 −𝑝
Which comes from the previous slide. We have already obtained the equation for dq upon dp’.
So, we are taking the minus sign inside we have

𝑞
𝑑𝜖𝑞𝑝 2
𝑀2
𝑝 =
𝑑𝜖𝑝′ 𝑝𝑦′
𝑀2 (2 − 𝑝′ )
Which can be further simplified as

𝑑𝜖𝑞𝑝 2𝑞
=−
𝑑𝜖𝑝𝑝′ 𝑀2 (2𝑝′ − 𝑝𝑦′ )
𝑞
𝑑𝜖𝑞𝑝 𝑀2
=−
𝑑𝜖𝑝𝑝′ 𝑝𝑦′
𝑀2 (𝑝′ − 2 )
q/p’ is nothing but stress ratio 𝜂 which we have seen in our previous discussions. And we have
𝑀2
also derived the equation for 𝑝’/𝑝’𝑦 which is equal , please refer to MCCM discussion for
𝑀2 +𝜂 2
this. So, these 2 relationships are known which can be substituted in this. So, if you substitute
𝑑𝜖𝑞𝑝 2𝜂
=
𝑑𝜖𝑝𝑝′ 𝑀2 (2 −
(𝑀2 + 𝜂2
)
𝑀2
So, you finally get this expression
716
𝑝
𝑑𝜖𝑞 2𝜂
=
𝑑𝜖𝑝𝑝′ 𝑀2 − 𝜂2
So knowing the critical state parameter and the stress ratio we can find out 𝑑𝜖𝑞𝑝 .
𝛥𝑞
Now the elastic part of the deviatoric strain, we already know that is 𝛥 𝜖𝑞𝑒 = 3𝐺 . Now equations
for strains are only valid for small changes in stress, I mean to say incremental form. So, the
moment you are talking about yielded point from first yielding to the next small increment of stress
that is represented by another yield curve. So, between the initial yield curve and the reference
yield curve which we are talking about there is a small stress increment and we are talking about
that small strain increments.
So, all the equations that we have discussed till now is referring to that small incremental strain.
Not possible to determine failure strains by substituting failure stress directly. So, if you want to
determine what is the failure strain, we should not simply substitute the failure stress and get the
result. We need to get the small increments; you keep on adding it and then get the final strain. So,
CSM is an elastoplastic model, so it is not elastic model, so you cannot simply substitute point
stress and get the strain.
So, strains should be calculated for small increment of stresses from yielding to failure. So, you
will have different components of strains from the first yielding to the failure. So, the total
717
components of strain increments are added to obtain the total plastic strain. So, these small,
components up to failure is added together to get the total plastic strain.
So, drained compression, if you want to determine the stress for MCCM that is corresponding to
drained compression. So, what we have done is, we have determined all the strain components,
knowing the MCCM model and the critical state parameters, now what we are referring to is. We
are just summing up to say how do you determine the stresses which we have already discussed
before. But in prediction when we discussed about this we did not specifically mention.
So, when I say determination of stresses in MCCM what are the stresses which we are concerned
about? We want yield stresses that is when the ESP meets the yield curve that is the initial yield
curve, so that yield point is important. So, 𝑞𝑦 𝑝’𝑦 corresponding to the first yielding, then we are
interested in the failure stresses, that is 𝑞𝑓 𝑎𝑛𝑑 𝑝’𝑓 . Along with that the strain and stiffness comes
into picture but what we are more interested in this particular lecture or this particular course is
the stress components.
So, we have yield as well as failure stress components in terms of q and p’, so how to determine
this? And we also need to find out whether it is for drained compression or for undrained
compression? So, that is what we will just quickly glance through how to obtain the stress
components in this particular slide. Now another aspect is the pore water pressure, the excess
718
powder pressure corresponding to yielding and the excess pore water pressure corresponding to
failure, how do we do that?
We know if it is undrained test, we know that there is a total stress path, so we know that it is at 3.
Draw that; get the difference between total stress and the effective stress. So, determining pore
water pressure is also a straightforward task. So, for drained compression, let us see 𝑝𝑜′ that is the
initial state of the soil 𝑝’𝐶 is the maximum isotropic compression point and the critical state
parameters are known.
So, this is the yield curve which corresponds to the maximum yield stress 𝑝’𝐶 , 𝑝𝑜′ is the initial point
at which it has been unloaded let us say it is an LOC point, this is the ESP. Now this is the point
where it yields which is given as 𝑝’𝑦 and q I. So, let us give an example how to determine 𝑞𝑦 and
𝑝’𝑦 ? We know from this figure ESP meets the ellipse at this particular point that is 𝑞𝑦 and 𝑝’𝑦 , we
know the equation of ellipse.
So, yield stresses 𝑝’𝑦 , q I can be determined by substituting this in the ellipse equation because it
intersects, ESP intersects the yield curve. So, now this is the equation that is
2 𝑞𝑦2
𝑝𝑦′ − 𝑝𝑦′ 𝑝𝑐′ + =0
𝑀2
From this geometry we can also write
𝑞𝑦 = 3(𝑝𝑦′ − 𝑝𝑜′ )
Substituting for 𝑞𝑦 in this particular equation, we will get
2
9(𝑝𝑦′ − 𝑝𝑜′ )
𝑝𝑦′ − 𝑝𝑌′ 𝑝𝐶′ + =0
𝑀2
Now this can be simplified into a quadratic equation in terms of 𝑝’2𝑦 . So, solving the quadratic
equation we will get the value of 𝑝𝑦′ and once we get 𝑝𝑦′ we can get 𝑞𝑦 , so knowing 𝑝𝑦′ 𝑞𝑦 can be
determined. Similarly 𝑝𝑓′ and 𝑞𝑓 can be determined from the intersection of ESP with CSL, this
particular point.
719
Now this is nothing but the geometry, so from the geometry of the figure we can determine 𝑞𝑓 and
𝑝𝑓′ . Rather we have discussed how to determine 𝑞𝑓 and 𝑝𝑓′ in the previous lecture, the same
geometry. Similar procedure can be adopted for undrained compression test. So, let us try to
summarize what we have done essentially in this lecture is to demonstrate how we can determine
the strains from the MCCM model.
The total change in volumetric strain is expressed in terms of critical state parameters. Volumetric
elastic and plastic strain is determined. Plastic strain increment vector and volumetric plastic strain
can be used to determine plastic shear strain, so this also we have discussed. Equations for strain
are only valid for small changes in stress specifically when it yields and when there is expansion
of yield curve.
Strains should be calculated for small increment of stresses beyond yielding. These small
increment of strains are added together to determine the total plastic strain. So, now we have
completed most of the aspects of the critical state module. Now we are left with the final aspect of
critical state where we are going to sum up all these information and understand what is known as
the boundary surface?
720
The state boundary surface where it limits all the possible soil states? So, the next lecture will be
on state boundary surface and with that we intend to finish this particular module 4. So, that is all
for now, thank you.
721
Lecture-54
State Boundary Surface
Welcome back all of you, with last lecture we have already completed what are the basic
requirements for critical state soil mechanics what I have actually planned for in this course. We
have discussed about various 2 dimensional representations. The initial point where the soil starts,
it is shearing, elastic behaviour, yielding and finally the failure. Now where are we heading towards
that will be covered in today's lecture and that is defining the state boundary surface.
In fact we have already discussed this, we have already stated what are the bounds on a two
dimensional representation. So, in today's lecture we will cover state boundary surface specifically.
And we would represent the surface means actually the surface means it is three dimensional, we
have been working with curves in a two dimensional representation. So, today we will integrate
the two dimensional aspects into three dimensional and define what is known as the complete state
boundary surface for the soil.
So, let us start today's lecture which is on state boundary surface. So, CSSM that is critical state
soil mechanics, provides the framework for defining the boundary for the possible soil state. That
722
means when the soil is loaded to be very specifically sheared from it is starting point, what is the
ultimate state up to which the soil would exist? So, this can be defined by critical state soil
mechanics, we have seen that and rather we can predict what is going to happen for a given initial
state of the soil and by knowing the critical state parameters.
So, we worked with two dimensional representation where we have qp’ plot and vp’ plot. And we
have already told ICL, NCL becomes the right most boundary and CSL is the point where it fails.
So, CSL in qp’ is given. ICL, NCL is the right most boundary, this we kept on telling, so right
most boundary is already defined. But still there is something which we need to understand in v
lnp’it is very easy to say that, that is the right most boundary.
But what is the bound when the soil gets sheared; I mean to say in qp’ boundary. So, now since
you have clearly understood qp’ and v ln p’, there is one additional point which I would like to
stress here. If you see vp’ representation, we are not talking about anything related to shearing;
shearing is totally dependent on qp’ plot, why? Because when if the soil is sheared we have the q
that is a deviatoric stress that coming into picture.
And vp’ or v lnp’ plot is exclusively when we talk about ICL or normal consolidation or one
dimensional consolidation where the shear stress has not yet developed. So, basically it is the initial
state or maybe when the soil is getting consolidated, so that is what we define in the bottom most
part. So, in that perspective ICL, NCL is the right most boundary is correct in terms of vp’ but
what happens when the shear stress comes into picture.
Now we also know peak envelope provide the top most boundary, that we have seen and peak is
very specifically useful for over consolidated or dense state of the soil. So, we have already defined
what is wet state and dry state and we know that it was defined with respect to critical over
consolidation ratio line; rather you can also define it in terms of critical state line as the difference
is not much.
In most of the textbooks you will find this wet and dry is defined with respect to critical state line
as well. There is nothing wrong in that, it is with respect to which line you are defining wet and
723
dry state. And probably one should understand that in the case of a normally consolidated sample,
the initial state will have water content more than the critical state line. And as it gets compressed
in the process of loading, it loses it is water content and it joins the critical state line, the stress
path joins the critical state line where it will have water content less than it is initial state, so that
is why it is called wet of critical state more specifically. And same is the case of dry state which is
essentially for over consolidated sample, where the water content is less than the critical state. So,
as the over consolidated sample is sheared, it undergoes dilation, so it has got a tendency to draw
water into it, that we have already explained.
So, the final state will have the water content more than that of the initial state. So, that is why
