1

1
Slope stability analysis
2
1m
1m
Unit of stress:
1 kilopascal kPa
= 1 kN/m
2
(Kilo Newtons per square
meter)
1,000N
Slightly above-average
American male
2
3
Unit weights of materials (per m
3
)
Water,
w
= g
w
= 9.81 kN/m
3
Solid rock,
s
= 26.0 kN/m
3
Soil-made up of solid grains and pores
Dry soil, idealized:

d
=
s
V
s
/(V
s
+V
v
) =
s
(1-n) = 15-20 kN/m
3
Saturated soil:

sat =
(
s
V
s+

s
V
s
)/(V
s
+V
v
) =20-23 kN/m
3
V
s
= volume solids
V
v
= volume voids
Porosity:
n=V
v
/(V
v
+V
s
)
air
water
4

z = z
d
+z
w
z
d
z
w
L
W
N
T

Fig. 1
Slope cross-section
Fig. 2
Isolate a column
Fig. 3
Force polygon
Z
w
cos
Flow lines
Equipotential lines
Fig. 4
Groundwater pressure
Infinite slope equation
3
5
F
c z z z
z z
d d sat w w w
d d sat w
=
+ +
+
' ( ) cos tan
( ) sin

w
= unit weight of water (9.81 kN/m
3
)

d
= dry unit weight of soil (15-20 kN/m
3
)

sat
= saturated unit weight of soil (20-23 kN/m
3
)
Note: the soil column is assumed
to be 1m wide perpendicular to paper
6
Assume no cohesion (c=0) and full saturation, z
d
= 0:
F
z
z
w sat w
sat w
w
sat
=

=
( ) cos tan
sin
( )
tan
tan
.
tan
tan

1 05
Assume no cohesion (c=0) and a dry slope, z
w
= 0:
F =
tan
tan

Conclusion: a dry cohesionless slope will be at the point of

failure when = ("angle of repose"). However, a saturated
slope with parallel seepage will be about half as steep.
4
7
Role of cohesion:
Assume fully saturated
slope, parallel seepage
z = z
w
= 1.5 m
= 32
F
c z z z
z z
d d sat w w w
d d sat w
=
+ +
+
' ( ) cos tan
( ) sin


Solve equation to find c when
F=1.0 (at failure)
Conclusion:
Slopes in mountainous
regions require cohesion
8
Undrained (short term) slope failure in clay
= 0
S
u
instead of c
Slope fully saturated (by capillary action) , z = z
w
F
c z z z
z z
d d sat w w w
d d sat w
=
+ +
+
' ( ) cos tan
( ) sin


sin ) ( z
S
F
sat
u
=
z
Undrained failures in clay
tend to be rotational
sand
clay
5
9
Shallow landslide susceptibility
Program LISA (US Forest Service):
apply Infinite Slope Equation on an
areal basis, in a probabilistic
manner, map probability of
failure (or Factor of Safety)
(Hammond et al., 1992)
f
r
e
q
u
e
n
c
y
10
Influence of logging on slope stability
Root
cohesion
0 5 10 15 20 25 30
YEARS AFTER LOGGING
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
R
E
L
A
T
I
V
E

R
O
O
T

R
E
I
N
F
O
R
C
E
M
E
N
T
6
11
Subjective slope stability mapping
(J. Ryder, Vancouver, 1998)
12
7
13
Deep-seated landslides: Method of Slices
General method:
1) Work out the equilibrium of each slice
2) Calculate the equilibrium of the slice assembly
3) Results depend on assumptions regarding the
interslice forces E and X
14
Available equations
X T N W + + = sin cos
F
uL N
F
L c
T
' tan
) (
'
+ =
0 cos sin =

