Ch11_Slope.pdf

11-1
11. THE STABILITY OF SLOPES
11.1 INTRODUCTION
The quantitative determination of the stability of slopes is necessary in a number of
engineering activities, such as:
(a) the design of earth dams and embankments,
(b) the analysis of stability of natural slopes,
(c) analysis of the stability of excavated slopes,
(d) analysis of deepseated failure of foundations and retaining walls.
Quite a number of techniques are available for these analyses and in this chapter the
more widely used techniques are discussed. Extensive reviews of stability analyses have been
provided by Chowdhury (1978) and by Schuster and Krizek (1978). In order to provide some
basic understanding of the nature of the calculations involved in slope stability analyses the case
of stability of an infinitely long slope is initially introduced.
11.2 FACTORS OF SAFETY
The factor of safety is commonly thought of as the ratio of the maximum load or stress
that a soil can sustain to the actual load or stress that is applied. Referring to Fig. 11.1 the factor
of safety F, with respect to strength, may be expressed as follows:
F =
τ ff
τ
(11.1)
where τff is the maximum shear stress that the soil can sustain at the value of normal stress of σn,
τ is the actual shear stress applied to the soil.
Equation 11.1 may be expressed in a slightly different form as follows:
τ =
c
F
+
σn tan φ
F
(11.2)
Two other factors of safety which are occasionally used are the factor of safety with
respect to cohesion, Fc, and the factor of safety with respect to friction, Fφ. The factor of safety
with respect to cohesion may be defined as the ratio between the actual cohesion and the cohesion
required for stability when the frictional component of strength is fully mobilised.

11-2
This may be expressed as follows:
τ =
c
Fc
+ σn tan φ (11.3)
The factor of safety with respect to friction, Fφ, may be defined as the ratio of the tangent
of the angle of shearing resistance of the soil to the tangent of the mobilised angle of shearing
resistance of the soil when the cohesive component of strength is fully mobilised. One way in
which this may be expressed is as follows:
τ = c +
σn tan φ
Fφ
(11.4)
A further factor of safety which is sometimes used is FH, the factor of safety with respect
to height. This is defined as the ratio between the maximum height of a slope to the actual height
of a slope and may be expressed as follows:
FH =
Hmax
H (11.5)
The factors of safety Fc, Fφ, FH are only occasionally used in slope stability analyses.
The factor of safety with respect to strength (F) as expressed in equation (11.2), is the one which
is almost universally used in calculations.
Fig. 11.1 Definition Diagram for Factor of Safety

11-3
Fig. 11.2 Culmann Aproach to Slope Stability
11.3 CULMANN METHOD
A technique for the calculation of slope stability based upon the assumption of a plane
surface of failure through the toe of the slope has been proposed by Culmann (see Taylor, 1948).
In Fig. 11.2 the line QS represents a plane potential failure surface. The forces acting on the
wedge QRS are indicated on the figure as the weight of the wedge W, the mobilised cohesive
force Cm and the mobilised frictional force P. φm is the mobilised angle of shearing resistance.
These three forces are placed in equilibrium to yield the following expression:
cm
ρgH
=
cos (i + φm - 2θ) - cos (i - φm)
4 cos φm sin i
(11.6)
where the symbols are indicated in Fig. 11.2. The term on the left hand side of this equation is
known as the stability number. Since QS is an arbitrarily selected trial plane inclined at an angle
θ to the horizontal, it is necessary to find the most dangerous plane along which sliding is most
likely. This is done by setting the first derivative with respect to θ of the expression above equal
to zero. This results in determination of the critical inclination θcrit given by the following
expression:
θcrit =
1
2
(i + φm)
Substitution of θcrit into equation (11.6) yields the maximum value of the stability number,