with respect to critical state line over consolidated point is on the dry side. So, whether you define
the dry state or wet state it can be defined with respect to the critical state line or the final point or
the failure point. So, peak envelope provide the top most boundary which we have already seen
and this is very specifically for the over consolidated state of the sample.
The peak envelope can be defined by Mohr-Coulomb envelope or power law. So, please refer to
our earlier lecture on peak state. So, we have defined in both ways, so this particular state line what
we have shown here comes from the Mohr-Coulomb failure model. But we have also stated that
we can represent the peak state using power law. So, that it passes through the origin and the
importance of it passing through the origin was also very well explained in that particular lecture,
please refer back.
And we also stated that there is a limit up to which the peak state can go, so it cannot just go and
join the q axis. The reason is because of the tension developed in the soil and we know that soil
cannot tolerate the tension that gets developed. So, peak definition also comes from the yield curve,
so peak definition it comes from the Mohr-Coulomb envelope which is a straight line it can come
from the power law curve or it can also come from the yield curve which is the yield curve
corresponding to failure.
So, why it can because we know that the effective stress path, it has to go and yield first before
failing. So, in that perspective in critical state models such as Cam Clay and modified Cam Clay
724
it defines it is boundary based on yield curve. So, yield curve also happens to be defining the peak
state and in the previous lecture we have also quoted this particular point. Peak state is a well
defined boundary for soil state on the dry side of critical state line, we have already discussed this.
Now the question is, is there a well defined boundary for soil state on the wet side of critical state
line? That is normally consolidated and lightly over consolidated state. Now initial state wise we
know that this is the right most boundary. But what about in qp’ plot? We are talking about such
a line that whether this forms a boundary towards the right side when the soil is sheared, also this
represents the normally consolidated line.
Now LOC will be less than this, so NCL forms the right most boundary. Now the question is
whether such a curve or such a surface do exist as the right most boundary? So, this is the question
and this is what we will be answering in today's lecture. Now we also know that the left most
boundary is defined by tension cut off zone, all these aspects we have already discussed.
So, we are just summing it up. Now we have clearly understood the two dimensional
representation, till last lecture we have consistently followed two dimensional representation of
critical state soil mechanics. So, for the first time we will integrate whatever we have learned in
two dimensional plane into a three dimensional surface. So, two 2D plots which we have learned
which is qp’ and vp’ is integrated to a three dimensional plot.
725
So, this is qp’ we have seen, and vp’ we have seen rather we worked with v lnp’. So, for
representation it is fine, we use vp’ as well. Now that we are now representing as a three
dimensional plot where q is shown qp’, so this plane corresponds to qp’. And vp’ that is which is
on the floor, this is floor or on the ground this is vp’ that corresponds to this.
So, this three dimensional plot is in fact broken into 2 different two dimensional plot. Ideally we
should have started with a three dimensional representation and then breaking it up into two 2D
plot. But if I explain the three dimensional plot in the beginning it is very difficult for you to
understand. So, till now we have followed a two dimensional approach, now these 2 two
dimensional approaches is now joined together where p’ axis is common.
Now if I represent critical state line in space, it will look like this. Now please remember this
critical state line which is represented it is in three dimensional space. It is earlier we have
discussed on two dimensional space. But now this is in imagine that this critical state line is in the
space, so somehow somewhere in the space like this. Now if you take the projection of critical
state line in space on vp’ we will get critical state line 1 which is lying on vp’ plot.
So, what we have marked here this particular line is critical state line 1. So, we never quoted it to
be 1 or 2, we discussed it as critical state line. So, now for your understanding what is this critical
state line 1 is the projection of critical state line in space on vp’ plane. Similarly you can have the
projection of critical state line on qp’ plot. So, that is critical state line 2 which lies on qp’ plot. So,
this qp’, the critical state line is what you can see here.
So, you are taking the projection of critical state line in space and that will give you CSL 2. Now
let us say that there are points A, B, C on critical state line which is located in the space. You can
get the same points A, B, C on CSL 1 and CSL 2. So, points A, B, C on CSL can be mapped onto
CSL 1 and CSL 2 rather that is what we were doing. We never mapped but we got the points on
critical state line 1 and 2.
726
So, if I drop perpendicular you can get A1, B1, and C1 which are the projections of A, B, C on
CSL 1. Similarly A2, B2 and C2 are the points on CSL 2 which is the projection of points A, B,
C. Similarly we can get the points on the critical state line in space which corresponds to the points
that we discussed in 2D plots when we discussed in our earlier lectures. So, all those points wherein
the soil goes and fail and what we have discussed in 2D plots, in fact it lies on critical state line in
space.
So, from wherever it is starting it goes and fails on the critical state line in the space. Projection of
this on CSL 1 and CSL 2 is what we have discussed till our last lecture. So, when we understand
this on the three dimensional representation we need to understand further how the effective stress
path looks like in the three dimensional space. So, for facilitating the understanding better we
define 2 different planes.
One is known as drained plane and the other one is known as undrained plane, now what does
these planes correspond to? These are the planes which accommodates the effective stress path
corresponding to a drained shearing or an undrained shearing. So, that is what we are going to see
now.
The first one is drained plane. Now let us consider initial state which is on normally consolidated
line with 𝑝𝑜′ as the initial point which is point A and the initial specific volume is 𝑣𝑜 . So, let us
727
consider the three dimensional representation 𝑞𝑝′ 𝑣. Now this NCL is on vp’ please remember
when we are talking about isotropic consolidation line or normally consolidation line, we are
talking about vp’ plane.
So, now the initial point is on the ground, let us say that if we consider vp’ as the floor or the
ground and q, the vertical direction upwards. Then what I mean to say is that CSL 1 and NCL or
ICL is on the ground. So, NCL, ICL is on vp’ plane which is on the ground with q = 0 isotropic or
one dimensional. Please remember, there is no shearing which has happened, we have compressed
the sample, we have consolidated the sample it is still it lies on ICL or NCL.
And when we say that there is no q, q = 0 for the initial conditions that means q that the point is
on the ground itself or the point is on ICL or NCL. So, let this point be A, it has the initial point of
v 0, that is the initial specific volume as 𝑣𝑜 and the initial mean stress of 𝑝𝑜′ . So, that is the starting
point, A is the starting point now and now when it is sheared q starts developing.
So, without knowing much we can say that the effective stress path will start moving in the vertical
direction, upwards. If it is for compression it will move downwards if it is for extension, so that is
what we are going to see now. So, now we are talking about a drained shearing, so we are talking
about drained plane, so we are referring to drained shearing. So, for drained shearing ESP will
originate from A that means it will originate from this point A.
And it will fail on the critical state line, this particular critical state line in space at B. So, let us see
how it happens? So, when shearing starts q comes into picture, that means the effective stress path
has to move upwards for drained compression. So, this is how it looks like, now please remember
this particular effective stress path AB is in space. So, it starts from the point on the ground and it
goes and meets the critical state line in the space which is in upwards.
Drained ESP AB will lie on a plane APQBS, what it means is that? This particular effective stress
path can be accommodated on a plane which is represented by APQBS, how does it look like? So,
it looks like this APQBS, so this is the plane which accommodates this effective stress path AB
728
you can very clearly see that this particular plane is with an inclination, without thinking much
you should be in a position to now answer what will be the slope of this particular drained plane.
Because that is what we were working with for the last few lectures, we did not understand this in
terms of plane. But definitely we have understood this in terms of the slope of drained effective
stress path. And we know that the effective stress path is at an inclination of 3 in qp’ plot, so that
is exactly what we are doing now also. So, APQBS that plane is inclined at slope 3 with vp’ plane,
so with vp’ plane is the bottom plane.
So, this plane APQBS is inclined at 3 with the vp’ plane. So, line p q, this line p q, this one is the
effective stress path in qp’ plane. So, please remember that is what we were working with, that
was discussed in 2D plot. So, every time we were referring to p q as the effective stress path in qp’
plane and that is with an inclination of 3, remember. So, in qp’ plot with CSL 2, so this line is p q
with an inclination of 3, this is what we have discussed.
So, it is one at the same, so it is just the projection of one of the side of the drained plane on qp’
plot. So, the plane APQBS is called drained plane, so that is what we need to understand. So,
drained plane is what? Drained plane is the plane which accommodates the effective stress path
corresponding to drained shearing or to be very specifically compression or extension. So, there
will be series of such drained planes possible, depending upon various initial state.
Now p 0 dash becomes p 1 dash, so you will have another drain plane because the effective stress
path then changes. Similarly a series of drained planes are possible in the case of drained shearing.
So, let us say that what it means when I say that there is a series of effective stress path or the drain
planes which are possible, so this is the plane what I have already explained with an inclination of
