Consolidation Characteristics of Clay Using Constant Rate of Deformation Test

You might also like

Download as pdf or txt
Download as pdf or txt
You are on page 1of 10

Golden Jubilee Conference of the IGS Bangalore Chapter, Geo-Innovations, 30-31 October 2014

Consolidation Characteristics of Clay Using


Constant Rate of Deformation Test
Amit Prashant1* and Gyan Vikash2
1
*Indian Institute of Technology, Gandhinagar, India. E-mail: ap@iitgn.ac.in
2
Shiv Nadar University, Dadari, India. E-mail: gyan.vikash@snu.edu.in

Abstract: Constant rate of deformation (CRD) test has several advantages over incremental
load (IL) test for finding compressibility characteristics and permeability of clays, including
the fact that CRD test is a much quicker test and provides an efficient option. Awareness
about the potential of CRD test has limited its use in practice despite its major advantages.
The duration of IL tests often makes it difficult for the commercial laboratories to perform all
the tests and pushes them to often use crude empirical relationships. At the same time, it is
imperative to note that the existing CRD theories used to interpret the results from this test
have some issues which pose constraints related to strain rates and sometimes produce
unrealistic interpretations. A review of the conventional CRD theory and the interpretation
scheme is presented in this article besides illustrating their limitations. A new theory has been
also illustrated which eliminates some of the conventionally existing limitations in CRD test
interpretations. This theory is based on the moving boundary concept under constant rate of
deformation which provides consistent solution for loading as well as unloading phase of
these tests.

Keywords: Clays, Consolidation, Compressibility, Constant rate of deformation,


Permeability

Introduction
Consolidation characteristics of clays play important role in finding the settlement of
foundations and simulation of other structures for their stress-deformation response. The
samples collected through geotechnical investigations are generally used in the laboratory to
find the compressibility and permeability of clay by performing IL tests. These tests take
somewhere around 7-12 days to perform one test. A complete project may have around 50 to
100 such samples to characterize, which makes it formidable task sometimes to perform IL
tests considering the time line of the project and the amount of effort required. In such a
situation, a commercial agency is inclined to use empirical equations to predict the
compressibility parameters based on the data from Standard Penetration Tests (SPT) or Cone
penetration test (CPT). Such correlations come with high uncertainty which makes the design
engineer to go for an uneconomical design with conservative settlement estimates.
On the other hand, the CRD test on clays can be performed in 4-5 hours and it also
provides significantly more data for more accurate interpretation of parameters. Hamilton and
Crawford (1959) were the first to introduce this test. Despite the fact that more than 50 years
have passed, there is lack of awareness about this test because of which it has not come to
practice as much as it should have. The testing method and interpretation is simple enough for
any laboratory to adopt. The theories proposed by Smith and Wahls (1969), Wissa et al.
(1971), and Yoshikunis et al. (1995) have been conventionally used to interpret the CRD
consolidation test data. Although there is no Indian code available for this test, ASTM D4186-
06 is available with good description of the test and its interpretation scheme. However, there
are some unresolved issues in this tests which include the issue of interpreting unloading-
reloading phases of test. The works of Gorman et al. (1978), Armour and Drenevich (1986),
and Sheahan and Watters (1997) also point out that the existing theories for interpretation of

1
Golden Jubilee Conference of the IGS Bangalore Chapter, Geo-Innovations, 30-31 October 2014

CRD test data sometimes provide unrealistic results for a significant duration at the early
stage of the test. Although the compressibility curve in terms of effective vertical stress and
void ration is not much affected by these issues, the interpretation of permeability can be
highly erroneous. There have been attempts to understand these issues, but they have still
remained a puzzle for the researchers until recently.
Vikash (2013) reviewed different theories of CRD interpretation and the issues raised by
various researchers regarding interpretations at the early stage data of the test. The rate of
deformation was understood in this study to have a significant impact on the pore-pressure
generation during the test, which was supposed to be considered in the interpretation of
results. At the beginning of test, the rate of loading changes from zero to a prescribed rate.
The same issue occurs when unloading or reloading is initiated during the test, when the
loading rate is reversed from positive to negative or negative to positive rate of deformation.
A new theory based on the moving boundary concept was proposed to account of such
changes in the deformation rates. This theory is supposed to provide consistent interpretation
of the permeability as well as the relationship between void ratio and vertical effective stress.
The interpretation scheme for CRD test based on this theory is summarized in this paper.
It is imperative to take up CRD test in practice to expect better quality of data being
reported in geotechnical investigation reports. Irrespective of what theory is used to interpret
the results, it will still provide more accurate results compared to using empirical relations and
it will have reasonable enough accuracy to eliminate the need of performing the time
consuming IL tests. Efforts have to be made from all fronts to bring this test into main stream
by training more people on this test method and by resolving any perceived regulatory issues.

