Chapter Five Slope Stability

CHAPTER FIVE
SLOPE STABILITY
5.0 Introduction
The term slope as used in here refers to any natural or manmade earth mass, whose surface
forms an angle with the horizontal. Hills and mountains, river banks, etc. are common
examples of natural slopes. Examples of manmade slopes include fills, such as
embankments, earth dams, levees; or cuts, such as highway and railway cuts, canal banks,
foundations excavations and trenches. Natural forces (wind, rain, earthquake, etc.) change
the natural topography often creating unstable slopes. Failure of natural slopes (landslides)
and manmade slopes have resulted in much death and destruction.
In assessing the stability of slopes, geotechnical engineers have to pay particular attention
to geology, drainage, groundwater, and the shear strength of the soils. The most common
slope stability analysis methods are based on simplifying assumptions and the design of a
stable slope relies heavily on experience and careful site investigation. In this chapter, we
will examine the stability of earth slopes in two dimensional space using limit equilibrium
methods.
When you complete this chapter, you should be able to:
 Understand the causes and types of slope failure.
 Estimate the stability of slopes using limit equilibrium methods.

Sample Practical Situation: A reservoir is required to store water for domestic use. Several
sites were investigated and the top choice is a site consisting of clay soils (clay is preferred
because of its low permeability – it is practically impervious). The soils would be
excavated, forming sloping sides. You are required to determine the maximum safe slope
of the reservoir.

5.1 Definitions of Key Terms

Slip plane or failure plane or slip surface or failure surface is the surface of sliding.
Sliding mass is the mass of soil within the slip plane and the ground surface.
Slope angle (or simply slope) is the angle of inclination of a slope to the horizontal. The
slope angle is usually referred to as a ratio, for example, 2:1 (horizontal: vertical)

5.2 Some Types of Slope Failure

Slope failures depend on the soil type, soil stratification, groundwater, seepage, and the
slope geometry. A few types of slope failure are shown in Figure 5.1. Failure of a slope
along a weak zone of soil is called a translational slide (Fig. 5.1 a). Translational slides are
common in coarse-grained soils.

HU/HiT/Civil Engineering Department Lecture Note Page 1

Chapter Five Slope Stability

Figure 5.1: Some types of slope failure (Budhu, pp. 524)

A common type of failure in homogeneous fine-grained soils is a rotational slide. Three

types of rotational slides often occur. One type, called a base slide, occurs by an arc
enclosing the whole slope. A soft soil layer resting on a stiff layer of soil is prone to base
failure (Fig. 5.1 b). The second type of rotational failure is the toe slide, whereby the
failure surface passes through the toe of the slope (Fig. 5.1 c). The third type of rotational
failure is the slope slide, whereby the failure surface passes through the slope (Fig. 5.1 d).
A flow slide occurs when internal and external conditions force a soil to behave like a
viscous fluid and flow down even shallow slopes, spreading out in several directions (Fig.
5.1 e).

5.3 Some Causes of Slope Failure

Slope failures are caused in general by natural forces, human mismanagement and
activities. Some of the main factors that provoke failure are summarised in Figure 5.2
below.

HU/HiT/Civil Engineering Department Lecture Note Page 2

Chapter Five Slope Stability

Figure 5.2: Some causes of slope failure (Budhu, pp. 526)

As shown in Fig. 5.2, some of the most common causes of slope failures are erosion,
rainfall, earthquake, geological features, external loading, construction activities (ex.
excavation & fill), and reservoir rapid drawdown.

5.4 Two-Dimensional Slope Stability Analysis

Slope stability can be analyzed using one or more of the following: the limit equilibrium
method, limit analysis, finite difference method, and finite element method. Limit
equilibrium is the most widely used method for stability analysis. In the following sections,
we will learn some of the commonly used slope stability analysis methods that are based
on the limit equilibrium.

HU/HiT/Civil Engineering Department Lecture Note Page 3

Chapter Five Slope Stability

5.5 Stability Analysis of Infinite Slopes

Infinite slopes have dimensions that extend over great distances. In practice, the infinite
slope mechanism is applied to the case when a soft material of very long length with
constant slope may slide on a hard material (e.g. rock) having the same slope. Let’s
consider a clean, homogeneous soil of infinite slope  s as shown in Figure 5.3. To use
limit equilibrium method, we must first speculate on a failure of slip mechanism. We will
assume the slip would occur on a plane parallel to the slope. If we consider a slice of soil
between the surface of the soil and the slip plane, we can draw a free-body diagram of the
slice as shown in Figure 5.3.

