Indian Geotechnical Conference IGC2016

15-17 December 2016, IIT Madras, Chennai, India

STABILITY ANALYSIS OF NON-HOMOGENEOUS SOIL SLOPES

USING NUMERICAL TECHNIQUES

D. Chatterjee1
A. Murali Krishna2
1
Research Scholar, 2Associate Professor, Department of Civil Engineering, Indian Institute of Technology Guwahati, Assam 781039.
1
dooradarshi@iitg.ernet.in, 2amurali@iitg.ernet.in

ABSTRACT: This paper presents the stability analyses of non-homogeneous slopes under different loading conditions. A
non-homogeneous soil slope with two different soil layers is considered. Special cases are created, varying the height of the
layers, to account for the non-homogeneity of the earth slope. A water table is considered in this study to account for the
seepage forces. Pseudo-static earthquake force is considered, taking into account both the forces in the horizontal and vertical
directions. A rigorous limit equilibrium method of slices i.e. Morgenstern-Price method is used to analyze the stability of
the slope. Finite element shear strength reduction technique is also used for displacement calculations and comparison with
limit equilibrium method. Safety factor, critical slip surfaces and displacements with different loading conditions are studied
and compared.
Keywords: Numerical model, non-homogeneous slope, displacements, critical surface

2 METHODOLOGY
1 INTRODUCTION
Various methods are available for stability
Slope stability is a major area of concern for practicing determination of earth slopes, they have their own
engineers for filled slopes or cut slopes. Factor of safety advantages and limitations. Limit equilibrium method
is an important parameter in stability evaluation of gives the factor of safety of slope and position of critical
slopes but the total displacements occurring can give a slip surface. It doesnt give information on deformations
better idea about the zone of failure. Position of critical occurring within the slope so to overcome this, finite
slip surface evaluation is another important issue in element method is selected as well.
slope stability. Slopes can be man-made or natural and
many external natural forces act on these earth structures 2.1 Limit Equilibrium Method
making it vulnerable to failure. Out of these external It is a statically indeterminate method, hence for the
forces, the effects of hydrology and seismicity are the solution; some assumptions are required to make it
most common. Rainfall affects the stability of slopes in determinate. Therefore, anon-circular failure surface is
numerous ways like the increase in height of water table assumed and the soil mass is divided into a number of
in an area due to heavy downpour or the formation of slices and each slice is checked for equilibrium of forces
phreatic line. In high seismic zones slopes suffer greater and moments. Figure 1 depicts the various forces acting
dynamic loadings which may eventually lead to on a typical slice.
instability. Varying soil properties make the slope non-
homogeneous in nature. Many studies are available in
literature like the effect of variation of cohesion with
depth has been investigated by Koppula (1984). Kim et
al. (2002) compared stability of complex soil slopes
using upper bound and lower bound limit analysis.
Hammouri et al. (2008) studied stability of layered
slopes and showed effects of drawdown, tension crack
and undrained clay soils. Qian et al. (2015) proposed
stability charts for two-layered cohesive slopes. This
paper deals with a two-layered non-homogeneous slope
whose layer height is varied, water table is introduced,
pseudo-static forces are applied and then the stability of
the slope is evaluated.
Fig.1 Slice representing various forces