CRD Test Setup


Figure 1 shows details of a CRD test setup. The consolidation cell and loading frame are two
main components this apparatus, which are the same as used to perform a triaxial test with
pore-pressure measurement. The stainless steel ring of IL tests is used in this test too. An
arrangement is made to attach the ring to the base of the consolidation cell as shown in Figure
1(b). A pressure line is connected at the top of the consolidation cell to apply the desired
amount of back pressure in the test specimen through surrounding water. The loading frame is
used to subject the specimen to a constant rate of deformation. During the test, pore pressure
at the bottom is measured using a transducer and load is measured using the load cell attached
to loading plunger at the top. The displacement is measured using a dial gauge or an LVDT.

CRD Test Procedure


1. The consolidometer ring is carefully pushed into the bulk sample and trimmed to flush
with the help of a wire saw to obtain smooth surface at top and bottom of the ring. A
small part of the remaining sample is collected for water content analysis.
2. The specimen in consolidometer ring with porous stones at top and bottom are
assembled in the cell in such a way that no air can entrap between the bottom porous
stone and the specimen. The perspex cell is then attached to the base of the
consolidation cell.
3. The whole assembly is then placed in the standard triaxial loading frame and distilled
and deionized water is filled in the cell to submerge the consolidation ring.
4. A suitable back pressure can then be applied into the cell and the specimen is left as it
is for 30 minutes or so which allows the pore pressure inside the specimen to
equilibrate with the applied back pressure while improving saturation of the specimen.
5. The dial gauge or LVDT is adjusted for the initial reading and a seating pressure of
about 10-15 kPa is applied at the top of the specimen to make proper contact between
the soil specimen and the top loading platen.

2
Golden Jubilee Conference of the IGS Bangalore Chapter, Geo-Innovations, 30-31 October 2014

(a) Complete Test setup

(b) Arrangement for consolidometer ring to the base (c) Parts of consolidometer ring.

Figure 1: Experimental setup of CRS consolidation test apparatus

6. ASTM D4186-06 suggests that the absolute value of the excess pore-water pressure
ratio should be greater than 0.03 at any time during the test. Considering this,
appropriate deformation rate is selected for loading as well as for unloading.
7. The specimen is loaded up to a desired level and then it is unloaded as desired. The
data is recorded manually or by using a data-acquisition system.
As stated earlier, several theories for CRS consolidation exist (Smith and Wahls, 1969;
Wissa et al., 1971; and Yoshikuni's et al., 1995) which can be used to determine important
consolidation and strain-stress parameters from the test. These parameters include the
relationship between void ratio (e) and vertical effective stress ('), coefficient of
permeability (k), and coefficient of consolidation (Cv). Wissa’s theory is generally adopted in
the existing codes and the interpretation of data using this theory is presented here.

Interpretation of CRD Data Using ASTM code and Wissa’s Theory:


Wissa et al. (1971) derived a comprehensive solution for CRS consolidation which provides
strain distribution across the depth of the specimen as a function of non-dimensional depth (X)
and time factor (Tv). The expression for the strain distribution is as follows,

3
Golden Jubilee Conference of the IGS Bangalore Chapter, Geo-Innovations, 30-31 October 2014

( , )= 1+ (2 − 6 + 3 )− ( , ) (1)
Where

( )
( , )= ∑ (− ) (2)

and r is the rate of strain, X = 0 at base and X = 1 at top of the specimen. The expression for
the strain distribution consists of two parts, the first part is steady state component and the
second is transient state component which is denoted by T(X,Tv). The transient state
completely dissipates for Tv > 0.5. Further, end of the transience is defined with function
F3(Tv) = 0.4 where,
( , ) ( )
( )= = (3)
( , )

The steady-state condition prevails for F3(Tv) > 0.4. For steady state condition, they derived
the expression for vertical effective stress and coefficient of permeability as,

