Professional Documents
Culture Documents
Stability Analysis of Embankments On Soft Ground-Low
Stability Analysis of Embankments On Soft Ground-Low
ON S O F T GROUND
By Bak-Kong Low, 1 Member, ASCE
INTRODUCTION
NOTATIONS
Trial Limiting
Tangent # 1
Trial Limiting
Tangent # 2
> 1111 > i > 11 > 111 > 11 > 11111 > 1 1 1 1 1 1 1 1 1 1 1
The factor of safety as shown in Fig. 2 has been derived using a number
of approximations (Low 1985). A brief outline is given below.
Expressions were obtained for the overturning moment M0 and the resist-
ing moment MR corresponding to an arbitrary slip circle that is tangent to a
horizontal plane at a depth D below the top of the clay layer. The coordinates
of the critical circle center were then determined by taking partial derivatives
of the expression for Fs (the ratio of MR to M0). The partial derivatives
showed that the center of the critical circle is located on a vertical line pass-
ing through mid-slope of embankment. By substituting the expressions for
the coordinates of the critical circle center into the expressions for MR and
M0 (Appendix I), the final equation for the minimum Fs was obtained.
Then, as shown in Fig. 2, the minimum Fs corresponding to a trial limiting
tangent at depth D is given by the equation
ey
Downloaded from ascelibrary.org by Universidad De Sevilla on 06/03/15. Copyright ASCE. For personal use only; all rights reserved.
T r i a l l i m i t i n g tangent
( :
A / m \
(Fs>D = N l + N 2 - •7i + X tan <f)m)
7H \7H /
where t^ = Nj_ ( — , cot /3 )
N N
2 = 2 ( — ' c°t|S)
X = X ( — , cot ]8 )
v
H I
C A = Average undrained shear strength
within the depth D, computed
using a derived equation.
FIG. 2. Notations for Embankment on Weak Foundation
The coefficients A^, N2, and 7 in Eq. 1 can be determined from the charts
shown in Figs. 3 and 4. Their values depend on D/H and cot p. The symbol
7 stands for embankment unit weight in both terms of Eq. 1. For base circles
the ratio D/H has to be greater than zero. In Figs. 3 and 4 the curves begin
with D/H values of about 0.5.
The term CA in Eq. 1 corresponds to the undrained shear strength of the
foundation material when the soil is uniform. A procedure will be described
later so that Eq. 1 can also be applied to soils with a nonuniform strength
profile.
The subscript "D" associated with the term Fs in Eq. 1 means that the
equation computes the minimum Fs corresponding to a given trial limiting
tangent of depth D. For cases where undrained shear strength of the foun-
dation increases with depth, it may be necessary to consider several different
tangent depths, in order to locate the one that results in the minimum factor
of safety.
Each of the terms on the right hand side of Eq. 1 is dimensionless. The
term CA/yH may be regarded as the normalized foundation strength param-
eter SF, and the term Cm/yH + 7 tan 4>,„ as the normalized embankment
213
D/H D/H
strength parameter SM, so that the factor of safety (FS)D is the sum of two
components:
<F.)D = NrSF + N2-SM (2)
Note from Figs. 3 and 4 that, for D/H > 3, the factor of safety is largely
due to the normalized foundation strength SF, unless the normalized em-
bankment strength SM is significantly greater than the normalized foundation
strength SF. Conversely, for D/H < 1 (i.e., shallow base circle), the em-
bankment strength may contribute significantly to the factor of safety, the
more so if SM is greater than SF.
It is perhaps worth emphasizing that the unit weight 7 which appears in
both the normalized foundation strength and the normalized embankment
strength terms in Eq. 1 is the unit weight of the embankment material. For
undrained conditions in the clay foundation, and for a horizontal foundation
surface, the unit weight of the foundation clay has no influence on the factor
of safety and hence does not appear in Eq. 1.
The procedure leading to Eq. 1 assumes circular slip surfaces and con-
siders only moment equilibrium. It differs from the general procedure of
Bishop's simplified method or the ordinary method of slices in that it does
not involve dividing the slope into vertical slices.
Fig. 5 defines the factor of safety as used in Bishop's simplified method.