Figure 5.3: Forces on a slice of soil in an infinite slope.

The forces acting on the slice per unit thickness are the weight, W = bz , the shear forces
X j and X j +1 on the sides, the normal forces E j and E j +1 on the sides, the normal
force N on the slip plane and the mobilized shear resistance of the soil, T , on the slip
plane. We will assume that forces that provoke failure are positive. If seepage is present, a
seepage force J s = i wbz develops, where, i is the hydraulic gradient. For a uniform
slope of infinite extent, X j = X j +1 and E j = E j +1 . To continue with the limit equilibrium
method, we must now use the equilibrium equations to solve the problem. But before that
we will define the factor of safety (FS) of a slope in the following subsection. The general
objective of infinite slope stability analysis is to determine either the critical slope or
critical height, or alternatively, the factor of safety of the slope.

5.5.1 Factor of Safety

The factor of safety of a slope is defined as the ratio of the available shear strength,  f , to
the minimum shear strength required to maintain stability (which is equal to the mobilized
shear stress on the failure surface),  m , that is:
f
FS = (5.1)
m

HU/HiT/Civil Engineering Department Lecture Note Page 4

Chapter Five Slope Stability

The shear strength of the soil is governed by the Mohr-Coulomb failure criterion.

5.5.2 Stability of Infinite Slopes in u =0, cu soil.

For the u =0, cu soil, the Mohr-Coulomb shear strength is given by:
 f = cu (5.2)
From statics and using Figure 4.3,
N = W cos s And T = W sin s (5.3)

The shear stress per unit length on the slip plane is given by:
T W sin  s cos s bz
m = = = sin  s cos s = z sin  s cos s (5.4)
l b b
The factor of safety is then,
cu 2c u
FS = = (5.5)
z sin  s cos s z sin(2 s )
At limit equilibrium, FS = 1. Therefore, the critical slope is

 c = 12 sin −1 (2cu z ) (5.6)

And the critical depth is:

2cu
zc = (5.7)
 sin(2 s )

5.5.3 Stability of Infinite Slopes in c’,  ' soil – with no seepage.

For a c’,  ' soil, the Mohr-Coulomb shear strength is given by:

 f = c' +  n' tan  ' (5.8)

The factor of safety FS is then:

c' +  n' tan  ' c'  n' tan  '
FS = = + (5.9)
m m m
The normal and shear stresses per unit length at the failure plane in reference to figure 5.3
are given by:
N T
 'n = And  m = (5.10)
l l

For a slope without seepage, Js=0. From Eqns. (5.4, 5.9 and 5.10) we get:
c' W ' cos s tan  ' c' tan  '
FS = + = + (5.11)
z sin  s cos s W sin  s z sin  s cos s tan  s
At limit equilibrium FS = 1. Therefore, the critical depth zc is given by

HU/HiT/Civil Engineering Department Lecture Note Page 5

Chapter Five Slope Stability

c'  sec2  s 
zc =   (5.12)
  tan  s − tan  ' 

For the case where  s   ' , the factor of safety is always greater than 1 and is computed
from Eqn. (5.11). This means that there is no limiting value for the depth z, and at an
infinite depth, the factor of safety approaches to tan  ' / tan  s . For a coarse-grained soil
with c’ = 0, Eqn. (5.11) becomes:
tan  '
FS = (5.13)
tan  s
At limit equilibrium FS = 1. Therefore, the critical slope angle is:
c = ' (5.14)
The implication of Eqn. (5.13) is that the maximum slope angle of a coarse-grained soil
with c’ = 0, cannot exceed  ' . In other words, the case c’ = 0 and  s   ' is always
unstable and cannot be applied to practical situations.

Example 5.1
An infinitely long slope is resting on a rock formation with the same inclination. The
height of the slope is 3.2 m. Determine a) the factor of safety, b) the shear stress
developed on the sliding surface, and c) the critical height.  s =250,  = 17.5 kN/m3, c’ =
12 kPa and  ' = 200.