= − (4)

= (5)

One should note that the expressions (4) and (5) apply only to the steady state condition, and
hence they cannot be used to interpret the test data of transient state.
The CRD test was performed on two different clays of different permeability. The
specimens were first loaded at the rate of 0.02mm/min up to total stress of 1000kPa and then
unloaded at the rate of 0.01mm/min up to 200 kPa. The observed total stress and base pore
pressure (ub) for both the clays are shown in Fig. 1. As the first step to interpret the test data,
F3 has been determined using equation (3), which lied between 0.7 to 0.8 for both Clay 1 and
Clay 2. This indicates that the transeince state did not occur in the present tests, and hence the
steady state condition prevailed right from the beginning of the test in both the clays. Figure 2
shows the variations of permeability k with the void ratio e as interpreted using equation (5)
for both clays during loading stage. The value of k decreases rapidly with e at the early stage
(OA) and then varies moderetly at a later stage (AB). Such a large decrease in value of k in
the early stage of test is unrealstic, which shows a linear trend at the later stage of the test.

Figure 1: CRD test measurements (a) total stress with time (b) Pore-water pressure with time
for Clay 1 and Clay 2.
4
Golden Jubilee Conference of the IGS Bangalore Chapter, Geo-Innovations, 30-31 October 2014

Figure 2: Predicted permeability k using Wissa’s theory during loading stage.

Similarly, the predicted values of permeability during unloading stage are plotted in Fig. 3,
which shows again a trend of high values of k at the early stage of unloading and then almost
a linear trend at later stage at lower k values. The values just before unloading should match
with the values at the end of loading stage, which show almost 100 fold variation. This
observation indicates that Wissa’s theory provides unrealsitic interpretation at the early stages
of the test, and also at the stage just after when the direction of the loading is changed. Similer
observation can also be made for the coefficient of consolidation from Figure 4.
The values of effective stress at any stage of test are computed using equation (4) based
on the assumption of parabolic distribution of stress from top to bottom of the specimen.
Figure 5 shows the predicted effective stress versus void ration relationship for both Clay 1
and Clay 2. It is worth noting, here that the experimental observations from Sheahan and
Watters (1997) indicated that the pore pressure distribution at early stage of loading is not
parabolic, but it becomes parabolic after certain stage of loading at constant rate of
deformation. Hence, the interpreted effective stress versus void ratio relationship may be
slightly erronious at the early stage of test, which was found to range beyond the pre-
consolidation stress of the specimen. Hence, the preconsolidation stress could be shifted a
little bit but not too significantly.

Figure 3: Predicted permeability k using Wissa’s theory during unloading stage.

5
Golden Jubilee Conference of the IGS Bangalore Chapter, Geo-Innovations, 30-31 October 2014

Figure 4: Variation of cv with e interpreted using Wissa et al.’s theory for Clay 1 and Clay 2 .

Figure 5: Compression response of Clay 1 and Clay 2 .

Remaining Issues in Interpretation using ASTM code and Wissa’s Theory:


The analysis presented in previous section indicates that the ASTM code and Wissa’s theory
of interpretation provide unrealistic results at the early stage of the test. Similar observations
were also reported by many researchers in the past (Gorman et al., 1978; Armour and
Drenevich, 1986; Lee et al., 1993; Shehan and Watters, 1997). They found that at the early
stage of the test, the existing theories provide unrealistic interpretation for k and Cv, and show
a notable disparity between e versus ' relationship obtained from CRD consolidation test and
that obtained from IL consolidation test. They argued that it might occur because of higher
strain rate, and hence they suggested various recommendations on the strain rate at which test
can be performed. Wissa et al. (1971) suggested that the selection of a proper strain rate
depends on the parameters required. If the compression response is required, the test should
be performed at a strain rate such that the base pore pressure ub remains close to zero. If k and
Cv are required, a strain rate is set such that the finite hydraulic gradient develops within the
specimen. On basis of these conditions, Wissa et al. (1971) suggested that pre pressure ratio
Ru (= ub/) could be between 0.02 to 0.05. This also forms the basis of recommendations in
ASTM D4186-6. Further, Lee et al. (1993) suggested that the rate of deformation is
acceptable if the coefficient of consolidation derived from an analysis of the drained face
equals that from the undrained face. They found that CRD consolidation test results and IL