11-4
cm
ρgH
=
1 - cos (i - φm)
4 cos φm sin i
(11.7)
The factor of safety with respect to strength may be determined from equation (11.7) by a
trial and error process similar to that described in section 11.7.
This method of slope stability analysis is not widely used since it has been found that
plane surfaces of sliding are observed only with very steep slopes, and for relatively flat slopes the
surfaces of sliding are almost always curved.
EXAMPLE
Referring to Fig. 4.2
H = 16m
tan i = 2/3
tan θ = 1/3
c = 10kPa
φ = 35˚
and the weight of the soil wedge QRS is 3.5MN/m. Calculate the factor of safety (F) against
sliding along the potential failure surface QS.
For this problem, equation (11.7) is not applicable since the angle (θ) has been specified
and this may not necessarily be equal to the critical value (θcrit). Equation (11.6) can be used and
substitution into this equation of the given information yields the following expression
cm
291.67
=
cos (φm - 3.18˚) - cos (33.69˚ - φm)
2.219 cos φm
(11.8)
If equation (11.2) is rewritten as
τ = cm + σn tan φm
it is seen that the factor of safety (F) may be expressed as
F = c/cm = tan φ/tan φm (11.9)
By using equations (11.7) and (11.9) and using successive approximation the values of φm and cm
may be determined.
φm = 15.30˚ and cm = 3.91 kPa
which leads to a factor of safety (F) of 2.56.

11-5
An alternative and possibly simpler technique that may be used for this problem is to
express the factor of safety (F) in terms of forces instead of stresses.
F =
maximum forces tending to resist sliding down the plane QS
forces tending to cause sliding down the plane QS
=
C + N tan φ
T
(11.10)
where C = cohesive force acting on plane QS
= c x length QS x 1
N = resolved part of W acting normal to plane QS
= W cos θ
T = resolved part of W acting down the plane QS
= W sin θ
∴ F =
10 x 50.6 x 1 + 3500 x 0.949 x 0.700
3500 x 0.316
= 2.56
11.4 THE φ
φ
φ
φ = 0 METHOD OF SLOPE STABILITY ANALYSIS
Since the surfaces of sliding for many slope failures have been observed to follow
approximately the arc of a circle, most of the commonly used analytical techniques for calculation
of slope stability involve the assumption of a circular failure arc. Most of the techniques
discussed in this chapter are based upon this assumption. For composite failure surfaces, analyses
have been developed by Morgenstern and Price (1965) and by Janbu (1973).
The problem is illustrated in the upper part of Fig. 11.3. The forces acting on the sliding
wedge of soil are the weight W, normal stresses which act around the failure surface and resisting
shear stresses τ which also act around the failure surface. The factor of safety F may be defined
for this situation as follows:

11-6
F =
sum of moments of maximum resisting forces
sum of moments of moving forces
=
τmax x arc length x R
Wd
(11.11)
In this equation the moments have been taken about the centre of the circle, part of which forms
the failure surface so that the normal stresses do not enter into the calculation.
It will be noted that in equation (11.11) the maximum shear stress (τmax) has been
assumed to be a constant. If this shear stress varies with the position along the sliding surface, it
would be necessary to integrate the shear stress around the arc for use in equation (11.11).
In the special case where the slope is formed of a saturated clay the angle of shearing
resistance (φu) will be zero for the short term case. The maximum resisting shear stress around
the failure arc will then be equal to the undrained cohesion (cu). If the undrained cohesion is a
constant around the failure surface then equation (11.11) may be rewritten as follows:
F =
cu R x arc length
Wd (4.12)
This total stress analysis is commonly referred to as the φu = 0 method. This method has been
widely and successfully used in practice for the evaluation of the short term stability of saturated
clay slopes. For example, Ireland (1954) has demonstrated the validity of this technique in the
analysis of the short term stability of a slope excavated in saturated soil.
For the case where the angle of shearing resistance is not equal to zero the situation is not
as straightforward as described above because of the necessity to determine the frictional
component of the resisting shear stresses. For this case the forces acting on the block are shown
in the lower part of Fig. 11.3. These forces (derived from effective stresses) are:
(a) the mobilised cohesion force (C'm), for which the line of action is known and the
magnitude of which can be expressed in terms of the effective cohesion (c') and the
factor of safety (F),
(b) the effective normal force (N'), the magnitude and line of action of which are unknown
since it depends upon the distribution of the normal effective stress around the circular
arc,