3.
Now parallel to this plane you will have another plane with the same inclination of 3 but with a
different initial point. Similarly you can see that how it progressed, so this is with maybe another
initial state. So, whatever B the pattern is same, inclination is 3, so this is what is meant by drained
plane. So, now it is very easy for you to understand what is an undrained plane, because if the
729
drain plane is the plane which accommodates the effective stress path corresponding to drained
ESP.
Then we know that undrained plane has to be the plane which accommodates the undrained ESP
for a given state. So, that is what we will see next.
So, we will see undrained plane so same initial state 𝑝𝑜′ and 𝑣𝑜 , let that point be point C. So, NCL,
the starting point is on the NCL, so point C with specific volume 𝑣𝑜 and 𝑝𝑜′ is the initial mean
stress. For undrained shearing ESP will originate from C and will fail on CSL in space at D, that
we know. Now how it will fail? So, it will fail like this, we know that for an undrained test, the
initial specific volume and the final specific volume is same.
So, what does that mean? The final point which lies on the CSL which is point D should have the
same specific volume as that of 𝑣𝑜 . So, that means 𝑣𝑓 = 𝑣𝑜 , now what does that mean? It gives
another hint that it should lay on a plane which is parallel to qp’ or which is perpendicular to q v
plane. So, then only this condition is satisfied, so whatever is the kind of effective stress path that
we have in plane the same projection we will get on the qp’ plane as well.
And this is the effective stress path which we discussed earlier in all our previous lectures. So,
undrained ESP CD will lie on a plane CTUDV. So, it looks like this CTUDV, so this is the plane
730
and this particular effective stress path CD is on the plane CTUDV. Now how does this plane looks
like? So, 𝑣𝑓 = 𝑣𝑜 so we know that CTUDV has to be parallel to qp’ plane, that means this plane
is parallel to qp’ plane then only the condition = 𝑣𝑓 will be satisfied.
Also for convenience we can say that CTUDV is perpendicular to v q plane as well. So, CTUDV
is perpendicular to q v plane, this plane CTUDV is called undrained plane similar to what we
discussed for drained. So, there will be a series of such undrained planes possible. This is the first
plane what we have discussed corresponding to 𝑣𝑜 Now for 𝑣1 , there will be another plane which
is again parallel to qp’ and another plane which is parallel to qp’ depending upon different initial
conditions. And all these planes will be parallel to qp’ and perpendicular to q v, so this how it
looks like.
So, now we have discussed about drained and undrained planes and how the effective stress path
lies on it. Now the question still remains, that can we define a right most boundary in terms of a
normal consolidated curve which is there on the qp’ representation. So, let us try to first find out
because this is relevant for NC soil, so drained and undrained SPs for NC soil. Let us say different
initial state A, B, C which is on the wet state.
So, we are discussing whether we are trying to find out whether there will be a unique boundary
on the right side. So, we have 3 different initial states A, B and C which is on the wet side. So, A,
731
B is isotropically consolidated, this A and B we know that it is on the p’ axis, so it will be
isotropically consolidated. A is sheared under drained condition with p’ constant, we can do the
way we want, we can regulate it.
So, A represents shearing under drained condition, no A is not representing, A is the initial point,
from A the soil is sheared under drained condition holding p’ to be constant. So, it is very easy to
represent, so this is the way in which the effective stress path will be. B is shared under undrained
condition; we know that for undrained shearing it will move towards left, that is how it moves the
effective stress path.
𝑞
And C is initially and isotropically consolidated at 𝜂 = 𝑝′ , remember, so we have already stated
this, this is a 𝐾𝑜 consolidation not isotropic consolidation. So, 𝐾𝑜 consolidation will lie above the
𝑞
p’ axis. So, that means there is a ratio 𝑝′ at which it happens, on the isotropic consolidation line q
𝑞
is 0, so 𝜂 = 0. Here the initial consolidation is under an isotropic condition where 𝜂 = 𝑝′.
𝑞
So, at this point C = η. Further the soil sample is sheared at the same anisotropic η. So, now the
𝑝′
shearing is also done in such a manner that 𝜂 is maintained or q by p’ ratio is maintained, which
means the effective stress path should move at an inclination of η, so that is what it means. So, this
𝑞
is the line where 𝜂 = 𝑝′ is maintained and the shearing is happening.
So, q is also increasing, p’ is also increasing and q by p’ is always kept constant at η, so that is
how the effective stress path moves. So, all ESPs passes through the same point P, now whether
this point P is unique or not, so this is the question. We have different effective stress path moving
in space depending upon the starting points A, B and C. The question is these effective stress path
crisscrosses each other at point P.
If so, whether the specific volume at that point P is same for all the effective stress paths. Now if
it is so, then it means that there is a unique surface, that means you are making something like a
net. There is some thread going in this direction, there is a thread going in this direction and another
732
thread going in this direction corresponding to A, B, C. Now the point is whether the effective
stress path going in this direction and going in this direction whether it is touching or whether it is
above.
So, when it is crisscrossing at a point P, if it is touching, this means that this has got a unique point
or unique specific volume. So, that is what we need to understand, so that is why this whole
discussion is. So, all the ESPs passes through the point P, so this is η = 0 on the vp’ plane. So, you
have the initial points A and B on NCL or ICL, so A is mapped and A and B is on the ICL line
because η = 0, why η = 0? It is isotropic consolidation.
C will lie on η that is q by p’ = η, now P is this particular point, now the effective stress path moves
through P. So, from A it moves downwards and it will fail at the critical state line, so for B it
moves in the left ward direction through the point P. And we know that this particular line it is still
moving through the point P, this η. So, that is how it is whether it follows this particular condition
or not, so that is what we need to understand.
So, here you can see one interesting fact that if you are loading with such an η, you can say that
the soil is not going to fail, it is very much similar to the ICL or NCL. But still the shear stress is
there in the soil, but that shear stress is not going to fail the soil. If we are able to and this particular
aspect we have already discussed with respect to the slope M. If the ESP is at a slope less than M,
it is not going to fail, so this η is such a similar condition.
To understand whether all the ESPs will have the same v at point P. Now this is very well explained
before in terms of normalization. So, when we normalize it, it all boils down to the same point, so
we have already explained this, in fact there is a unique point or a unique specific volume that is
possible. So, I am putting it again this particular question whether different drained and undrained
ESPs will have same specific volume where it crosses each other.
That is 𝑃1 , 𝑃2 , 𝑃3 in three dimensional space, what I mean to say is this. These are the series of
undrained tests and these are the series of drained test. Let us say that the point where it crosses
each other is 𝑃1 , 𝑃2 , 𝑃3 so this 𝑃1 , 𝑃2 , 𝑎𝑛𝑑 𝑃3 . Now the point is whether the specific volume at
733
𝑃1 , 𝑃2 , 𝑎𝑛𝑑 𝑃3 will be same for undrained and drained stress path? If yes, then there is a unique
surface, if not, then it is not, so it is very difficult for us to understand and tell this it has to be
based on experimental evidence.
And lot of studies have been done in the past by very prominent scientist and they have come up
with a clear understanding on this, and that is what we are going to discuss now. If the point is
unique, that means wherever it is crossing each other if the specific volume v is same then this
would result in a unique surface in vqp’ space. Otherwise every thread is lying up and down, then
there is no unique surface, every thread is lying at the same if it is crisscrossing like this at the
same point for corresponding to the given state, then only a unique surface is possible.
So, that is where we will discuss the state boundary surface for normally consolidated soil which
is the wet side of critical state line, and that particular surface is called Roscoe surface. So, you
probably understand that the answer is yes, there is a unique surface which is possible in vqp’
space. In fact this answer comes from the normalization which we have already studied in our
previous lecture.
We have shown normally consolidated line as a point on the p’ axis, we have shown critical state
line as a point, so that is how we have normalized the peak line. So, please refer back, so if it
normalize and it exist at a particular point the critical state line exist at a particular point then it in
734
fact shows the uniqueness. So, experimental data which is published and which has been
investigated in detail, suggested that the point of crossing of drained and undrained ESPs have
same specific volume vpq’ space.
Which means that 𝑃1 , 𝑃2 , 𝑃3 in fact have the same specific volume. So, drained and undrained
ESPs form a unique surface between normally consolidated line in v:p’ and critical state line in
v:q:p’ space. So, this is the starting point, so all the starting point of ICL, NCL lies on the ground
that is on a v:p’ plane. And the final failure will be on this critical state line in the space, so what
it means is that?
All the drained and undrained ESPs from different starting points, it forms a unique surface
between normally consolidated line and critical state lines. So, between NCL and CSL there is a
unique surface. So, this is how it looks in the space, like you have these undrained stress paths and
then followed by drained stress paths. So, you can see that if you have multiple such initial points
it will look like a net, so it is a well net surface.
So, 𝑃1 , 𝑃2 , 𝑃3 is shown here the same two dimensional representation is shown in the space. This
forms a unique surface is the state boundary surface for normally consolidated soil in v:q:p’ space.
So, the SBS for normally consolidated soil is state boundary surface for normally consolidated soil
is known as Roscoe surface. So, this particular unique surface is what is represented by Roscoe
surface and that Roscoe surface is the state boundary surface for normally consolidated soil.
So, Roscoe surface forms one of the boundary surfaces beyond which soil state is impossible. So,