6
Golden Jubilee Conference of the IGS Bangalore Chapter, Geo-Innovations, 30-31 October 2014

consolidation test results are nearly the same at this deformation rate. Wissa's theory gives the
following expressions for Cv at the drained and undrained face. Cv at the drained face is given
by

= (6)

where h is the current thickness of the specimen, ub is the excess pressure at the undrained
face, i.e., at the base, and d' is the change in the effective stress at the drained face in time
t. Cv at the undrained face is given by

= ∆
(7)

where ud' is the change in the effective stress at the undrained face. According to Lee et al.
(1993), the test should be performed at a deformation rate for which the two coefficients,
given by equations (6) and (7), converge.
The criteria given by Lee et al. (1993) is further analyzed herein to investigate the concept
behind it. Since the excess pore-water pressure is zero at the drained face, therefore d' will
be equal to. ud' will be equal to  - ub since the excess pore-water pressure at the
undrained face is ub. Hence, equations (6) and (7) can be modified to equation (8).


− = (8)

Equation (8) depicts that the coefficient of consolidation at the drained and undrained face
will converge only when ub/t becomes close to zero. This condition may exist sometimes at
the later stage of test when the slope of the curve between ub and t becomes moderate. This
condition is better satisfied at extremely low deformation rate, but not at early stage of the
test. It can be deduced that the current theories are unable to describe the rapid variation of ub
with t at the early stage, and this is why these theories yield unreliable results at the early
stage of the test. Hence, if the mechanism explaining the evolution of ub with time t at the
early stage is incorporated in the theory for CRD consolidation, then the issue of unrealistic
interpretation at the early stage could be resolved. Consequently, the criterion for allowable
strain rate will not be required, and hence CRD consolidation test could be performed at
higher deformation rate.

Interpretation of CRD Data Using New Proposed Theory:


The theory proposed by Vikash (2013) assumes that at the beginning of CRD consolidation
test the diffusion process occurs only in the region close to the top pervious surface. The
depth of this region increases as the test progresses, and after a sufficient time, the whole
depth begins to participate in the diffusion process. The proposed mechanism is depicted in
Fig. 6, which indicates that the specimen can be divided into two regions at the early stage of
the test. The region close to the top drained surface participates in the diffusion process;
whereas; the lower remaining part of the specimen has no diffusion. The depth of these
regions have been denoted by Hd and Hud, respectively. Hd is nearly zero at the beginning of
the test, and it increases with the progress of the test and eventually covers the whole depth of
the specimen after some time.
It is assumed that the entire specimen participates into the diffusion process at time t = .
Therefore, at any instant t < , the undrained boundary will lie at the depth Hd from the top of
the specimen and Hd remains at the bottom of the specimen for t ≥ . The value of  can be
calibrated using the base pore pressure data as shown in Fig. 7 for Clay 1 and Clay 2. It is
assumed to have a bilinear response of base pore pressure between the early stage of the test
with moving boundary and the later stage of the test when complete specimen participates in

7
Golden Jubilee Conference of the IGS Bangalore Chapter, Geo-Innovations, 30-31 October 2014

Figure 6: Schematic illustration of moving boundary condition in the pore-water diffusion


during CRD consolidation,(a) at early tage, (b) at an intermediate instant, (c) at instant when
the whole depth participates into diffusion process.

Figure 7: Determination of the characteristics time for diffusion process using the test data of
first loading stage.

the diffusion process. Further, Terzaghi’s solution with parabolic pore pressure distribution is
assumed to be valid for the region above the moving boundary, where the pore pressure
remains constant in the region below up to the specimen bottom. The governing differential
equation considering the moving boundary of diffusion provided the following solution for k.

= (9)

This equation comes out to be similar to Wissa’s solution in equation (5), with one difference
that the thickness of the specimen H is replaced here by Hd, the depth of undrained boundary
at any stage. At any stage of test, the value of = / . Using this formulation, the
predicted values of permeability are plotted in Fig. 8 for both Clay 1 and Clay 2, which shows
almost linear trend of variation of k with e in semi-log scale. The values of predicted k for
unloading stage of Clay 1 are also shown in Fig. 9, which indicates that the predicted value of
k before and after the onset of unloading stage are of the same magnitude. Hence, the solution
from the proposed new theory is consistent for prediction k throughout the test.