The numerator and the denominator represent the resisting and the over-
turning moment respectively. The term N' is the effective normal force on
214
<4°L
\ N'
B i s h o p ' s Method:
£ ( c ' / + tan (f)' • N')
F
s =
,w s i n a
W — S i n cc - ujtcos cc
F
where: N' = s
+ t a n (t> ' s i n a
c o s cc
the slip surface. Bishop's simplified method assumes the interslice forces to
be horizontal. By considering vertical equilibrium, an expression for N' can
be obtained, which is a function of the factor of safety Fs. Since the term
Fs occurs on both sides of the equation in Fig. 5, successive approximation
must be used to evaluate Fs. In addition, an adequate number of trial failure
surfaces have to be examined in order to obtain the minimum value of Fs.
The method is capable of handling complex slope geometry and soil con-
ditions.
In the procedure summarized in Fig. 2, the normal force on the slip sur-
face is considered only for that part of the arc within the embankment. The
approximation used for this normal force is described in Appendix I. No
important discrepancy arises from the different assumptions regarding the
normal force. This may in part be attributed to the fact that the frictional
215
in the procedure proposed in this paper and also to that of the Ordinary
Method of Slices.
The proposed procedure as summarized in Fig. 2 is much easier to use
than the method of slices, and is particularly convenient in hand computa-
tion. However, the simplicity of the equation is achieved at the expense of
some versatility. For instance, if the friction angle of the foundation material
is not equal to zero, or if the shear strength varies laterally, then a more
general computer method like Bishop's simplified method has to be used.
10.0
embankment
216
[See, for example, figure 35.3 midpoint circle curves in Terzaghi and Peck
(1967)]. The corresponding values between the two figures are almost iden-
tical. For the type of problems under consideration, the values of Ni and iV2
plotted separately as in Fig. 3 are more useful than their combined values
shown in Fig. 6. The stability can be calculated when the shear strength of
the embankment and of the foundation are not the same.
Fig. 6 shows that when cot (3 > 1 and the undrained shear strength is
constant for both the embankment and the foundation, as would be the case
for a cut slope in uniform clay, the critical circle will extend to the bottom
of the uniform clay (i.e., D/H tends to infinity). However, when the shear
strength of the embankment is different from that of the foundation, the
depth of the critical circle will depend mainly on the relative strength of the
embankment and of the foundation material and to a lesser extent on cot (3.
For two embankments resting on the same soft clay, the critical failure circle
will extend deeper down for the embankment with the higher normalized
embankment strength because both N2 and 7 decrease with depth (Figs. 3
and 4).
hcA
Downloaded from ascelibrary.org by Universidad De Sevilla on 06/03/15. Copyright ASCE. For personal use only; all rights reserved.
WWm Trial L
Lii m i t i n g Tangent
g'_ p\
iiiiiiiiiiuiiiiiiniiiiiiiiu/ IIIIIWHIIIIIIIHIIIIIUIII,
ACA=0.35(DC/D)1-1 ACT
7TT\
iiii/iifiiiiii/iini iiiiniiniuiiiniuiiinun,
C'K = 0 . 3 5 C T + 0 . 6 5 C D + 0.35(Dc/D)l-lACT
CA + ACA
A
HI M
mm• D
1 Trial Limiting Tangent
iiiiDiiiiiniiiim, 77777777777777/7777777777777777;
—-Trial Limiting
Tangent Depth
A p p l y Eqn 4 A p p l y Eqn 5
3 l
•A12M+A23K
77777777777777777777
Note that both Eq. 4 and Eq. 5 take into account the curvature of a circular
arc, This can be seen from Eq. 4 where more weight is given to CD than
to CT. These two equations are to be used to average the shear strength of
the foundation clay only.
Downloaded from ascelibrary.org by Universidad De Sevilla on 06/03/15. Copyright ASCE. For personal use only; all rights reserved.
ILLUSTRATIVE EXAMPLES
Example Number 1
Three horizontal clay layers each 4.5 m thick are shown in Fig. 10. The
values of c for the upper, middle, and lower strata are, respectively, 30, 20,
and 150 kN/m 2 . The unit weight is 18 kN/m 3 . A cut is excavated with side
slopes of 1V:3H to a depth of 6 m. For the purposes of computation and
the following discussion, the portion of the slope above the bottom of the
cut is treated as the embankment. It may be observed that the contributions
of the embankment and of the foundation material towards the total factor
of safety are 35% and 65%, respectively (Fig. 10). The relatively significant
contribution from the embankment is due to the relatively shallow base circle
(small D/H) and the relatively high normalized shear strength of the em-
bankment which in this case is greater than the normalized shear strength of
the foundation. For comparison, results from computer program (Duncan
and Wong 1984) which systematically locates the critical slip circle are also
shown in the figure. The Ordinary Method of Slices and Bishop's simplified
method are identical in this example because <$> is zero.