1
Stability Analysis of Non-Homogeneous Soil Slopes Using Numerical Techniques

Depending upon the type of equilibrium satisfied, permeability values of the two soils are 410-7m/s and
various methods are formulated for the factor of safety 8.210-6m/s, respectively.
of slope such as ordinary method of slices (Fellenius,
1936), Bishops Simplified method (Bishop 1955) and
Janbus generalized method of slices (Janbu 1968).
Complete equilibrium is followed by Spencer (1967)
and Morgenstern and Price (1965) methods. Then a
search procedure is used to find the critical slip surface
giving the minimum factor of safety of the slope.
Morgenstern and Price (1965) method with a half sine
function has been used throughout this study. Circular
slip surface is assumed for homogeneous profiles while Fig.2 Non-homogeneous slope model
non-circular is taken for non-homogeneous profiles. The
number of slices is fixed at 20 while the tolerance is Stability of slopes being an unconfined problem, the
selection of the dilation angle doesnt make much of a
taken to be 0.005.
difference in the analysis (Griffiths and Lane, 1999). A
2.2 Finite Element Method non-associated flow rule has been used throughout this
study. The soil model consists of 2700 uniform
Zienkiewicz et al. (1975) studied c- slopes using the triangular 6-noded elements and 5500 nodes. The height
finite element method showing good comparison with of the top layer is varied from zero to total depth, DH.
slip circle results. With more use and confidence gained, In essence the first and the last case becomes a
researchers like Matsui and San (1992), Ugai and homogeneous slope with either of the soil layers. Table
Leshchinsky (1995), Griffiths and Lane (1999) and 1 shows the variation of the layer height. The h values
Cheng et al. (2007) used it for stability analysis of shown in table represent the depth of layer from the
slopes. The safety factor, F is defined as the factor by ground surface. When the ratio is increasing the soil in
which the strength parameters need to be reduced to the upper region is actually forming the major portion of
bring the slope to a failure point. the slope. A pseudo-static seismic coefficient of 0.2 is
,
= (1) taken representing violent earthquakes. Both horizontal
and vertical pseudo-static acceleration are taken equal as
tan = (2) suggested by Shukha and Baker (2007). A water table is
introduced and classified as case-2 while the other case
This method is known as the shear strength reduction i.e., case-1 is free of pseudo-static force and water table.
technique. Here failure is taken as that point where the
Table 1 Variation of h/D ratio
solutions fail to converge within a specified number of
iterations. The stress distribution fails which implies that h/D 0 0.2 0.4 0.6 0.8 1
the Mohr-Coulomb failure criterion and the global h (m) 0 8 16 24 32 40
equilibrium is not satisfied. Failure of slopes is
accompanied by a sudden rise of displacements. 4 RESULTS
3 DEVELOPMENT OF NUMERICAL The slope has been analyzed using two different
MODEL approaches, their results and comparison is shown in
Tables2 and 3 for homogeneous and non-homogeneous
A numerical model (Figure 2) has been developed in to
cases, respectively. Difference in safety factor values
represent the homogeneous and nonhomogeneous slope
between the two methods range from 0.19-12.45% for
in Rocscience software suite using Phase and Slide
the non-homogeneous slopes.
modules (Rocscience, 2016). Soil-1 is silty clay while
soil-2 is silty sand in nature. The slope angle is taken as Table 2 Homogeneous slope with two different soils
2:1 (H:V) and the slope height is 20m and depth factor
Slope Factor of Safety
D is 2. The Mohr-Coulomb failure criterion is used for
Model LEM FEM Difference
modelling the soils. Drained values are selected for
(%)
shear strength parameters of the two soils. Youngs
modulus value for the two soils are taken as 5MPa while Case- Case Case Case Case Case-
1 -2 -1 -2 -1 2
the Poissons ratio is 0.3. Total unit weight of 16 kN/m3
is applied to the layered soil model. Cohesion values for Soil-1 1.11
soil-1 and 2 is 10kPa and 0kPa, respectively. Friction h=D 1.687 4 1.68 1.05 0.41 5.74
angle values are 30o and 36o, respectively. The Soil-2 0.99
h=0D 1.453 9 1.48 0.98 1.85 1.90

2
 Indian Geotechnical Conference IGC2016
15-17 December 2016, IIT Madras, Chennai, India

Table 3 Non-homogeneous slope with two soil layers equilibrium method is small and confined to the toe
region.
Layer Factor of Safety
Height LEM FEM Difference(%)
Case- Case- Case- Case- Case- Case-
1 2 1 2 1 2
h=0.2D 1.478 1.018 1.55 1.02 4.87 0.19
h=0.4D 1.578 1.088 1.64 1.05 3.92 3.49
h=0.6D 1.736 1.188 1.68 1.05 3.22 11.61
h=0.8D 1.736 1.188 1.68 1.05 3.22 11.61

The factor of safety increases with layer height and

becomes constant after h=0.6D. The increase in content
of the silty clay layer within the slope increases its safety Fig.4 Critical surfaces for various layer heights for case-2
factor value. In terms of the two cases considered, the using the limit equilibrium method
factor of safety from case-2 show a huge decrease from
case-1. Figure 3 depicts the critical slip surfaces for The maximum displacement is 0.067m. The same is
various layer heights corresponding to case-1. Critical observed for h=0.4D but here the zone of failure
surface for homogeneous slope with silty clay soil show increases but the slip surface from limit equilibrium
a toe type of failure. Shallow slope failure is observed method is much smaller. For this height the maximum
for layer heights h=0.2D, 0.4D and homogeneous slope displacement is 0.11 m. For h=0.6D and h=0.8D the
with silty sand soil. Critical surface for layer heights zone of failure increases and it is seen spreading
h=0.6D and 0.8D coincide with each other having a outwards.
deeper slope failure. The increase in the silty clay layer
shifts the critical slip surface deeper into the slope a
simultaneously increasing the factor of safety as well.

c
Fig.3 Critical surfaces for various layer heights for case-
1using the limit equilibrium method
Critical slip surfaces for various layer heights
corresponding to case 2 is shown in Figure 4. For
h=0.2D and h=0.4D the slip surface is shallow and
parallel to the slope while for the other two heights they
coincide with each other having a deep slip surface. d
From the figure it is evident that the slope with more
silty sand soil has a critical slip surface which is shallow
and parallel to slope while with more silty clay soil the
failure is a deep slope failure. The total displacement
contours and the critical slip surface from the limit
Fig.5Contours of total displacements from FEM and
equilibrium method for case-2 and four layer heights are critical surface from LEM for case-2 and layer
shown in Figure 5. The displacement increases with height (a) h=0.2D (b) h=0.4D (c) h=0.6D and (d)
increase in layer height i.e., the zone of failure increases h=0.8D
and parallel to the slope while the slip surface from limit