there is no soil state which is possible beyond this particular surface.
735
So, similarly we will also discuss state boundary surface for OC. We know that it is a peak line,
and this surface in three dimensional plane is called Hvorslev surface and this is for OC soil which
corresponds to dry side of critical state line, and the surface is known as Hvorslev surface. A series
of drained and undrained ESPs of OC soil meets peak or yield curve and then fails at critical state
line, we know this.
This is exactly what I am telling and we have already discussed this. So, CU fails so U yields and
then fails similarly the consolidated drained ESP goes and yields and then fails at this particular
point. Peak can also be defined by power curve, so this peak line can also be defined by power
curve. Peak line forms the state boundary surface for over consolidated state, this state boundary
surface for OC is called Hvorslev surface.
𝑞
Now Hvorslev surface meets Roscoe surface in the normalized plot, so this if you normalize by 𝑝′
𝐶
𝑝’ 𝑞 𝑝’
′ it will meet at this particular point, so that is how it looks like 𝑝′ ′ . We have already seen this,
𝑝𝐶 𝐶 𝑝𝐶
only thing is we have learned Roscoe surface as normally consolidated state. So, this corresponds
to normally consolidated point, this point corresponds to critical state line, so this is Hvorslev
surface, this is Roscoe surface projection on the normalized plot. Hvorslev surface is represented
by
𝑞𝑝 𝑝’𝑝
= 𝐺𝑝𝑣 + 𝐻𝑝 ,
𝑝𝑐′ 𝑝𝐶 ′
736
this comes from the peak line that particular lecture. Please refer back this is the equation for
Hvorslev line. Hvorslev surface in 3D representation, so how does it look like? So, this is CSL,
this is the line which gets terminated, so this is the tension cut off zone, we know that the peak line
terminates at this point, so this line is the termination of peak line.
So, this is the peak line corresponding to various initial states, so this, this, this is the same line
which is shown here. So, the peak line in space forms a surface and this is the tension cut off which
is shown here in three dimensional representation. So, this top surface, so this surface is what is
known as Hvorslev surface, and this surface is the tension cutoff zone.
So, we have defined the right most boundary, the top boundary and left side boundary, and bottom
is the ground, what is that ground? Ground is vp’ plane, so all the sites it is bounded. So, soil state
starts from the ground and it goes and meets at the state boundary surface. State boundary surface
is basically defined by the critical state parameters what we have already learned, and that is how
the three dimensional representation takes place.
So, now we will define complete state boundary surface, we have discussed everything in
piecewise. Like we have started off with 2D, we have defined the bounds, now we have represented
3D for Roscoe and Hvorslev surface. Now we will put all these things together and that will give
737
us the complete state boundary surface. A complete state boundary surface CSBS is a boundary
that defines all possible states of a remolded soil.
So, that is how v it will look like, so this is the Roscoe surface, this is the Hvorslev surface, so this
one is Roscoe this is Hvorslev and finally this is the tension cut off. So, this forms the complete
state boundary and whatever soil state is there on vp’ plane that is the ground finally it comes and
joins at any of these points on the state boundary surface. State of soil cannot exist outside critical
this complete state boundary surface.
If soil is unloaded from the surface, the state will move back and SBS can also act like yield
surface. So, this is one sort of representation we also know that at yield point if it is not failed, it
can trace back. So, it can go further down, so it can go down the state boundary surface, so state
boundary surface can also become a yield surface. For example in the case of Cam Clay model or
modified Cam Clay model we have a kind of yield surface.
So, if it is ellipse like this then it becomes the modified Cam Clay. So, this type of plane also forms
the state boundary surface because it touches at the critical state line in the space. So, then this is
the kind of surface, so this surface is also complete state boundary surface but when it is considered
in terms of modified Cam Clay. So, logarithmic spiral or ellipse are the state boundary surface in
Cam Clay model and modified Cam Clay model respectively.
So, it can be an yield curve that is developed into a yield surface or it can come from this kind of
representation as well. So, whatever be the complete state boundary surface is defined in critical
state framework.
738
So, let us summarize what we have learned today. The state boundary surface SBS defines all
possible state of soil. Soil state is not possible beyond the state boundary surface. ICL, NCL is the
rightmost boundary without any doubt. Peak envelope or yield curve provide the top most
boundary. Tension cut off zone is the left most boundary, drained plane encompass drained ESP
with a slope of 3 with v:p’ plane.
Undrained plane accommodates ESP and it is parallel to q:p’ plane rather it is perpendicular to v
q plane. SBS for NC soil is Roscoe surface, SBS for over consolidated soil is Hvorslev surface.
Integrating Roscoe surface Hvorslev surface and tension cut off zone gives the complete state
boundary surface for soil. Logarithmic spiral and ellipse are the state boundary surface in Cam
Clay model and modified Cam Clay model respectively.
So, all of you, so now we have completed all the relevant lectures which are needed for module 4.
And probably all the lectures which are needed for this particular course advanced soil mechanics.
Now we are left with lecture on using the problems or rather working out the problems with respect
to the critical state concepts which we have learned. Then we will summarize module 4, and further
will be the closure of advanced soil mechanics.
So, we will work out some problems in the next lecture, you may please go through all the lectures
what we have done for module 4 till now once again. So, that you will understand the problem and
739
it is solution in a better manner. So, that we will see in the next lecture, so that is all for now, thank
you.
740
Lecture-55
CSSM Problems
Welcome back, all of you. So, with the last lecture we have completed more or less all the required
concepts for critical state soil mechanics which I planned for this particular course. That does not
mean that we have completed everything related to critical state soil mechanics, there are a lot
more to learn further. But what is required to start with and for this particular course as advanced
soil mechanics we have finished.
And we have touched upon the stress, the plastic behaviour as well as the strain. But the very next
stage is what is actually needed for practicing it in for solving the actual problems in geomechanics,
so that part we are not getting on to in this particular course. Now based on the concepts that we
have learned, we will work out some problems. So, that whatever concepts you have gone through
it will help you to understand better.
So, today we will be solving some problems relate which will use the concepts that we have
discussed in the past few lectures. So, let us move on to the first problem.
741
The problem is soil sample was isotropically consolidated to 400 kilopascal and the initial specific
volume is 2.052. So, this initial specific volume corresponds to this particular isotropic
consolidation. The soil was subjected to drained compression, determine failure stresses, final
specific volume and volumetric strain at failure? Critical state parameters 𝑎𝑟𝑒 𝑁𝑜 = 3.25, 𝜆 =
0.2, 𝛾 = 3.16, 𝑎𝑛𝑑 𝑀 0.94, so what has been given?
There is a soil sample whose critical state parameters are known, then we have been given with
the initial state, that it has been isotropically consolidated. Now knowing the initial state and the
critical state parameters we will demonstrate or that is what we started off with, why do we need
a critical state framework? For predicting the behaviour of soil now what do you mean by actually
prediction? We have not underlined and specifically stated what actually it is.
Even though we have shown the figures we have shown from initial state how the stress path will
move, how it will yield and how it will fail, we have discussed this. But then this problem is a very
good demonstration of you can use critical state concept for predicting the behaviour of soil or
where it will yield, where it will fail depending upon the known initial condition. And obviously
the initial condition will be known, and based on the known critical state parameters for that
particular soil.
That is what the question is all about. So, now it is given, it is isotropically consolidated, that
means it is having the maximum yield stress. So, for example if I draw q p’, what I mean is that?
This is the yield curve, this is 400, so it is already at it is yield point, it is not unloaded, unloaded
means we are creating lightly over consolidated to heavily over consolidated state. Here in this
case it is sheared under drain condition soon after isotropic consolidation.
So, that means it is a normally consolidated point, so please keep this in mind. So, 400 kilopascal
here it refers to NC normally consolidated state. So, isotropically consolidated but it is normally
consolidated. Now from here if there is a shearing which is done under it is given drained
compression. So, drained compression we know that it will move like this, so it is at a inclination
of 3, this line is actually marked here, this is the one.
742
So, here I have not shown the yield curve rather but it will look like this, so that is we please keep
in mind when you read the question you should be able to visualize or rather you should be drawing
sketching the required details. And please understand about the aspects of there are 3 aspects one
is whether you have to draw it in q p’ v ln p’, it is better you draw it on both because that will give
you additional understanding.
Second thing is the effective stress path, how it will be whether it is drained or undrained? Placing
of the yield curve, now for all practical purpose let us consider modified Cam Clay and hence you
can draw an ellipse. So, half ellipse you can draw, so that will give the entire picture of that
particular problem. So, 𝑝’𝑓 can be directly determined knowing
3𝑝𝑂′
𝑝𝑓′ =
3−𝑀
refer back to the lectures for knowing this. This expression has come from the geometry of the
figure. So, if you substitute
3 ∗ 400
𝑝𝑓′ =
3 − 0.94
𝑝𝑓′ = 583 𝑘𝑃𝑎
Now once you know 𝑝’𝑓 , we can obtain q f, because we know the relationship
So, substitute for 𝑝’𝑓 we will get
𝑞𝑓 = 0.94 ∗ 583
𝑞𝑓 = 548 𝑘𝑃𝑎
So what has been asked, determine the failure stresses? So, one part is over, failure stresses can be
obtained based on the critical state framework. Then we also have the relationship
𝑣𝑓 = 𝛤 − 𝜆 ln 𝑝𝑓′
So, substituting we have