11-7
Fig. 11.3 Forces Involved in Calculations for Stability of Slopes
Fig. 11.4

11-8
(c) Rφ which is the frictional force acting around the arc. This force is normal to the force
N', has a magnitude equal to N' tan φ'/F where φ' is the effective angle of shearing
resistance, but the line of action of this force Rφ is unknown. U, the pore pressure force,
is known from seepage or other considerations.
This means that there are four unknowns altogether, the factor of safety F, the magnitude
of N', the direction of N', and rφ to locate the line of action of the frictional force Rφ. Since there
are only three equations of static equilibrium, this problem is indeterminate to the first degree. In
order to solve the problem some assumption has to be made to remove one of the unknowns. One
commonly made assumption involves the distribution of the normal effective stress around the
failure arc. This will enable the direction of the effective normal force N' to be evaluated.
EXAMPLE
Evaluate the short term stability for the dam shown in Fig. 11.4. The embankment
consists of a saturated soil for which the angle of shearing resistance φu = 0 and the undrained
cohesion cu = 70kN/m2. The calculation is to be carried out for the reservoir depth of 18m and
for the case where the reservoir has been completely emptied.
In the calculations, the forces Ww and U will have moments about the centre of the circle
and therefore must be evaluated.
Evaluating the forces (per m) acting on the block:
W = 720 x 9.81 = 7060kN = 7.06MN
Ww =
1
2 x 36 x 18 x 1.0 x 9.81 = 3180kN = 3.18MN
U =
1
2 x 18 x 1.0 x 9.81 x 18 = 1590kN = 1.59MN
maximum cohesive force C = cu x arc length
= 70 x 41.2 x 1.32
= 3.8MN
When moments are taken about the centre of the circle, there will be no moments due to the
normal stresses and pore pressures acting around the arc; so these stresses can be ignored.

11-9
Factor of Safety F =
Σ moments of maximum resisting forces
Σ moments of moving forces
=
3.8 x 41.2
7.06 x 14.8 + 3.18 x 2 - 1.59 x 34
=
156.6
104.5 + 6.4 - 54.0
= 2.75
On occasions the force U is considered as a resisting force, in which case,
F =
156.6 + 54.0
104.5 + 6.4
= 1.90
This illustrates that a different answer can be obtained depending upon the precise definition of
the factor of safety. The former calculation yielding a value of F of 2.75 is the more usual one
that is performed.
When the water is removed,
F =
156.6
104.5
= 1.50
so the slope will still be stable.
11.5 ORDINARY METHOD OF SLICES
In cases where the effective angle of shearing resistance is not constant over the failure
surface, such as in zoned earth dams where the failure surface might pass through several
different materials, the friction circle method cannot be used. A 'slices' method, is more
appropriate in this situation. With a method of slices the sliding wedge PQS as shown in
Fig. 11.5 is subdivided

11-10
Fig. 11.5 Illustration of Ordinary Method of Slices
vertically into slices. The factor of safety is determined by examining the contributions to the
moving and resisting forces provided by each slice.
The forces acting on a typical slice are shown in Fig. 11.5. These forces are the weight
of the slice W, the normal and tangential forces acting on the lower boundary of the slice and the
side forces indicated by X and E which act on the sides of the slice. With the Ordinary Method of
Slices sometimes known as the Fellenius Method or the Swedish Circle Method (Fellenius
(1936), and May and Brahtz (1936)), a number of simplifying assumptions are made in order to
render the problem determinate.
Firstly it is assumed that the side forces X and E may be neglected and secondly, that the normal
force N, may be determined simply by resolving the weight W of the slice in a direction normal to
the arc, at the mid point of the slice, as shown in the lower part of Fig. 11.5.
N = W cos α
where α is the angle of inclination of the potential failure arc to the horizontal at the mid point of
the slice