3
Stability Analysis of Non-Homogeneous Soil Slopes Using Numerical Techniques

For the last two layer heights the critical slip surface References
from limit equilibrium method is same and they coincide
Bishop, A.W. (1955) The use of the slip circle in the
while finite element method depicts that their
stability analysis of slopes, Geotechnique, 5(1), pp7-
displacements are not the same, layer height h=0.8D has
17.
more displacement than height h=0.6D. The maximum
displacement for the last two layer heights are 0.122m Cheng, Y.M., Lansivaara, T. and Wei, W.B. (2007) Two-
and 0.126m respectively. The variation of factor of dimensional slope stability analysis by limit equilibrium
safety with layer height to depth factor ratio for case-1 and strength reduction methods, Computers and
and case-2 is shown in Figure 6. Out of the two cases Geotechnics, 34, pp137-150.
considered it is seen that for case-1the factor of safety Fellenius, W. (1936) Calculation of the stability of
values are higher than that of case-2. The factor of safety earthdams, Proc. 2nd congr. large dams, Washington
increases with layer height to depth factor ratio up to DC 4.
height 0.6D and stays constant after that. Griffiths, D.V., and Lane, P.A. (1999) Slope stability
analysis by finite elements, Geotechnique, 49(3),
pp387403.
Hammouri, A., Malkawi, A.I.H., and Yamin, M.M.A.
(2008) Stability analysis of slopes using the finite
element method and limit equilibrium approach, Bull
Eng. Geol. Environ, 67, pp471-478.
Janbu N (1968) Slope stability computations, Soil Mech.
Found. Engg. Report. Trondheim: Technical University
of Norway.
Kim, J., Salgado, R., and Lee, J., (2002) Stability analysis
of complex soil slopes using limit analysis, J. Geotech.
Geoenviron. Eng., 128(7), pp 546-557.
Koppula, S.D., (1984) Pseudo-static analysis of clay
slopes subjected to earthquakes, Geotechnique, 34(1),
Fig.6 Variation of safety factor with height ratio for case-1 pp71-79.
and 2
Matsui, T., and San, K.C., (1992) Finite element slope
5 CONCLUSIONS stability analysis by shear strength reduction technique,
Soils and Found, 32(1), pp5970.
A two-layered non-homogeneous slope with a silty clay
layer at the top and silty sand layer at the bottom is Morgenstern, N.R., and Price, V.E., (1965) The analysis
prepared. Two cases are considered here of which case- of the stability of general slip surfaces,
2 is more critical where a water table and pseudo-static Geotechnique,15(1), pp79-93
force has been introduced. Considering only the non- Qian, Z.G., Li, A.J., Merifield, R.S., and Lyamin, A.V.,
homogeneous profiles, the factor of safety and strength (2015) Slope stability charts for two-layered purely
reduction factor is getting increased with increase in cohesive soils based on finite-element limit analysis
layer height from 0.2D to 0.6D then it stays constant. methods, Int. J. of Geomechanics, 15(3), 06014022.
The minimum safety factor is observed for layer height Rocscience Inc.. RS2 v9. Toronto, Canada: Rocscience
h=0.2D. It is clear that the silty clay soil imparts strength Inc.; 2016.
to the two-layered slope and makes it more stable. For
Shukha, R., and Baker, R., (2007) Design implications of
layer height 0.2D and 0.4D the critical surfaces are the vertical pseudo-static coefficient in slope analysis,
shallow and parallel to slope while the surface coincides Computers and Geotechnics, 35, pp86-96.
for heights 0.6D and 0.8D which has a deeper surface
depicting a slope failure. Hence it is found that slopes Spencer, E., (1967) A method of analysis of the stability
with sandy soil have a shallow slope failure parallel to of embankments assuming parallel inter-slice forces,
Geotechnique, 17(1), pp1126.
slope surface and slopes with clay soil have a deeper
slope failure. Displacements show an increasing trend Ugai, K., and Leshchinsky, D., (1995) Three-dimensional
with increasing layer height i.e., with increase in layer limit equilibrium and finite element analysis: a
height the silty clay soil in the slope increases and with comparison of results, Soils Found, 35(4), pp17.
it the differential displacement also increases indicating Zienkiewicz,O.C., Humpheson, C., and Lewis, R.W.,
a deeper zone of failure. (1975)Associated and non associated visco-plasticity
and plasticity in soil mechanics,Geotechnique,25(4),
pp67189.