𝑣𝑓 = 3.16 − 0.2 ln 583
𝑣𝑓 = 3.16 − 0.2 ln 583
𝑣𝑓 = 1.89
743
So you can see that it is a drained compression and hence the specific volume will reduce, so it has
come to 1.89. So, the final specific volume has been determined. Now what is the volumetric strain
at failure? At failure the volumetric strain 𝜖𝑣𝑓
𝛥𝑣
𝜖𝑣𝑓 = − ∗ 100
𝑣𝑜
Change in void ratio, so it change in specific volume upon initial specific volume, initial specific
volume is given 2.052, 𝛥𝑣 can be determined because the final one is known so 2.052 is the initial
minus final that will give the change in specific volume.
1.89 − 2.052
𝜖𝑣𝑓 = − + 100
2.052
𝜖𝑣𝑓 = 7.9%
So, this how we can use a very simple example of how we can use the critical state framework for
predicting the final conditions. And here the conditions can be failure condition or may be final
specific volume. Now here it is the case of drained compression, so we know that it undergoes
volume change and initial specific volume will be different from the final specific volume ok.
So, let us move to the next example. Soil sample A was isotropically consolidated to 400 kilopascal
and the initial specific volume is 2.052. So, sample A is that from the previous discussion we know
744
that this is a normally consolidated state. Soil sample B was isotropically consolidated to 863
kilopascal, and then it is allowed to swell to 40 kilopascal where the specific volume is again 2.052.
Please note here, sample B has been subjected to a very high stress 863, it is allowed to swell to a
very low stress say 40 kilopascal. So, without looking anywhere you can find, you can see that this
particular soil sample B will be heavily over consolidated, there is no doubt about it. Determine
failure stresses, final pore water pressure for both samples for undrained shearing? So, now earlier
example was drained shearing, here we have undrained shearing along with that we have both the
samples NC and OC.
Now very important aspect which you have to keep in mind is that the initial specific volume of
both the samples are same, sample A and sample B which has 2.052. Critical state parameters are
same for both the samples 𝜆 = 0.2, 𝛾 = 3.16, 𝑎𝑛𝑑 𝑀 = 0.94. Now the very first condition
which what we know is, that it is undrained. So, the initial specific volume will be same as final
specific volume, so 𝑣0 = 𝑣𝑓 which is equal to 2.052 for both samples A and B.
So, this both samples 2.052 initial condition means it gives us a lot of clue about how this test
would take place and that is the beauty of the idealized critical state, we will see how? Both
samples A and B would fail at the same point on critical state line, how? If you have understood
the critical state concept clearly this sentence it is not very difficult to understand. But for that it is
always better that we sketch the diagram.
So, it is q p’, v p’, the first sample is 400, so now this is a normally consolidated point, so this is
the starting point, this is the required point. Now the next sample it has been unloaded to 40
kilopascal from 863. So, for us loading to 863 is not very important, it will be somewhere down
and from there it will be unloaded. So, that it reaches the final state of on the unloading curve, so
this unloading curve comes from somewhere down.
So, I am not drawing it here but we need to understand that. So, here along the 40 kilopascal we
know that the initial specific volume is 2.052. So, whatever is the specific volume for NC the same
should be the specific volume for HOC as well, so somewhere here, so 2.052, so it will be
745
somewhere in this line. Now we know that this is an undrained test, so there is no change in the
specific volume, so initial equal answer.
So, the path will move in this direction and fail on the critical state line. Now same horizontal line
will be the HOC response also, you can see that this is the starting point of HOC. So, this is HOC
and this is NCL, now this is in the same horizontal line, so it will fail at this particular point. And
that is what is meant by both samples A and B would fail at the same point on critical state line,
why it is so?
It is so, because the starting specific volume is same for both the samples which is equal to 2.052.
Here the failure stress will also be same, so this is the failure stress 𝑝’𝑓 , 𝑝’𝑓 is the failure stress
that also remains same, if 𝑝’𝑓 is same then 𝑞𝑓 is same for both the samples. And it is all possible
because we know that the initial specific volume is same 𝑞𝑓 , 𝑝’𝑓 it is all same for both.
Now here this is the kind of effective stress path for sample A, and what will be the sample B? See
if we do not know the concept of yielding what we will do is, we will draw something like this,
this is also possible or it will be drawn like this. But now we know that it has to yield before it has
to fail. So, with that understanding and we know that there is an undrained shearing, so A will be
equal to 1/3 till it yields.
So, it is all possible and it is in this direction because we know the concept. So, that is the point of
yielding, I am not showing the yield curve here separately. So, yielding has happened but it comes
and fails at the same point. Now we are asked to find out the final pore water pressure as well. And
for that any stress condition on the TSP minus the homologous condition on ESP will give us the
excess pore water pressure even at failure or even at whatever the yield point of wherever we need
we will get that.
For example in this particular case, let us say this is ESP and this is TSP. Now if I draw a horizontal
line this point and this point they are called homologous points. So, the difference between this
will give the u excess, this is the pore water pressure. So, that is what we need to find out and
746
hence we need to have the total stress path, and it is very clear total stress path will be at an
inclination a slope of 3, that also we know.
So, at both the initial points at 400 as well as 40 we need to find out the final pore water pressure.
Now here this difference, this is the failure point, so from here are this difference will give 𝑢𝑓 of
A, that is the final pore water pressure of the sample A. I cannot mark for B because it is very
clumsy, but it is already discussed in the lecture. It is the point between this and this, if you draw
a horizontal line, so it will be this total minus effective.
So, again the same like what we have for 𝑢𝑓𝐴 . So, if that is the case we have the expression
𝑣𝑓 = 𝛤 − 𝜆 ln 𝑝𝑓′
And we also know 𝑣𝑓 = 𝑣𝑜 , unless we know these conditions it will be difficult to solve this
problem. So, substituting the initial specific volume for 𝑣𝑓 , we have
2.052 = 3.16 − 0.2 ln 𝑝𝑓′
𝑝𝑓′ = 255 𝑘𝑃𝑎
𝑞𝑓 = 0.94 + 255
𝑞𝑓 = 239 𝑘𝑃𝑎
So, we have determined the failure stresses. Now we are left with final pore water pressure at
failure. So,
𝑢𝑓𝐴 = 𝑝𝑓𝐴 − 𝑝𝑓′
so we need to determine what is 𝑝𝑓𝐴 ? Now 𝑝𝑓𝐴 is this, that is a point, so this is the point which we
need to find out, so what is 𝑝𝑓𝐴 ?
𝑞𝑓 239
𝑝𝑓𝐴 = 400 + = 400 +
3 3
𝑝𝑓𝐴 = 480𝑘𝑃𝑎
𝑢𝑓𝐴 = 480 − 255 = 225 𝑘𝑃𝑎
𝑞𝑓 239
𝑝𝑓𝐵 = 40 + = 40 + = 120 𝑘𝑃𝑎
3 3
So, once we know 𝑝𝑓𝐴 we know 𝑝𝑓′ , so 𝑢𝑓𝐴 can be determined,
747
𝑢𝑓𝐴 = 𝑝𝑓𝐵 − 𝑝𝑓′ = 120 − 255
𝑢𝑓𝐵 = −135 𝑘𝑃𝑎
And this is very correct because it is an HOC, so the final pore water pressure will be negative, so
that is exactly what we are getting and that is equal to -135 kilopascal. So, we have in the previous
problem we predicted final condition of the specific volume, volumetric strain and the failure
stresses. So, in this problem since it is an undrained test, we have determined the failure stresses
as well as the final excess pore water pressure.
So, we should keep in mind using critical state concept, we can even determine the pore water
pressure at failure, at yield or at any point intermittent point from the starting to the failure. So,
these are all possible, provided we follow critical state framework, we know critical state
parameters and we know what is the initial state of the sample.
So, let us move on, now most of the things have been covered, now what is not covered? We can
see we have not determined the yield conditions. So, now we will demonstrate a problem wherein
we try to determine the yielding. Soil sample was isotropically consolidated under a cell pressure
of 300 kilopascal and it is unloaded to a mean effective stress of 200 kilopascal. So, now from the
question itself we can make out it is not a normally consolidated, it can be LOC or HOC.
748
Drained shearing was performed, so volume change is going to take place. Determine principal
stresses at yield? So, we need to understand what is yield, at what point it will yield and we need
to find out what are the principal stresses, it is not q and p’, it is principal stresses at yield. Now
for finding out principal stresses at yield, we need to determine what is the q and p’ value at yield,
deviator and mean stresses at failure, that is 𝑞𝑓 and p’.
The conditions are initial void ratio is 1.1, now initial void ratio of 1.1 means it corresponds to the
unloaded portion, it is not the isotropically consolidate, it is unloaded now. So, 1.1 corresponds to
unloaded portion, critical friction angle is given 30 degrees, slope of NCL in e-ln p’ is given as 𝜆
which is 0.3. So, here we need to understand that 𝜆 𝑜𝑟 𝜅 is the slope, that we know, that
corresponds to either NCL or the unloading-reloading line.
Now we have discussed most of the things in terms of specific volume and p’. But we can the same
thing can be plotted in e versus p’, only thing is there will be the scale Difference of 1 + e. So, here
please note in the question it is given λ and κ corresponds to the slope in e ln p’ plot. So, you can
go ahead with y axis taken as e instead of specific volume v, so I would just like to show you this
as well.
So, you need to see the question properly and λ and κ do not go by that blind understanding that,
it is always related with specific volume v, no. Whatever is the slope if it is defined in the question
λ corresponds to this, please go ahead with that. The slope of the unloading line is 0.05, so all the
details are given. So, M we know that it is
′
6sin𝜙𝑐𝑠
𝑀= ′
3 − sin 𝜙𝑐𝑠
Now here the critical friction angle is already given and that is equal to 30 degrees, if you substitute
we will get M the value to be 1.2,
6sin30
𝑀= = 1.2
3 − 𝑠𝑖𝑛30
so that is known, 𝑝’𝐶 , what is 𝑝’𝐶 ? We need to understand here, the significance of 𝑝’𝐶 , 𝑝’𝐶 is the
maximum isotropic consolidation. So, that is the maximum yield stress or pre-consolidation
749
pressure, we need to understand, that is the extreme point of the major axis of the ellipse, it is
unloaded to 200 kilopascal.
𝑝𝑐′ = 300 𝑘𝑃𝑎
𝑝𝑜′ = 200 𝑘𝑃𝑎
𝑝𝑐′ 300
𝑂𝐶𝑅 = = = 1.5 (𝐿𝑂𝐶)
𝑝𝑜′ 200
OCR is less than 2 we can consider this to be a lightly over consolidated sample, so it is an LOC,
so this is what it means. So, this is the yield curve, ellipse yield curve for modified Cam Clay. So,
this is the extreme yield stress or the pre-consolidation pressure 300 unloaded to 200, then it is
subjected to drained shearing. So, this the effective stress path fails at point F, so this is the point
of interest to us.
′
Because we need to determine principal stresses at yield. So, this 𝑞𝑦 and 𝑝𝑜𝑦 we need to determine.
So, same thing is shown here, this corresponds to 300, here is the unloaded point 200 and this is
the point where the unloading line crosses the critical state. And according to the definition of
MCCM and the elliptical yield curve, we know that this is the midpoint. So, now we do not have
the gamma parameter given, so we can compute gamma parameter rather in this case it is 𝑒𝛤 .
𝑝𝑦′
𝑒𝛤 = 𝑒𝑜 + (𝜆 − 𝜅) ln + 𝜅 ln 𝑝𝑜′
2
So, 𝑒𝛤 is given, the expression we have already derived, please look back you will find in the
lecture that we have derived this particular equation. 𝑝𝑦′ is the yield stress of pre-consolidation
pressure, λ is given, κ is given, 𝑒𝑜 is given, the initial void ratio is 1.1.
300
𝑒𝛤 = 1.1 + (0.3 − 0.05) ln + 0.05 ln 200
2
𝑒𝛤 = 2.62
750
So, now we can determine the yield condition. We know that the effective stress path meets the
′
yield curve at this particular point. So, we can always substitute q y and 𝑝𝑜𝑦 in the elliptical yield
curve equation. So,
𝑞2
𝑝′2 − 𝑝′ 𝑝𝑦′ + =0
𝑀2
′ ′
𝑞𝑦2
𝑝𝑜𝑦 − 𝑝𝑜𝑦 ∗ 300 + =0
1.22
So, 𝑞𝑦 from the geometry we know that it is
′
𝑞𝑦 = (𝑝𝑜𝑦 − 200)
′ 2
′ 2 ′
(3𝑝𝑜𝑦 − 600)
𝑝𝑜𝑦 − 𝑝𝑜𝑦 ∗ 300 + =0
1.22
2 2
′ ′
′ 2 ′
(9𝑝𝑜𝑦 − 3600𝑝𝑜𝑦 + 360000)
𝑝𝑜𝑦 − 𝑝𝑜𝑦 ∗ 300 + =0
1.44
Again expanding we will be left with a quadratic equation. So, here we will have 2 roots
′2 ′
7.25𝑝𝑜𝑦 − 2800𝑝𝑜𝑦 + 250000 = 0
′
2800 ± √28002 − 4 ∗ 7.25 ∗ 250000
𝑝𝑜𝑦 =
2 ∗ 7.25
′
One of the root for the 𝑝𝑜𝑦 is 246.1. And if you substitute for 𝑞𝑦 will be 138.2
′
𝑝𝑜𝑦 = 246.1
𝑞𝑦 = 138.2
751
′
The another root of 𝑝𝑜𝑦 is 140.1 and 𝑞𝑦 is -179.6
′
𝑝𝑜𝑦 = 140.1
𝑞𝑦 = −179.6
Now 𝑞𝑦 = -179.6 is not possible rather it is not useful for us because it is a compression test.
We know that it is a compression test and hence there is no possibility of 𝑞𝑦 , it achieving an