5.5.4 Stability of Infinite Slopes in c’,  ' soil – steady state seepage.
We will now consider groundwater at the ground surface and assume that seepage is
parallel to the slope. The seepage force is:
J s = i wbz
Since seepage is parallel to the slope, i = sin . From statics,
N ' = W ' cos s =  ' bz cos s (5.15)
And
T = W ' sin  s + J s
=  ' bz sin  s +  w bz sin  s = ( '+ w )bz sin  s (5.16)
=  sat bz sin  s

Therefore, the shear stress at the slip plane is:

T  sat bz sin  s cos s
m = = =  sat z sin  s cos s
l b
From the definition of factor of safety (Eqn. 5.1), we get:

HU/HiT/Civil Engineering Department Lecture Note Page 6

Chapter Five Slope Stability

c'  ' bz cos s tan  '

FS = +
 sat z sin  s cos s  sat zb cos s tan  s
(5.17)
c'  ' tan  '
= + 
 sat z sin  s cos s  sat tan  s

At limit equilibrium, FS=1. Therefore, the critical height is:

c' csc2  s
zc =
 tan  s −  ' tan  '
At infinite depth the factor of safety in Eqn. (4.17) becomes:
 ' tan  '
FS =  (5.19)
 sat tan  s
Eqn. (5.19) can also be used for calculating the factor of safety for a coarse-grained soil
with c’ = 0. At limit equilibrium FS = 1, and hence, the critical slope for a coarse-grained
soil with c’ = 0 is given by:
'
tan  s = tan  ' (5.18)
 sat
For most soils,  '  sa t  12 . Thus, seepage parallel to the slope reduces the limiting slope
of a clean, coarse-grained soil by about one-half.
If the groundwater level is not at the ground surface, weighted average unit weights have to
be used in Eqns. (5.17 and 5.18).

Example 5.2
A long slope of 4.5 m deep is to be constructed of material having the following properties:
 sat =20 kN/m3,  dry =17.5 kN/m3, c’=10 kPa, and  ' =320. Determine the factor of safety
a) when the slope is dry, b) there is a steady state seepage parallel to the surface with the
water level 2 m above the base and c) the water level is at the ground surface.

5.6 Rotational Slope Failure

The infinite slope failure mechanism is reasonable for infinitely long and homogeneous
slopes made of coarse-grained soils, where the failure plane is assumed to be parallel to the
ground surface. But in many practical problems slopes have been observed to fail through
a rotational mechanism of finite extent. As shown in Fig. (5.1), rotational failure
mechanism involves the failure of a soil mass on a circular or non-circular failure surface.
In the following sections, we will continue to use the limit equilibrium method assuming a
circular slip surface. Those methods, which are based on non-circular slip surface, are
beyond the scope of this course.

HU/HiT/Civil Engineering Department Lecture Note Page 7

Chapter Five Slope Stability

5.6.1 Stability of Slopes in cu, u =0 soil – circular failure surface.

The simplest circular analysis is based on the assumption that a rigid, cylindrical block will
fail by rotation about its center and that the shear strength along the failure surface is
defined by the undrained strength cu. Figure 4.4 shows a slope of height H and angle  s .
The trial circular failure surface is defined by its center C, radius R and central angle  .

Figure 5.4: Slope failure in cu, u =0.

The weight of the sliding block acts at a distance d from the center. Taking moments of the
forces about the center of the circular arc, we have:
cu LR cu R 2 0 
FS = =  (5.19)
Wd Wd 1800
Where L is the length of the circular arc, W is the weight of the sliding mass and d is the
horizontal distance between the circle center, C, and the centroid of the sliding mass. If cu
varies along the failure surface then:
R 2 (cu11 + cu 2 2 + ... + cun n ) 
0 0 0
FS =  (5.20)
Wd 1800
The centroid of the sliding mass is obtained using a mathematical procedure based on the
geometry or the sub-division of the sliding mass into narrow vertical slices.

Example 5.3
Find the factor of safety of a 1V:1.5H slope that is 6 m high. The center of the trial mass is
located 2.5 m to the right and 9.15 m above the toe of the slope. cu = 25 kPa, and  =18
kN/m3. Take d = 3.85 m.