11-11
Effective normal force N' = N - U
= W cos α - u ∆X sec α
total maximum resisting force Tmax = Σ(c' + σ' tan φ')∆X sec α
= Σ (c' ∆X sec α + N' tan φ')
= Σ (c' ∆X sec α + tan φ' (W cos α - u∆X sec α))
Factor of Safety =
sum of moments of maximum resisting forces
sum of moments of moving forces
=
Σ Tmax R
Σ Wd
=
Σ Tmax R
Σ W R sin α
=
Σ Tmax
Σ W sin α
=
Σ maximum resisting forces around the arc
Σ moving forces around the arc
F =
Σ (c' ∆X sec α + tan φ' (W cos α - u ∆X sec α))
Σ W sin α
(11.13)
This procedure would then be followed for a number of trial failure surfaces until the
lowest factor of safety is found.
Some difficulties may be experienced with equation (11.13). Negative values of the
effective normal force N' may be encountered for large values of the angle α when pore pressures
are present. This method is widely used by dam constructing authorities even though it has been
demonstrated by Whitman and Moore (1963) and this method of analysis is unsound and yields
factors of safety which are smaller than the correct values.

11-12
EXAMPLE
Using the Ordinary Method of Slices, determine the factor of safety for the slope
undergoing seepage and for the failure surface shown in Fig. 11.6. The soil properties are as
follows:
total density = 2Mg/m3
effective cohesion c' = 30kN/m2
effective friction angle φ' = 30˚
Fig. 11.6

11-13
Fig. 11.7 Stresses and Forces Acting on a Typical Slice
The sliding wedge has been subdivided into six slices as shown in Fig. 11.7. The
weights of the slices have been determined and the average pore pressures acting on the bases of
the slices have been determined from the flownet which is drawn in Fig. 11.6. The effective
normal force N' may be determined either graphically as shown in Fig. 11.5 or mathematically as
shown in Table 11.1. The remaining calculations for this problem are set out in Table 11.1.
11.6 BISHOP METHOD OF SLICES
A slices method of slope stability analysis which involves a different procedure and
gives different answers compared with the Ordinary Method of Slices has been proposed by
Bishop (1955). With this method, the analysis is carried out in terms of stresses instead of forces
which were used with the Ordinary Method of Slices. The stresses and forces which act on a
typical slice and which are taken into account in the analysis are shown in Fig. 11.8. The major
difference between the Bishop Method and the Ordinary Method of Slices is that resolution of
forces takes place

11-14
Fig. 11.8 Stresses and Forces Acting on a Typical Slice
TABLE 11.1
CALCULATIONS FOR ORDINARY METHOD OF SLICES
Slice Slice
Width
(∆x)
m
Sin α
α
α
α Weight
of Slice
(W)
kN
Pore
Pressure
Force (U)
kN
W sin α
α
α
α
kN
W cos α
α
α
α
(N)
kN
N - U
(N')
kN
N' tan φ
φ
φ
φ'
kN
c' ∆X
sec α
α
α
α
kN
1
2
3
4
5
6
8
8
8
8
8
6
- .111
.049
.242
.436
.630
.775
450
1118
1590
1742
1590
570
0
150
370
450
340
20
- 50
54
384
760
1000
442
447
1117
1542
1568
1235
360
447
967
1172
1118
895
340
258
558
676
645
516
196
243
243
246
270
318
315
Totals 2590 2849 1635