negative value. So, even though the root is there that is not the useful term that we have to use. So,
′
we will go with the final value of 𝑝𝑜𝑦 and 𝑞𝑦 , 246.1, 138.2 is the required yield stresses.
′
Now we can easily find out the principal stress at failure, why? Because 𝑝𝑜𝑦 and 𝑞𝑦 is known,
substitute the expression. So, principal stresses at yield condition, we have determined, this will
give
′ ′
𝜎1𝑦 + 2𝜎3𝑦
= 246.1 𝑘𝑃𝑎
3
′ ′
𝜎1𝑦 − 𝜎3𝑦 = 138.2 kPa
Solving these 2, we will get
′
𝜎3𝑦 = 200 kPa
′
𝜎1𝑦 = 338.2 kPa
3𝑝𝑜′
𝑝𝑓′ =
3−𝑀
752
So, this is the required solution for principal stresses.
Now the failure condition, we are asked to find out the 𝑞𝑓 and 𝑝’𝑓 which we already know. So, it
is the same solution
3𝑝𝑜′
𝑝𝑓′ =
3−𝑀
200
𝑝𝑓′ = 3 ∗
3 − 1.2
𝑝𝑓′ = 333.3𝑘𝑃𝑎
𝑞𝑓 = 1.2 ∗ 333.3
𝑞𝑓 = 400𝑘𝑃𝑎
So, now we have completed more or less the predictions of stresses. Now we also have found out
we can obtain the stiffness from the critical state framework. We can also determine the strain
from the critical state framework. So, we will do a problem which will demonstrate that particular
aspect.
Soil sample was isotropically consolidated to mean effective stress of 225 kilopascal, sample was
unloaded to mean effective stress of 150 kilopascal. The initial void ratio 𝑒𝑜 at 150 kilopascal is
753
1.4, so the initial void ratio is given, determine total volumetric and deviatoric strain for an increase
in deviatoric stress of 12 kilopascal after initial yield under drained condition. So, what we know?
What we know is that it is loaded to a maximum or the pre-consolidation pressure of 225
kilopascal. Now there is a elliptical yield curve corresponding to this and the major principle x
major axis of the ellipse will be 225 kilopascal.
It is unloaded to 150 kilopascal and the initial void ratio is given. Now what is asked, we need to
find out what is the total volumetric strain and deviatoric strain for this given problem, when the
effective stress path moves beyond the initial yield curve. And that how much it is given? It is
given there is an increase in deviatoric stress of 12 kilopascal. So, when the deviatoric stress moves
12 kilopascal from the initial yield curve. I mean to say the effective stress path what will be the
total volumetric and deviatoric strain? So, we need to count from the initial point to the point where
effective stress path goes past beyond the initial yield curve by an increment in deviatoric stress
by 12 kilopascal. So, you will come to know better when we show it on a diagram. Slope of NCL
in e-ln p’ λ is given 0.16, slope of unloading line is 0.05, critical friction angle is 25.5 degrees,
Poisson's ratio is 0.3.
So, now we need to use the stiffness concept, we need to use the strain concept. M we can find out
there is no need for explanation,
′
6𝑠𝑖𝑛𝜙𝑐𝑠 6𝑠𝑖𝑛25.5
𝑀= ′
= =1
3 − 𝑠𝑖𝑛𝜙𝑐𝑠 3 − 𝑠𝑖𝑛25.5
𝑝𝑐′ 225
𝑂𝐶𝑅 = = = 1.5 (𝐿𝑂𝐶)
𝑝𝑜′ 150
Yield condition, so we need to have the yield first, we will come to the figure a bit later, let us first
resolve what is yield of the effective stress path. So, what I mean to say is, this is q p’, we have
225, so once we have 225 we can always draw the yield curve and then it is unloaded to 150. So,
here this is the point which I want to say and when there is a drained shearing. So, we need to
obtain, since it is given total volumetric and deviatoric, so if you want to have volumetric strain it
is understood that it has to be drained. So, here this is the ESP, now this is the point which we are
′
talking about where we need to find out 𝑞𝑦 and this is 𝑝𝑜𝑦 .
754
So, this is what we need to determine and that we have already shown it in the last problem. So,
just substitute it
𝑞2
𝑝′2 − 𝑝′ 𝑝𝑦′ + =0
𝑀2
′2 ′
𝑞2
𝑝𝑜𝑦 − 𝑝𝑜𝑦 ∗ 225 + =0
12
′
𝑞𝑦 = 3(𝑝𝑜𝑦 − 150)
′
𝑝𝑜𝑦 = 180 𝑘𝑃𝑎 𝑞𝑦 = 90𝑘𝑃𝑎
So, to be very precise, this is if I draw it again on a bigger scale, this is 150, this is set to 3, so
′
𝑝𝑜𝑦 = 180 𝑘𝑃𝑎 and 𝑞𝑦 = 90𝑘𝑃𝑎. Now we are asked if it moves past if this effective stress path
moves by this is 12 kilopascal, deviatoric stress is 12 kilopascal. So, this is the point which it
corresponds to when it deviates by this what will be the kind of strain, total strain that it has
undergone. So, this will include the strain due to this elastic portion and the strain due to this
increment, so both we need to find out, so that is what we will solve now.
So, first let us do the elastic strain determination. So, 𝛥𝜖𝑝𝑒′ we have already derived this, please
refer back strain from critical state that particular lecture wherein it is clearly given 𝛥𝜖𝑝𝑒′ . That
means the elastic part of the volumetric strain and for the initial yield we do not have any plastic
strain. Till it reaches the yield point it exhibits completely a volumetric elastic strain.
755
So, that is what we have found out here, that is
′
𝜅 𝑝𝑜𝑦 0.05 180
𝛥𝜖𝑝𝑒′ = ln ′ = ln = 38 ∗ 10−4
1 + 𝑒𝑜 𝑝𝑜 1 + 0.4 150
Which is the elastic volumetric strain when the effective stress path moves from the initial point
to the first yielding. And 𝛥𝜖𝑞𝑒 that is the elastic deviatoric strain is
𝛥𝑞
𝛥𝜖𝑞𝑒 =
3𝐺
Now what is 𝛥𝑞 and for G we need to find out we have the expression for G
3𝑝′ (1 + 𝑒𝑜 )(1 − 2µ)
𝐺=
2𝜅(1 + µ)
µ is the Poisson's ratio, 𝑒𝑜 is known, p’ please remember here since it is a line starting from q = 0
to q = 𝑞𝑦 better to take the average of the stresses corresponding to this. That is p’ will be the
average value, so
′
(𝑝𝑜𝑦 + 𝑝𝑜′ )
𝑝′ =
2
180 + 150
= = 165 𝑘𝑃𝑎
2
So, instead of using 180, use the average stress condition. So, that will give
3 ∗ 165(1 + 1.4)(1 − 2 ∗ 0.3)
𝐺= = 3655 𝑘𝑃𝑎
2 ∗ 0.05(1 + 0.3)
So, then we can find out what is the deviatoric elastic strain till the first yield point, that is
𝛥𝑞 90 − 0
𝛥𝜖𝑞𝑒 = = = 82 ∗ 10−4
3𝐺 3 ∗ 3655
So, both the elastic strain deviatoric and volumetric, till the first yield we have determined. I mean
to say the first yield means 𝑞𝑝 ’, when the effective stress path starts from here. So, this is the first
yield, so by the time it reaches here what is from here for this stretch what is the elastic strain it
has undergone? Both volumetric and deviatoric strain we have determined.
756
Now we need to determine the strain for expanded yield surface or when the effective stress path
crosses the first yield point, then we have both. Both means plastic strain as well as elastic strain,
now this is an elastoplastic response. It has yielded at the first yield point, it is moving by 12
kilopascal of deviatoric stress.
So,
𝛥𝑞 = 12𝑘𝑃𝑎
𝛥𝑞
=3
𝛥𝑝′
we need to find out if 𝛥𝑞 = 12 kilopascal, what will be delta 𝛥𝑝′ ?
𝛥𝑞 12
𝛥𝑝′ = = = 4𝑘𝑃𝑎
3 3
Now this is the final form of the figure, so this is our initial, the green one is our initial yield curve
and that is 𝑝’𝑦 . So, we have already done, now effective stress path moves from 𝑝𝑜′ , this is the first
yield point the conditions we have everything it is clear. Now it expands from A to B because there
is an increment of 𝛥𝑞 = 12.
Now corresponding increment in delta p’ for this point B is known which is 4 kilopascal. And now
we need to find out this particular point, what is the maximum yield stress corresponding to the
expanded yield curve? Remember the soil sample has not failed, it is intermittent, so this is an
intermittent yield circle. The same thing is shown please refer back to the lecture for more
explanation.
757
Now the total volumetric strain which is inclusive of both elastic as well as plastic is given. So,
we have already derived this, we know this how to obtain,
′
𝛥𝜖𝑝′ = [(𝜆 − 𝜅) ln ′ + 𝜅 ln ′ ]
′ ′
Now what is 𝑝𝑦𝐶 ? 𝑝𝑦𝐶 is the extreme maximum yield or pre-consolidation pressure, 𝑝𝑦′ is the
previous pre-consolidation pressure, 𝑝𝐵′ is the new stress state that is after yielding and 𝑝𝐴′ is the
initial yielding.
So, all these characteristics we need to know, we need to find out 𝑝𝐵′ is known, 𝑝𝐴′ is known, 𝑝𝑦′ C
′
is not known. So, we need to find out 𝑝𝑦𝐶 ,
Known are
𝑞𝐴 = 90 𝑘𝑃𝑎; 𝑞𝑏 = 90 + 12 = 102 𝑘𝑃𝑎
𝑝𝐴′ = 225 𝑘𝑃𝑎; 𝑝𝐵′ = 180 + 4 = 184 𝑘𝑃𝑎
So, what is the stress rate of B is now known. Now if we substitute for the bigger yield curve. So,
what we can write is
2 ′
𝑞𝐵2
𝑝𝐵′ − 𝑝𝐵′ 𝑝𝑦𝐶 + =0
𝑀2
2 ′
1022
184 − 184 ∗ 𝑝𝑦𝐶 + =0
1
′
𝑝𝑦𝐶 = 240.5𝑘𝑃𝑎
′ ′
We get 𝑝𝑦𝐶 which is equal to 240.5 kilopascal. So, 𝑝𝑦𝐶 , so all what is needed for 𝛥𝜖𝑝′ is now
determined.
758
So, just we need to substitute these values in this equation, once we substitute it we will get
′
Δϵp′ = [(𝜆 − 𝜅) ln ′ + 𝜅 ln ′ ]
1 240.5 184
Δϵp′ = [(0.16 − 0.05) ln + 0.05 ln ]
1 + 1.4 225 180
Δ𝜖𝑝′ = 35 ∗ 10−4
Please remember here this is the total volumetric strain when the effective stress path moves from
A to B. And this is inclusive of both elastic and plastic strain, only for that incremental portion.
Now we need to find out the q component that is 𝜖𝑞 what is the 𝜖𝑞 when it moves? So, both elastic
as well as plastic we need to understand.
′
𝑝 𝜆 − 𝜅 𝑝𝑦𝐶
Δ𝜖𝑝′ = ln
1 + 𝑒𝑜 𝑝′ 𝑦
So, we just need to substitute it here, all the parameters are known.
𝑝 0.16 − 0.05 240.5
Δ𝜖𝑝′ = ln
1 + 1.4 225
𝑝
Δ𝜖𝑝′ = 31 ∗ 10−4.
759
So, we have determined the total volumetric strain for the incremental portion, knowing that we
have determined what is the plastic volumetric strain for the incremental portion. Now we are left
with deviatoric stress, so
𝑑𝜖𝑞𝑝 2𝑞
=
𝑑𝜖𝑝𝑝′ 𝑀2 − 𝜂2
𝑞
So, 𝜂 = 𝑝𝐵′ because we are considering the point B.
𝐵
𝑞𝐵 102
𝜂= = = 0.554
𝑝𝐵′ 184
M is known, substitute
𝑑𝜖𝑞𝑝 2 ∗ 0.554
= = 1.
𝑑𝜖𝑝𝑝′ 12 − 0.5542
So,
𝑑𝜖𝑞𝑝 = 1.6 ∗ 31 ∗ 10−4
𝑑𝜖𝑞𝑝 = 50 ∗ 10−4
Please remember, we have obtained the plastic part of the deviatoric strain, that is 𝑑𝜖𝑞𝑝 = 50 ∗
10−4. We need to obtain the elastic part of the deviatoric strain for the incremental portion of A to
B. So, assuming G to be constant for the stress increment, that is from A to B, we can write
Δq 12
Δϵ𝑒q = = = 11 ∗ 10−4
3G 3 ∗ 3655
760
So, all what is needed we have computed now. So, total volumetric strain that is from the initial
point to the point B is Δϵp′ = Δϵ𝑒p′ that is from the starting point to the point A to Δϵp′ which is the
total volumetric strain, both elastic as well as plastic it is inclusive for the volumetric strain up to
the point from A to B.
𝑇𝑜𝑡𝑎𝑙 𝑣𝑜𝑙𝑢𝑚𝑒𝑡𝑟𝑖𝑐 𝑠𝑡𝑟𝑎𝑖𝑛 Δϵp′ = Δϵ𝑒p′ + Δϵp′ = (38 + 35) ∗ 10−4 = 73 ∗ 10−4
Now total shear strain Now the plastic portion comes only from for the portion A to B, the
remaining is elastic strain. Please note, here in this particular case we could obtain this 35
directly, but in the case of a deviatoric strain we have obtained the plastic strain and the elastic
strain separately for the portion A to B. So, that is why 82 is the elastic deviatoric strain which
undergoes when it starts from the initial to point A, 11 is the elastic strain which it is subjected to
when it moves from A to B. So, the total elastic strain will be
𝑝
𝑇𝑜𝑡𝑎𝑙 𝑠ℎ𝑒𝑎𝑟 𝑠𝑡𝑟𝑎𝑖𝑛 Δϵq = Δϵp′ + Δϵ𝑝q = ((82 + 11) + 50) ∗ 10−4 = 143 ∗ 10−4
So, that is what, here in this particular case this issue did not come up because we are getting the
total volumetric strain from A to B. So, all the concepts with respect to critical state we have tried
to cover in these problems, I would advise you strongly to go through more such problems for
practice and that will clear all your concepts. So, that is all for this lecture, thank you.
761
Lecture-56
Summary of Module 4
Welcome back, so in the last lecture we have seen some model problems related to critical state
model. And it is usefulness in predicting the final state or the failure state of soil, knowing the
initial state and the critical state parameters. So, with that we have finished what we have targeted
for module 4 in this course. So, in today's lecture we will summarize what we have learned in
module 4.
So, it is a summary of module 4. Critical state model CSM is a simple elastoplastic model with
easily measurable parameters. Now there are several other models as well, but what makes it
different is that it is simple to understand. And the parameters are easily measurable from whatever
routine tests that we know, so that makes it a bit different from the other models. CSM integrates
two 2D variations, the void ratio versus effective stress and shear stress versus effective stress.
Now probably you will understand the uniqueness of these two. Like the one dimensional
consolidation, it shows how the volumetric change happens. And how the yield curve expands or
the reference yield curve depending upon the pre-consolidation pressure or the yield stress, where
762
the shear stress developed is zero, or it is negligible. Then the failure condition is defined by q p’,
once you start shearing the shear stress develop and that progression is given by q p’.
So, integrating these 2 is the very basics of critical state, relationship among shear stress, normal
stress and void ratio at critical state in v or e, q, p’ plot that also we have seen. Equation for normal
consolidation line and swelling-reloading line in critical state model and the relevant parameters
𝜆, 𝜅, 𝑁𝑜 are the soil parameters and it essentially depends on soil type, mostly you can treat it to
be a fundamental property.
Then we discussed about OCR and yield stress ratio, we have brought out the difference between
the 2 and where one is advantageous than the other. Then we have seen the critical state for
normally consolidated LOC and HOC. Then we defined critical over consolidation ratio line