HU/HiT/Civil Engineering Department Lecture Note Page 8

Chapter Five Slope Stability

5.6.2 Culmann Method

This method of analysis is based on the assumption that failure occurs on a slip plane
through or above the toe of the slope. Fig 5.5 shows the forces acting on the sliding mass.

(a) (b)Forces acting on the sliding mass (c) Force Polygon

Fi g 5.5: Culmann’s Method

A value of FS ranging from 1.25 to 1.5 is considered to be satisfactory for an earth slope.
For economic reasons, a value greater than 1.5 is not desired. Hence, FS=1.5 may be
considered to be necessary as well as sufficient. Referring to the force polygon shown in
Fig 5.5(c)

Where, the quantity c/γH is known as the stability number. The condition for impending
slip occurs when c/γH is a maximum. Thus, from eqn. above.

HU/HiT/Civil Engineering Department Lecture Note Page 9

Chapter Five Slope Stability

The method gives good results for very steep or vertical slopes; it does not provide
satisfactory results for relatively flat slopes.

Example 5.4:
Find the factor of safety of the slope if γ=17KN/m3, c=26KPa, φ=180, β=480 and H=15m.

5.6.3 Effect of Tension Cracks

Tension cracks may develop from the upper ground surface to a depth z 0 that can be
estimated using chapter 2 Eqn. (4.13). The effect of the tension crack can be taken into
account by assuming that the trial failure surface terminates at the depth z0, thereby
reducing the weight W and central angle  . Any external water pressure in the crack creates
a horizontal force that must be included in equilibrium considerations.

Example 5.5
Rework Example 5.3 by taking into account tension cracks. Geometric data are:  =66.60,
area of sliding mass = 27.46 m2, and d = 3.48 m.

5.6.4 Stability of Slopes in c’,  ' soil – Method of Slices.

The stability of a slope in a c’,  soil is usually analyzed by discretizing the mass of the
failure slope into smaller slices and treating each individual slice as a unique sliding block
(Fig. 5.5). This technique is called the method of slices.

Figure 5.6: Slice discretization and slice forces in a sliding mass.

In the method of slices, the soil mass above a trial failure circle is divided into a series of
vertical slices of width b as shown in Fig. 5.6 (a). For each slice, its base is assumed to be a
straight line defined by its angle of inclination  with the horizontal whilst its height h is
measured along the centerline of the slice.

HU/HiT/Civil Engineering Department Lecture Note Page 10

Chapter Five Slope Stability

Figure 5.7: a) Method of slices in c’,  ' soil, b) Forces acting on a slice.
The forces acting on a slice shown in Fig. 5.6 (b) are:
W = total weight of the slice =  ×h×b
N = total normal force at the base = N’ + U, where N’ is the effective total normal
force and U = ul is the force due to the pore water pressure at the midpoint
of the base length l.
T = the mobilized shear force at the base =  m  l , where  m is the minimum
shear stress required to maintain equilibrium and is equal to the shear
strength divided by the factor of safety:  m =  f FS .
X1, X2 = shear forces on sides of the slice and E1, E2 = normal forces on sides the
slice. The sum of the moments of the interslice or side forces about the
centre C is zero.
Thus, for moment equilibrium about the center C (note the normal forces pass through the
centre):
i =n i =n i = n ( l ) i =n

 Ti R = R ( m l ) = R =  (W sin ) i R
f i
(5.20)
i =1 i =1 i =1 FS i =1

Where n is the total number of slices. Replacing  f by the Mohr-Coulomb shear strength,
we obtain:

 (c'+ 
i =n i =n
'
n tan  ' )l i  (c' l + N ' tan  ' )i
FS = i =1
i =n
= i =1
i =n
(5.21)
 (W sin )
i =1
i  (W sin )
i =1
i

The term c’l may be replaced by c'b / cos . For uniform c’, the algebraic summation of c’l
is replaced by c’L, where L is the length of the circular arc. The values of N’ must be
determined from the force equilibrium equations. However, this problem is statically
indeterminate – because we have six unknown variables for each slice but only three
equilibrium equations. Therefore some simplifying assumptions have to be made. In this

HU/HiT/Civil Engineering Department Lecture Note Page 11

Chapter Five Slope Stability

chapter two common methods that apply different simplifying methods will be discussed.
These methods are called the Fellenius method and Bishop simplified method.