11-15
F =
Σ (c' ∆X sec α + N' tan φ')
S W sin α
=
1635 + 2849
2590
= 1.73
in the vertical direction instead of a direction normal to the arc (a direction which is different for
each slice). This means that with the Bishop Method the side forces E acting on the sides of the
slices will not enter into the analysis. In the simplified Bishop Method which is described here, it
is assumed that the shear side forces X may be neglected without introducing serious error into
the analysis. A more rigorous method in which the side forces X are taken into account is found
to yield answers only slightly different from that obtained from the simplified Bishop Method.
The simplified analysis is as follows:
τ =
1
F
(c' + σ' tan φ')
To find σ' resolve forces in the vertical direction to obtain
W -
1
F (c' + σ' tan φ') ∆X tan α - (σ' + u)∆X = 0
∴ σ' =
W - u ∆X -
1
F c' ∆X tan α
∆X (1 +(tan ø' tan α)/F)
Now F =
Σ maximum resisting forces around arc
Σ moving forces around arc
=
Σ (c' + σ' tan φ') ∆X sec α
Σ W sin α
=
[ ]
∑
∑






∆
−
+
∆
α
φ
α
sin
1
tan
)
( '
'
W
M
X
u
W
X
c
(11.14)

11-16
where Mα = cos α +
sin α tan φ'
F
(4.15)
The factor of safety F appears on both sides of equation (11.14). Fortunately the solution
converges rapidly, only two or three trials for F being necessary in solving the equation. A plot of
Mα as given by equation (11.15) is presented in Fig. 11.9 to assist in the solution of equation
(11.14).
Fig. 11.9 Graph for Evaluating Mα
α
α
α
To facilitate the analyses of slope stability for a large number of potential failure surfaces
and for a variety of conditions, use is made of computer programs.
The Bishop Method yields factors of safety which are higher than those obtained with the
Ordinary Method of Slices. Further, the two methods do not lead to the same critical circle. It
has also been found that the disagreement increases as the central angle of the critical circle
increases. Analyses by more refined methods involving consideration of the forces acting on the
sides of slices show that the simplified Bishop Method yields answers for factors of safety which
are very close to the correct answer.

11-17
EXAMPLE
Using the simplified Bishop Method, determine the factor of safety for the problem
illustrated in Figs. 11.6 and 11.7. This is the same problem that has been solved in this chapter by
means of friction circle method and by means of the Ordinary Method of Slices.
The sliding wedge has been subdivided into the same six slices that were used for the solution by
means of the Ordinary Method of Slices. The evaluation of the factor of safety by means of the
Bishop Method is carried out in tabular form as shown in Table 11.2.
TABLE 11.2
CALCULATIONS FOR BISHOP METHOD OF SLICES
F(1) =
4852
2590
= 1.88
1 2 3 4 5 6
Slice
Slice
Width
Weight
of Slice
(W)
Pore
Pressure
u
W sin α
α
α
α c ∆X +
(W - u
∆X)
Mα
α
α
α col. (5)
col. (6)
- kN
(∆X)
m
kN kN/m2 kN tan φ
φ
φ
φ') kN trial (1)
F = 2.0
trial (2)
F = 1.85
trial (1)
F = 2.0
trial (2)
F = 1.85
1
2
3
4
5
6
8
8
8
8
8
6
450
1118
1590
1742
1590
570
0
18.7
45.1
50.0
32.4
2.0
-50
54
384
760
1000
442
500
799
950
1015
1010
502
.96
1.01
1.03
1.02
.97
.86
.96
1.01
1.03
1.03
.98
.87
520
790
920
995
1042
585
520
790
920
985
1030
576
Totals 2590 4852 4821

11-18
F(2) =
4821
2590 = 1.86
∴ Factor of Safety = 1.86
Fig. 11.10 Variation of Safety Factor with Time for Soil Beneath a Fill
(After Bishop and Bjerrum, 1960)
11.7 SHORT TERM AND LONG TERM STABILITY
In carrying our slope stability analyses for design purposes it is wise to check both short
term and long term conditions. For the short term conditions an effective stress analysis could be
used, but this will require an estimate of the pore pressures that will be developed. Alternatively a
total stress analysis could be used, but this would only be applicable in cases where the pore
pressure changes are entirely dependent upon stress changes. For long term conditions an
effective stress analysis is normally carried out, since the pore pressures are usually independent
of stress changes. For this analysis estimates of the pore pressures, for example, by means of
flownets, are required.
In examining the stability of a foundation soil following embankment construction, the
short term case is often more critical that the long term case. As discussed by Bishop and
Bjerrum (1960), this is illustrated in Fig. 11.10 with an examination of the stress changes at a
typical point P beneath the embankment.