COCRL, there is a bit of difference between this line and the critical state line. And let me again
repeat here in some of the books when wet and dry state of the soil is defined, it is also defined
with respect to critical state line, but COCRL if that is the line which actually separates, normally
consolidated LOC from HOC, but there is a small difference between both. Critical state in q p’
with parameters M and γ defining failure conditions, so the other one was where q = 0 one
dimensional consolidation, from where the λ κ 𝑁𝑂 comes and the failure is essentially defined by
M and γ.
Then we have normalized q p’ plot and we have proven that NCL becomes a point, CSL becomes
a point and that leads to the further state boundary surface concept, where we treat this to be unique.
So, on normalization it all boils down to a particular point, so in that sense we have understood the
uniqueness of normal consolidation line and critical state line.
763
State of soil and state parameters, we have extensively discussed that. 2D analysis of normally
consolidated LOC, HOC in vln p’ and q: p’ plot for CU and CD. So, for a typical triaxial
compression test we have discussed all these cases. Then the peak state which is relevant for the
dry state of the soil that is for HOC has been discussed in detail. And equation for normalized peak
line also we have seen.
In addition to the Mohr Coulomb representation of peak line, we have also seen the power curve
with the concept that it should pass through the origin. Then we have made the comparison
between power low curve and the linear Mohr Coulomb envelope. Comparison between stress
dilatancy and Mohr Coulomb envelope, in the same lecture we have done that. Then a very
important concept of soil yielding and yield curve was discussed.
Now without this the explanation is not complete and that is where this course is different from
the UG concepts what you have learned. Rather the strength, the failure those things we could still
conceive from our undergraduate background, but then adding these points like yield curve and
interpretation of the results with respect to yield curve. And how it fails from further the
elastoplastic state from where it starts, so these things were slightly additional in this particular
course.
764
Then we discussed about shape and size of yield curve, where we have seen different types of
shapes, where logarithmic spiral was one, and the other prominent one was ellipse. Then now there
are different other models which are critical state models and modified versions of even modified
Cam Clay by including various features, they are also existing. But in this particular course we
have restricted ourselves to modified Cam Clay.
But there are several other models as well, the participants are interested, you should read through
those and those are very well explained in the literature. Two critical state models we have dealt
in this course, one is the Cam Clay model and the other one is modified Cam Clay model. The
elliptical and the logarithmic spiral yield curves have been discussed and the equations were
obtained.
Flow rule, plastic potential hardening rule, plastic strain increment vector which is needed for
understanding the plastic response of geo material has been introduced. I would not say that the
discussion is extensive for this, because there is a huge amount of information that is needed further
to appraise better the plastic behavior of geo materials. But we have set the stage for further
understanding, so we have just discussed very briefly about these concepts and how it is relevant
for Cam Clay model.
765
Then we have went on to explain prediction of soil failure from modified Cam Clay model. Then
we have done the determination of soil stiffness from MCCM and further elastic and plastic
volumetric and shear strain from MCCM. So, how by knowing the initial state and the critical state
parameters, we can predict how the soil progresses towards failure using MCCM. This was
demonstrated, this was studied and this was demonstrated in the form of some model problems.
Then we could understand the state boundary surface SBS based on critical state soil mechanics.
The 3D representation of critical state, the drained and undrained plane which becomes relevant
for while defining the state boundary surface. Then we have seen Roscoe surface, which is the
state boundary surface for normally consolidated soil, that is a wet side. And that forms the
boundary on the right, then we have Hvorslev surface which is a state boundary surface for HOC
soil which is on the dry side and this forms the boundary on top.
And the tension cut off zone on the left, and the bottom we have the axis, axis of isotropic line. S
and the same is the mirror image downwards if you talk about extension. So, we have discussed
basically for compression in this particular course. Then we discuss the complete state boundary
surface for soil. Then we have shown if not in terms of tension cut off, peak and normally
consolidated line in the case of Cam Clay model and modified Cam Clay model even the yield
curve acts as the or is the representative state boundary surface.
766
Now let us discuss there are a few things in addition to what we have learned. Critical state model
was developed mostly for remolded soils which we have already discussed. And it was found to
be applicable for natural soils as well. It closely relates in most of the cases but there are cases
where it deviates as well. Now we need to understand critical state model is basically applicable
for isotropic soil.
The yield curve defined in CCM and MCCM may not be applicable for natural soils with
anisotropy. Now here is the point in natural soils we tend to have anisotropy because of the way
in which it is formed. Most of the soils are formed by sedimentation, so sedimentation means
layering, and when there is layering there is a possibility of anisotropy. So, this anisotropy it creates
some issues in interpreting the Cam Clay and the modified Cam Clay, because the yield curve
which has been idealized may not be representative.
So, there will be a clear deviation in the shape of the yield curve, and that is how the modified
Cam Clay model was also modified further to account for these, so this we are not discussing in
this course. Determination of parameters necessitate 1D compression where we say q = 0 and
triaxial shear test where q is greater than 0. So, these are the 2 concepts one with respect to q = 0,
and the other one with respect to q greater than 0.
A time dependent creep behaviour is not considered in this course, rather it is not captured in
critical state. Then one important aspect which I thought I should share with you, you can see that
in some literature it reports peak and pre-failure strain softening for loose saturated sands under
undrained shearing, this is not stated in this course. Now we have always gone by the notion that
the peak is always related to a dense state of the sand or over consolidated soil.
But there are results which are there in the literature wherein the peak is also reported for loose
saturated sand and that too under undrained condition. Why we did not discuss this? is that, for
sense generally the undrained condition is not that prominent, unless otherwise we talk about
liquefaction. So, that is why we have not chosen to explain this, and what important difference is
that this peak is considered to be due to pre-failure strain softening, where the trend which we have
discussed is post failure strain softening.
767
This is different from the post peak strain softening behaviour of dense sand which is stated in this
course. So, that is how, in fact we should not say it is post peak it is like post yielding. If you see
the lecture you will see that this is based on the post peak yielding. So, that after yielding it start
exhibiting, so that is how we have interpreted. But then it is more or less once the yielding starts
we know that it is towards failure.
And there is a very small kind of differences between these terminologies, so this is all for module
4. And now in all sense we have completed the course, we are left with one lecture where we will
basically discuss about the closure of this course, so that is all for now, thank you.
768
Lecture-57
Closure of Advanced Soil Mechanics Course
Welcome back participants. So, this is the closure lecture for advance soil mechanics MOOC. In
the past 12 weeks we had a journey together trying to advance the concept of soil mechanics. I
hope some of you would have got benefited from the lectures which we have seen as part of this
course. But I just want to add whatever I have been telling and what I told right at the beginning
that please treat this course as a beginning for further advanced learning in geo mechanics.
So, in this course basically we have started with introduction to continuum mechanics as the first
module. And this was followed by shear strength, stress path and critical state soil mechanics.
Now at this stage, you may find a disconnect between first module and rest of the course
modules, why I would like to add this point here is while teaching this to students one comment
which I use to receive is that the first module is a standalone and there is a disconnect between
that with the rest of the course.
769
In fact they are correct, I fully agree to it, now where the problem is that how to take this course
further? That will help us to understand where is the actual connect between module 1 and the
rest of the course. So, how to take this learning forward? Now we have stopped here as part of
this course because we need to have some sort of understanding and I think with further reading
you will be able to have your concepts clear.
Now having done that how to take this course forward. Now further learning will involve the
extension plastic behaviour, more into constitutive modelling. Not only Cam Clay, there are
various plastic behaviour and the fundamentals of plastic model or geo mechanical behaviour.
Wherein we also have to learn more about no associated flow rule, definition of plastic potential
and how do we use this for solving a problem?
In this part we have not touched upon. Now if you really want to advance the learning in plastic
behaviour, you will see that most of the discussions would involves stress tensor and invariants
which you have learnt in module 1. So, now you can see it is more like a circle, we started at a
point from module 1 and then we have not completed yet. So, if I would have had module 5
maybe that is going to further turn back and reach module 1.
So, without module 1, we cannot advance further for this particular course. So, for solving
geomechanical problem the concepts of elasto-plastic constitutive modelling need to be
integrated with the concepts learned in module 1. So, the basics that we have learned module 1
that particular basics they are needed for solving or for understanding the constitutive behaviour
in when you talk about the plastic behaviour of the geomaterials.
So, this is how we integrate module 1 with the rest of the courses. But that is further advancing
this particular course. Ideally that is the next module of learning. Now let me also tell you that
this course is not enough for a holistic understanding of the plastic behaviour of geomaterials and
it is implementation in the numerical modelling. Now many of you would have already used
some of the very popular numerical modelling softwares like FLAC, PLAXIS, Abaqus, geo-
slope kind of.
770
So, where there are lots of analyses of the structures which are done the slope stability analysis,
the stress analysis and we also discussed about the failure states. Now how it is implemented in
that particular framework? So, for that we need to complete the other part or we need to know
the further extension of what we have learned now. So, with this present learning it is not enough
but the stage is set for you to learn further, now that is all about this particular course.
Now before winding up I would acknowledge all the participants of this course for their valuable
support. My TA students Bharat Rattan, Charakho, Vineet, Tharun, Biswajyoti and Jishnu
Choudhury who has help me quite a lot. I would also like extend my thanks to centre for
Education Technology, IIT Guwahati. And finally to NPTEL, IIT Madras for a very good
organization and keeping us updated time to time and for helping us in completing the course.
So, that is all, if you find any inconsistency, mistakes in the lectures, please drop an email at
srees@iitg.ac.in, why I am very specific about this is that? So, this is my first MOOC and when I
am recording the videos, I have seen that I am using some words where it is written may be
sigma, maybe I am talking epsilon. So, these kinds of mistakes sometimes it may not be within
my notice.
But as a listener, as a participant when you are going through if you find there is some
inconsistency or mistakes, please feel free to inform me about that, so that I can get this
771
corrected. So, with this note let me wind up this course, I wish you all the best to all the
participants and thank you for attending this course.
772
THIS BOOK IS
NOT FOR SALE
NOR COMMERCIAL USE
(044) 2257 5905/08 nptel.ac.in swayam.gov.in