8
Golden Jubilee Conference of the IGS Bangalore Chapter, Geo-Innovations, 30-31 October 2014

Figure 8: Predicted k with e from Wissa theory and the new proposed theory during loading

Figure 9: Predicted k with e from Wissa theory and the new proposed theory during
unloading stage of CRD test

The effective stress at any stage can be calculated using the following equation which is
derived according to the expected pore pressure profile with the moving boundary.

= − 1− (10)

The base pore pressure ub was relatively small in comparision to the total stress during these
tests. This resulted into almost the same values of v' from equations (4) and (10) based on
two different theories. The consistancy of solution in the new theory, however, allows the test
to be performed at much higher deformation rates which can produce much higher values of
ub. In such a condition, the difference in the interpreted v' from two theories will also
increase. Hence, the new theory is expected to allow for faster testing with reasonably more
accurate interpretation of e versus ' relationship.

Conclusions
The CRD test provides a good alternative to IL test of consolidation for characterization of
soil in major projects. The testing method and interpretation schemes are simple enough for

9
Golden Jubilee Conference of the IGS Bangalore Chapter, Geo-Innovations, 30-31 October 2014

the practice to adopt. Despite certain limitations of interpretations posed by existing theories
adopted in codes, this test still provides reasonable values of the soil properties which is far
better than depending on the uncertain predictions from empirical relations based on SPT or
CPT data. The limitations of the existing theories pose challenge mainly at the early stage of
the test or in the unloading reloading phases of the test; however, the stress-strain relationship
or stress versus void ratio relationship is till reasonably predicted which is sufficient for
prediction of settlement in foundation or to compute deformation response of other structures.
The new proposed theory based on moving boundary concept, however, provides solution to
the issues of early stage discrepancies in the interpretation of CRD data and interpretation of
unloading-reloading phases of tests. By eliminating the issues, this proposed can be used to
interpret the results from CRD test at much higher deformation rates than the prescribed limits
in the codes. This further reduces the duration of the test and makes the test much faster to get
results quickly. It is expected that the practice will adopt this method for faster
characterization of soils and feel encouraged to put the representative experimental data in
geotechnical investigation report.

References
Armour, D. W. J. and Drnevich, V. P. (1986). “Improved techniques for the constant -rate-of-
strain consolidation test.” Consolidation of Soils, STP 892, American Society for Testing and
Materials, Philadelphia,, 170–183.

ASTM:D4186-06 (2008). “Standard test method for one-dimensional consolidation


propertiesof saturated cohesive soils using controlled-strain loading.
Gorman, C. and Drnevich, V. P. (1978). “Constant-rate-of-strain and controlled-gradient
consolidation testing.” Geotechnical Testing Journal, 1(1), 3 – 15.

Hamilton, J. J. and Crawford, C. B. (1959). “Improved determination of preconsolidation


pressure of a sensitive clay.” Special Technical Publication No. 254, American Society for
Testing and Materials, Philadelphia, 254–270.

Lee, K., C. V. L. S. H. and Queek, S. H. (1993). “Constant rate of strain consolidation of


singapore marine clay.” Geotechnique, 43(3), 471–488.

Sheahan, T. C. and Watters, P. J. (1997). “Experimental verification of crs consolidation


theory.” Journal of Geotechnical and Geoenvironmental Engineering, ASCE, 123(5) , 430–
437.

Smith, R. E. and Wahls, H. E. (1969). “Consolidation under constant rate of strain.” Journal
of the Soil Mechanics and Foundations Division, ASCE, 95(2), 519–539.

Vikash, G. (2013) “A New CRS Consolidation Theory and A Compressibility Model for
Kaolin Clay with Microfabric Effects”, Ph.D. Thesis, Indian Institute of Technology Kanpur.

Wissa A. E. Z., Christian, J. T. D. E. H. and Heiberg, S. (1971). “Consolidation testing at


constant rate of strain.” Journal of the Soil Mechanics and Foundations Division, ASCE,
97(10), 1393–1413.

Yoshikuni, H., M. T. I. S. and Xo, T. (1995). “Direct determination of permeability of clay


from constant rate of strain consolidation tests.” Proceeding of the International Symposium
on Compression and Consolidation of Clayey Soils, Vol. 1, PP, 609-614.
10

You might also like