4 . 5 m C = 3 0 kN/B
7 = 18 kN/m i-1.5 m C = 20 kN/m 2
3
2~W
V 4 . 5 m c = 150 kN/m 2
20 27.5
Minimum F_ 5.1 x (Eq 1)
18 X 6 18 x 6
= 0.944 + 0.509
= 1.45
Compared with computer programs : F s = 1.44 (Ordinary Method of
slices)
F s = 1.44 (Bishop's simplified
method)
220
8m K 4m
4m
/ / / / / J / / / / / / / / / / / / / ' s f t * / / / / / " / J * /
\.
3 0 kN/m^
Downloaded from ascelibrary.org by Universidad De Sevilla on 06/03/15. Copyright ASCE. For personal use only; all rights reserved.
N
C
A »1 2 X < F S>D
D
(FigfS) (Figs 3,4) (Eq 1)
Example Number 2
Fig. 11 shows an embankment of stiff clay which has been constructed
on soft clay. The undrained shear strength of the soft clay varies with depth.
The hand-computed value for the factor of safety against circular rotational
failure is 1.38. The contributions of the foundation and of the embankment
materials to the factor of safety are shown in the last column of the table.
Computer program (systematic searching for the critical circle) gives val-
ues of 1.36 and 1.14 based on Ordinary Method of Slices and Bishop's
simplified method, respectively.
The discrepancy between the values given by the Ordinary Method of
Slices and by Bishop's simplified method has been investigated for cj> = 0
in the embankment material. Both methods indicated a factor of safety of
1.31. That both methods give identical value for the factor of safety when
cp = 0 is theoretically consistent. That the value given by the Ordinary Method
of Slices is reduced (from 1.36 to 1.31) when <> j drops to 0 is also reasonable.
It is, however, illogical that the value given by Bishop's simplified method
increases from 1.14 to 1.31 when friction angle for the embankment material
drops from 15° to 0°, all other conditions remain unchanged. The value of
1.14 is therefore misleading.
Errors similar to that associated with the value of 1.14 given by Bishop's
simplified method may arise when there are slices with steeply inclined bases,
or when the embankment material has a high cohesion. In these circum-
stances the computed normal force on the base of some slices may be either
negative or unreasonably large (Chirapuntu and Duncan 1975; Whitman and
Bailey 1967). For the case in Fig. 11, the seeming paradox that the Fs value
given by Bishop's simplified method decreases with increasing angle of in-
ternal friction of the embankment can be explained by considering the
expression for N' in Fig. 5, with no pore pressures. The numerator in the
expression (and hence the normal force N') can be negative for slices in the
221
ACKNOWLEDGMENTS
Overturning Moments
The factor of safety for rotational failure can be defined in terms of overall
moment equilibrium. In Fig. 12, the origin of the coordinate system lies at
the level of the trial limiting tangent, on a vertical line passing through the
Downloaded from ascelibrary.org by Universidad De Sevilla on 06/03/15. Copyright ASCE. For personal use only; all rights reserved.
Resisting Moment
The resisting moment MR along IGEJ consists of two components: that
contributed by the available undrained shear strength along arc IGE, and
J
?8 ^ —
y = R
"cot/3 ^ X j B /
H
E 1
1 / A \ /
Trial 1
Limiting Tangent V • •
^ Origin
777 777 777 777 77? 777 777—
Resisting Moment along IGEJ :
M„=[ix ( / - x ) - | - / 2 + y (D+lj-itD+fj^^JTH
3.14
3.00
1.00
0.00 0.00
that contributed by embankment cohesion and friction along arc EJ. To ob-
tain the available resisting moment, it is necessary to know the arc length
of the circular slip surface. In Fig. 13, L = 2Qy, where 9 is in radians. The
exact equation for 6 involves the inverse cosine, cos -1 . To facilitate the par-
tial differentiation that will be done later, the angle (in radians) 26 is ap-
proximated by 3.06(D/v)053. The lower plot in Fig. 13 shows that the ap-
proximation is quite accurate. Therefore
For foundation material with constant undrained shear strength CA, the
available resisting moment can be expressed as
MR = y • (CA • 26y) + y • (C„, + \yH • tan <|>m) • QHy . (8)
where the notations are as shown in Fig. 12.