Siol Mechanics

Uploaded by

Document Information

Original Description:

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Siol Mechanics

Uploaded by

Copyright:

Available Formats

INDEX

The reason is encompassing strength compressibility and flow characteristics, it will be a

3 types of tests were unconsolidated undrained, consolidated undrained and consolidated

(Refer Slide Time: 21:29)

it linearly transforms to vectors that is cauchy's formula.

So we have a Cartesian coordinate x y z. So you can consider positive x plane, now x is an

There are 6 shear components or shear stresses where (i j)

can be written based on Cauchy’s equation as , normal component of traction vector

(Refer Slide Time: 07:22)

(Refer Slide Time: 02:48)

So, how do we get that?

by the normal vector.

what is the magnitude of .

(Refer Slide Time: 06:22)

Such as such a transformation can be obtained by diagonalizing σ. So it is one of the same

the equation takes the form as follows

we get 3 roots of the equation these are called Eigen values.

Now if we substitute these Eigen values in the equation

So we are going to consider transformation of under transformation of coordinate axes.

So S1 = σ 1 –σ m, similarly S3 = σ 3 –σ m. So the expression for J1 is S1 + S2 + S3, I1 is σ 1 + σ 2

So OD makes equal inclination means the inclination is with σ 1, σ 2, σ 3 axis. You

we know the relationship So we know these are same and equal, n1 = n2 =

This distance is given by p’ and distance MP = . So both p’ or σ m or the hydrostatic

hence it can also be written equal to . So both = . So please try to derive

Now σ m = (σ 1 + σ 2 + σ 3)/ 3. So if you substitute it here this will be equal to 0. So that is

The next relationship is to derive

The next relationship is I2 = S1 S2 + S2 S3 + S3 S1+ 3 σ 2

(Refer Slide Time: 12:10)

Now by replacing it by suitable expression. We know

So this is one relationship between the both invariants.

Now substituting J2 expression we get

Simplifying further we get

If you write that |σ - λ I | will give you lambda I, I is an identity matrix. So

So, strain displacement relationship helps to assess the compatibility requirements of a

So u, v, w are the displacement of a point in x, y, z directions and infinitesimal displacement

So cause-effect relationship it is written in matrix form where ε is the strain, E is the

So expression for bulk modulus corresponding to hydrostatic stress condition is just to

So undrained condition, we will be discussing about drained, undrained condition in detail,

So, pre-failure behaviour is defined by elastic parameters. Further, it needs parameters to

For this in this particular model, the deviatoric strain is given by

Why? Because here σ2 = σ3 hence, it will become - 2 μ σ3 / E. So that is denoted as equation 1.

That is if you rearrange this σ3 and - μ σ3 / E we get (1 – μ) σ3 / E - μ σ1 / E. So, based on ε1 and

(Refer Slide Time: 11:30)

we obtained in axisymmetric case, now, what is given .

The next invariant is mean stress

Volumetric strain εv = ε1 + ε2+ ε3 and deviatoric strain

So, to summarize this particular lecture axisymmetric 2D idealization is predominant in soil

If it is cementation we call it as a true cohesion. But it can also be due to certain

Accordingly we know p’ that is the mean stress is

So depending upon whether it is compression or extension, we need to take appropriate

(Refer Slide Time: 09:19)

So here it is t = a’ + s’ tan α , α is the inclination. So it is more or less very identical to Mohr

This causes consolidation which we know consolidation leads to progressive densification

So, H corresponds to volume of water here, 𝑚𝑣 is coefficient of volume compressibility in

nVCw Δ u Now this is already obtained in the previous slide.

So, if we equate these two we get

We can replace the effective stresses by total stresses

(Refer Slide Time: 22:54)

completely saturated case, if it is completely saturated now we have assumed water to be

component is no more 0 it is an appreciable amount.

or with increase in deviator stress that means.

sample is saturated or not when will it be saturated when b value is equal to 1.

stress has significant influence on Δ𝑢.