5.6.4.1 Fellenius or Ordinary or Swedish Method

The ordinary or Swedish method of slices was introduced by Fellenius (1936). This
method assumes that for each slice, the interslice forces X1=X2 and E1=E2. Based on this
assumption and from statics, the forces normal to each slice are given by:
N = W cos = N ' + ul  N ' = W cos − ul (5.22)
Substituting N’ into Eqn. 5.21, we obtain:
i =n

 (c' l + (W cos − ul) tan  ' ) i

FS = i =1
i =n
(5.23)
 (W sin )
i =1
i

For convenience, the force due to pore water is expressed as a function of W:

ub
ru = i i (5.24)
Wi
Where ru is called the pore water pressure ratio. Consequently, we have:
i =n

 (c' l + W (cos − r u sec ) tan  ' )i

FS = i =1
i =n
(5.25)
 (W sin )
i =1
i

The term ru is dimensionless because the term ub =  w  hw  b  1 represents the weight

of water with a volume of hw  b 1 . Furthermore, ru can be simplified as follows:
ub  w hw b  w hw
ru = = = (5.26)
W hb h
In the case of the steady state seepage the height of water above the midpoint of the base is
obtained by constructing the flow net. Alternatively, an average value of ru may be
assumed for the slope. By doing so it is assumed that the height of water above the base of
each slice is a constant fraction of the height of each slice. If the height of the water and the
average height of the slice are equal, the maximum value of ru becomes  w  , which for
most soils, is approximately 0.5. Note that the effective normal force N’ acting on the base
is equal to N ' = W cos − ul or N ' = W (cos − ru sec ) . If the term (cos − ru sec ) is
negative, N’ is set to zero because effective stress cannot be less than zero (i.e. soil has no
tension strength).
The whole procedure explained above must be repeated for a number of trial circles until
the minimum factor of safety corresponding to the critical circle is determined. The
accuracy of the predictions depends on the number of slices, position of the critical surface,
and the magnitude of ru. There are several techniques that are used to reduce the number of

HU/HiT/Civil Engineering Department Lecture Note Page 12

Chapter Five Slope Stability

trial slip surfaces. One simple technique is to draw a grid and selectively use the nodal
points as centers of rotation.

Example 5.6
Using Fellenius’ method of slices, determine the factor of safety for the slope of example
5.3 for ru = 0 and 0.4. Take the number of slices as 8, each having 1.5 m width (check the
width of the last slice). Soil properties are c’ = 10 kPa,  ' =290, and  =18 kN/m3.

5.6.4.2 Bishop Simplified Method

This method assumes that for each slice X1=X2 but E1  E2. These assumptions are
considered to make this method more accurate than the Swedish method. An increase of
5% to 20% in the factor of safety over the Swedish method is usually obtained. Referring
to Figure 5.6 b, and writing the force equilibrium in vertical direction (in order to eliminate
E1 and E2), the following equation for N’ can be found:
c' l sin 
W − ul cos −
N' = FS (5.27)
sin  tan  '
cos +
FS
In addition to the force in the vertical direction, Bishop Simplified method also satisfies the
overall moment equilibrium about the center of the circle as expressed in Eqn. (5.21).
Putting l = b / cos and ub = ruW , and substituting Eqn. (5.27) into Eqn. (5.21), we
obtain:
1 i =n
 c' b + W (1 − ru ) tan  ' 
FS = i =n 
i =1 
 (5.28)
i
 (W sin )
m
i
i =1

Where,
sin  tan  '
m = cos + (5.29)
FS
Equation (5.29) is non-linear in FS (that is FS appears on both sides of the equations) and
is solved by iteration. An initial value of FS is guessed (slightly greater than FS obtained
by Fellenius’ method) and substituted to Eqn. (5.29) to compute a new value for FS. This
procedure is repeated until the difference between the assumed and computed values is
negligible. Convergence is normally rapid and only a few iterations are required. The
procedure is repeated for number of trial circles to locate the critical failure surface with
the lowest factor of safety.

Example 5.7
Re-work Example 5.5 for ru = 0.4 using Bishop’s simplified Method.

HU/HiT/Civil Engineering Department Lecture Note Page 13