T N
0 =
T N w
Tr Nr Wr
1) Vertical equilibrium of a single slice (n equations)
2) Mohr-Coulomb strength (n equations)
3) Horizontal force equilibrium for the slice assembly (1 equation)
4) Moment equilibrium for the slice assembly
(1 equation)
Mobilized strength
Interslice forces cancel out
in these equations
rs are radii of rotation
8
15
Bishops Simplified Method:
Fredlund Krahn (1978) Modification for Non-circular surfaces
N
r
N
Add moment of the normal forces
16
Possible solutions
For n slices, we have the following unknowns:
n N forces + n T forces + n X + n E +1F =4n+1 unknowns
Bishops simplified method:
1) assume X = 0 (no shear between slices)
2) use only Equations 1,2 and 4 (neglect horizontal force
equilibrium) E not needed, problem determinate
3) Good for circular sliding surfaces, conservative for others
4) Not good if large horizontal external forces involved
9
17
More solutions
Janbu simplified method:
1) assume X = 0 (no shear between slices)
2) use only Equations 1,2 and 3 (neglect moment equilibrium)
3) Good for shallow sliding surfaces, tends to be more conservative
than Bishop (correction needed)
Spencers method:
1) assume X/E = constant (constant interslice friction)
2) Must add another equation (horizontal force equilibrium on each
slice). Use all five equations (rigorous solution)
3) Requires iterative solution, may not converge
Morgenstern-Price method:
1) assume X/E varies according to a prescribed function
(rigorous solution)
2) Requires iterative solution, may not converge
18
Simplified and rigorous method comparison
0 20 40 60 80 100 120 140 160 180 200
DISTANCE (m)
0
20
40
60
80
100
120
E
L
E
V
A
T
I
O
N

(
m
)
WEAK
WEAK
Bishop: F = 1.00
Spencer: F = 1.22
0 20 40 60 80 100 120 140 160 180 200
DISTANCE (m)
0
20
40
60
80
100
120
E
L
E
V
A
T
I
O
N

(
m
)
WEAK
STRONG
Bishop: F = 1.00
Spencer: F = 1.31
0 20 40 60 80 100 120 140 160 180 200
DISTANCE (m)
0
20
40
60
80
100
120
E
L
E
V
A
T
I
O
N

(
m
)
WEAK
STRONG
Bishop: F = 1.00
Spencer: F = 1.03
Simplified method OK, if
slide head stronger than toe
(classic compound slide) not
good if the opposite
10
19
Sarma method
/F
W
N
T

i
/F
Force polygon
1) Divide sliding body into blocks
2) Assume that a constant friction angle,
i
/F will be mobilized on
all block interfaces
3) Solve graphically from first block to last. Will only work out for
one specific value of F (iterations)
4) Problem: can we justify the given value of internal friction?
Danger of non-conservative error! Good for structurally-
controlled slides in rock
20
Limit Equilibrium Methods, Summary 1
Vertical
or
Inclined
Yes Yes Yes Rigorous Sarma
Morgenstern-
Price
Spencer
Janbu
Bishop
Method
Rigorous
Rigorous
Simplified
Simplified
Type
Vertical Yes Yes Yes
Vertical Yes Yes Yes
Vertical No Yes Yes
Vertical Yes No Yes
Slices Moment
Equilibrium
Horizontal
Force
Equilibrium
Vertical
Force
Equilibrium
11
21
Limit Equilibrium Methods, Summary 2
-less efficient, may not converge
-the assumption of fully mobilized internal
friction could lead to incorrect (non-
conservative) results, if not justified (e.g.
in rotational slides)
- good for structured slides (esp.
rock)
Rigorous Sarma
-less efficient, may not converge
-choice of interslice function required
-any geometry and loads
-can simulate internal shearing
-often cited as a benchmark
Rigorous Morgenstern
-Price
-less efficient, may not converge
-often more conservative than MP
-any geometry and loads
Rigorous Spencer
-usually more conservative than other
methods
-requires correction factor
-very efficient
-good for shallow slides
-horizontal external loads are OK
(includes horizontal force
equilibrium)
Simplified Janbu
-conservative with cases involving internal
distortion
-can be incorrect with external horizontal
loads (including earthquake loads)
-very efficient
-accurate for circular surfaces and
some non-circular (with Fredlund-
Krahn modification)
Simplified Bishop
Disadvantages Advantages Type Method
22
Search for the critical sliding surface
Automatic
Simplex
search
12
23
Grid search
24
Compound sliding surface (3D)
Weak surface
13
25
Specified non-circular sliding surface
Toe
submergence
26
Pore pressure conditions
f f w w
h B h u + =
w w
h u =
u
r h u =
fill
r
u
= pore pressure ratio
14
27
Strength reduction method
Calculation steps (time)
D
i
s
p
l
a
c
e
m
e
n
t1
st
strength
reduction
2nd strength
reduction
failure
Strength reduction:
c
mob
=c/F
mob
= /F
Every strength reduction increases
displacements. Start with F=1, apply
successively higher reduction in cycles,
until failure occurs.
FLAC: Fast Lagrangian Analysis of Continua