11-19
In this case the pore pressure at point P at the end of construction is determined largely
by the stress changes produced by the embankment. The pore pressure for the long term case, on
the other hand, is determined by the ground water conditions. In this idealized example the factor
of safety is considered as the ratio between the soil strength and the applied shear stress. It is seen
that in this case the minimum factor of safety is obtained at the end of construction, that is, for
short term conditions. As time elapses and the construction pore pressure dissipate the factor of
safety increases as illustrated in the sketch.
On the other hand Fig. 11.11 illustrates the stress changes at a typical point P beneath an
excavated slope. Here the pore pressures at the end of excavation are determined by the stress
changes produced by the process of excavation. The long term pore pressures, as in the previous
example are determined by the ground water conditions. In this case, it is seen that the soil
strength decreases with time and the factor of safety also decreases with time, which makes the
long term stability condition the critical one to be examined.
Fig. 11.11 Variation of Safety Factor with Time for Excavation of a Slope
(after Bishop & Bjerrum, 1960)

11-20
REFERENCES
Bishop, A.W., (1955), “The Use of the Slip Circle in the Stability Analysis of Slopes”,
Geotechnique, Vol. 5, pp 7 - 17.
Bishop, A.W. and Bjerrum, L., (1960), “The Relevance of the Triaxial Test to the Solution of
Stability Problems”, ASCE Conf. on Strength of Cohesive soils, pp 437 - 501.
Bishop, A.W. and Morgenstern, N., (1960), “Stability Coefficients for Earth Slopes”,
Geotechnique, Vol. 19, No. 4, pp 129 - 150.
Chowdhury, R.N., (1978), “Slope Analysis”, Elsevier Scientific Pub. Co., Amsterdam, 423 p.
Fellenius, W., (1936), “Calculation of the Stability of Earth Dams”, Trans. 2nd Cong. on Large
Dams, Vol 4, p 445.
Ireland, H.O., (1954), “Stability Analysis of the Congress Street Open Cut”, Geotechnique, Vol.
4, p 163.
Janbu, N., (1973), “Slope Stability Computations”, Embankment Dam Engineering, Ed. by
Hirshfeld and Poulos, John Wiley & Sons, New York.
May, D.R. and Brahtz, H.A., (1936), “Proposed Methods of Calculating the Stability of Earth
Dams”, Trans. 2nd Cong. on Large Dams, Vol. 4, p 539.
Moore, P.J., (1970) “The Factor of Safety against Undrained Failure of a Slope”, Soils and
Foundations, Japanese Society of Soil Mechanics and Foundation Engineering, vol. 10, No. 3, pp
81-91.
Morgenstern, N. and Price, V.E., (1965), “The Analysis of the Stability of General Slip Surfaces”,
Geotechnique, Vol. 15, No. 1, pp 79 - 93.
Schuster, R.L. and Krizek, R.J., (Eds.), (1978), “Landslides - Analysis and Control”,
Transportation Research Board, Special Report 176, Washington D.C.
Taylor, D.W., (1948), “Fundamentals of Soil Mechanics”, John Wiley & Sons, 700 p.
Whitman, R.V. and Moore, P.J., (1963), “Thoughts Concerning the Mechanics of Slope Stability
Analysis”, Proc. 2nd Pan-Am Conf. Soil Mech. & Found. Eng., Brazil, Vol. 1, pp 391 - 411.

Ch11_Slope.pdf

More Related Content

Similar to Ch11_Slope.pdf (20)

Recently uploaded (20)

Ch11_Slope.pdf