The first term on the right comes from the foundation material while the
second term derives from the embankment. The term XyH • tan <f>„, represents
224
11111111111111111 I I I I I I I I I I I I I I I I I
CASE A CASE B
JE almost vertical JE almost horizontal
F J E =. Frictional force on JE F
JE - Frictional force on JE
= Average an x tan 0 m = Average <7n x tan 0 m
X Length JE x Length JE
1 1
a— y HK„tan 4>m — — Y H tan </>m X Length JE
2 2
X Length JE
1 1
i.e. X = — K 0 i.e X =-
2 2
the average frictional stress along that part of the arc length which lies in
the embankment. By considering two extreme possibilities, one can deter-
mine the possible range of the coefficient X which appears in this term. In
Case A of Fig. 14, arc JE is close to being vertical. The average normal
stress on JE is 0.5 K0yH, where K0 = coefficient of lateral earth pressure
at rest. In Case B of the same figure, JE is almost horizontal, with average
normal stress equal to 0.5yH. Therefore the coefficient X lies within the
range KB/2 to 0.5. Parametric studies have been carried out to calibrate X
as a function of D/H and cot p. The equation is shown in Fig. 15. As plotted
in Fig. 4, the value of X varies from 0.19 for deep circles to 0.38 for shallow
circles.
Substitution of 26v and BH v of Eq. 8 by Eq. 7 results in the final expres-
sion for the available resisting moment MR as shown in Fig. 12. Note that
both the available resisting moment MR and the overturning moment M„ in
Fig. 12 are for an arbitrary slip circle with center at (x,y) and tangential to
a trial limiting tangent at depth D. Dividing the resisting moment by the
overturning moment gives an expression for the factor of safety for the par-
ticular circle with its center at (x,y):
MR
Fs(x,y,D,...) = (9)
M„
/ / i i i / /
( F s ) D = N i S + N 2 -7H
(
7i+Xtan^
D,0.53 1 4 7
w h e r e N n = 3 . 0 6 (— "l " /
Q=2
N2=l.53[(l+l)0-53-(|)0-53]^-47/a2
D 1 COt 2 )3 + 1
C K r = 1 . 5 6 4 ( - ^ + i ) + 0 . 1 3 03 -
1+0.5
0.02cot/3
\^0.19+ / o r F i gy . 4
D/H
(For D/H>0.5)
respond to a stationary value of the Fs for the particular trial limiting tangent.
By the nature of the problem, this stationary value is a minimum. The two
partial derivatives lead to the conclusion that, corresponding to a trial lim-
iting tangent at depth D,
y (cot S)2 + 1 D \
- = 0.1303 ^ — ^ + 1.5638 - + -
H D \ \H 2
+ 0.5
H
Substitution of the above expressions for x and y into Eq. 9 (where MR and
M 0 are as expressed in Fig. 12) leads eventually to the Fs equation and the
stability numbers iV, and N2 as shown in Fig. 15. The minimum Fs corre-
sponding to a trial limiting tangent at depth D is thus given by the equation
226
The equations of Nlt N2, and X in Fig. 15 have been used to plot the
charts shown in Figs. 3 and 4.
Downloaded from ascelibrary.org by Universidad De Sevilla on 06/03/15. Copyright ASCE. For personal use only; all rights reserved.
Chirapuntu, S., and Duncan, J. M. (1975). "The role of fill strength in the stability
of embankments on soft clay foundations." Geotechnical Engineering Report No.
TE 75-3, Univ. of California, Berkeley, Calif.
Duncan, J. M., and Wong, K. S. (1984). "STABR: A computer program for slope
stability analysis with circular slip surfaces, microcomputer version." Geotechnical
Engineering Report NO. UCB/GT/84-09, Univ. of California, Berkeley, Calif.
Leonards, G. A. (1982). "Investigation of failures." J. Geotech. Engrg. Div., ASCE,
108(GT2), 185-246.
Low, B. K. (1985). "Analysis of the behaviour of reinforced embankments on weak
foundations." Thesis presented to the Univ. of California, at Berkeley, Calif., in
partial fulfillment of the requirement for the degree of Doctor of Philosophy.
Terzaghi, K., and Peck, R. B. (1967). Soil mechanics in engineering practice. John
Wiley and Sons, Inc., New York, N.Y.
Whitman, R. V., and Bailey, W. A. (1967). "Use of computers for slope stability
analysis." J. Soil Mech. and Found. Div., ASCE, 93(SM4), 475-498.
227