Example of A Landslide Stabilization

Download as pdf or txt
Download as pdf or txt
You are on page 1of 45

Dr.

Daniel Pradel, PE GE
Principal Engineer at Group Delta Consultants,
Inc.
www.groupdelta.com
and Adjunct Associate Professor at UCLA
www.cee.ucla.edu
1

Example of a landslide stabilization

Example: Estrondo Landslide


Stabilization in Encino, CA

Slope after the repair


was completed 3

Background
Slope is a 1 :1 fill slope graded ~1960 with a
maximum height of ~110-feet that filled the end
of a canyon
The slope had failed several times before 1998
and the lower 75-feet of the slope failed during
the 1998 El Nio rains
Between 1998 and 2002 a total of 16 borings
were drilled (including small, large & slope
inclinometers)
4

Approximate limits of 1998 failure

Repair elements

2
5
1

groundwater
pressure
contours

5
3

1) Protection wall at toe of slope

2) Construction of
the mid-slope
caissons & wall
8

Slope inclinometer

3) Installation of
tiebacks (temp. in
City of LA)

10

4) Drainage
curtain / gallery
Caissons holes filled
with gravel (with
intercepting bells)

11

5) Excavation & wall construction

12

6) Second
row of
tiebacks

13

2005 rains!!!

14

Slope
inclinometer

15

Slope after the repair


was completed

16

17

1st Option:
designs
based on
Active
pressures

Example 1:
wall in Culver City, CA

18

March 24, 1995

Example 2:
wall at toe of 1995
La Conchita
landslide

19

Wall in 2004

20

10

Wall after 2005 failure

21

1998
design

22

11

2nd Option: Unbalanced Force method

FFOS=1.5

Often required by reviewing agencies in CA


23

What does a FOS=1.5 mean for design?


24

12

For exclusively soil failure mechanisms we


know very well what it means

25

Unbalanced
Force
Method

FFOS=1.5

Questions: Can we easily get FFOS=1.5?


Do all members of the design team understand how
to use it similarly?
Let us look at the most analogous case:
Sliding design for retaining walls
26

13

Traditional retaining wall analysis


for stability against sliding:
In school we are taught to calculate earth pressures

based on Rankine and Coulomb theories

Important:
in this step
soil strength
is not
factored,
i.e., FOS=1
27

Traditional retaining wall analysis


for stability against sliding:
We calculate earth

pressures based on
Rankine and Coulomb
theories
Using force equilibrium we
calculate forces acting
along base
We calculate FOS = Resisting
Driving
and verify that it
meets the design criteria
and codes (1.5 , 2.0 , ).
Otherwise

28

14

Similarly, let us explore how the force


FFOS=1.5 is used by the design team?

FFOS=1.5

29

Some design
approaches I have
seen used, in the
structural design of
shear pins:

FFOS=1.5

The Structural Engineer takes the FFOS=1.5 and

does a conventional design using this


geotechnical load (similar to the design of a
conventional retaining wall)
The FFOS=1.5 load is divided by FOS=1.5 and
this reduced load is used in a conventional
structural design
The FFOS=1.5 load is considered as having
certain structural load factors included (similar
to ACI design)

Duplication
of FOS?
Yes

?
30

15

Another question we may ask:


Is FOS = 1.5 appropriate for
landslide stabilizations?

31

Typical Geotechnical Uncertainties


Uncertainties
Shear strength
Geotechnical profile
Pore water / seepage
Densities
Weights and loads
from structure

Comparison New Slope


vs. a Landslide Repair
Much less uncertainty
Much less
Typically less
Similar?
Same?

Assumes static conditions only, access to construction records and a thorough investigation!
32

16

In summary:
The unbalanced force method is different from

conventional retaining structures design by nature of


the FOS included in the calculation of the lateral load
Application of the method can lead to duplication of
FOS, uneconomical designs, and/or very different
designs by different consultants
Experience suggests that slopes properly designed
using the unbalanced force approach generally
perform very well
IMO, clear guidelines on use of unbalanced force are
needed for Structural design
33

34

17

3.1 De Beer / NAVFAC method

35

36

18

3.2 Active/Passive wedge method


Pa
R1
R2

Practical questions:
Should it be used with residual or back-calculated strength?
Should the active load be calculated with FOS=1.0 like in
earth pressure theory or 1.5 like in slope stability?
37

3.3 The strong slice method


Strength of

slice is
increased until
desired FOS is
achieved
Pile load is
taken as the
base shear or
derived from
the inter-slice
forces
38

19

39

40

20

3.4 Base shear Force method

Resisting
Driving

OMS:

< FOSmin
41

OMS modification:

Fpile

OMS:

Note: method
requires an angle
correction for
structural design

+ Fpile

> FOSmin
42

21

FOSmin = 1.5

For FOS=1.5:
Need: 1.5 x 80977 lbs/ft
We have:
101609 lbs/ft
Extra needed: 80591 lbs/ft
43

This method has been


implemented in computer codes,
e.g., in Slide from Rockscience
and can be used with other
methods, e.g., Spencer

44

22

45

3.5 Application of a Force to


achieve Moment Equilibrium
Since a shear pin
creates a stabilizing
force, the resisting
moment created
can be easily
incorporated
mathematically into:
Method of
moments
Bishops
simplified method
46

23

Necessary equations
can be found in
numerous books and
manuals

47

Example of
implementation
in computer
programs:

y only
specified
48

24

x and y
specified
49

Question: what is the elevation of the


unbalanced force? (important for moments)

Pa

Method 2 (active wedge): Coulomb => typ. H/3


50

25

Question: what is the elevation of the


unbalanced force? (important for moments)

Method 3 (strong slice): generally at or near the base


(depends on method)
51

Thrust line from


Spencer

52

26

Question: what is the elevation of the


unbalanced force? (important for moments)

Method 4 (Force method): Unknown or near the


basal surface

53

Using Spencers Method

Thrust line

54

27

Question: what is the elevation of the


unbalanced force? (important for moments)

Method 5 (Moment method): Imposed by user


55

56

28

Goal: to find a pressure distribution


that is compatible with p-y method

57

58

29

Poulos

59

Flow mode
Short pile mode

Poulos ERCAP

Intermediate mode
60

30

Comments
Pile spacing, diameter, rigidity and length are

needed as input!
Numerous methods have been published
All methods provide structural demands
(shear force and bending moments) but a few
also incorporate/provide FOS for the
improved slope
Complex methods that often require use of
proprietary programs!
61

3.7 Numerical methods FEM and FDM:


a) Strength Reduction Method
(Griffiths and Lane, 1999)

All strengths are


divided by R , a
strength
reduction factor
R is increased in run after
run until failure occurs or
we have num. instability

f = cd + .tand

62

31

Slope
analyzed by
Strength
Reduction
Method
FOS=1.13

30 deg.
H=13.7m

Soil:
c = 23.94 kPa
= 10 deg.
= 19.63 kN/m3
E=50 Mpa
=0.25

D=3.0m

Bedrock:
c = 23.94 kPa
= 40 deg.
= 19.63 kN/m3
E=50 Mpa
=0.25

shear pin

500 psf

soil

10

bedrock =
=

K
=
G
=
E
=
=

125 pcf
3.3E+07 Pa
2.0E+07 Pa
5.0E+07 Pa
0.25

2.0E+11 Pa

Epile

40

63

60

64

32

max: Maximum Slope Displacement (m)

S/L=0.44

L=23.73m
4

S=10.45m

0
1

1.1

1.2

1.3

1.4

1.5

1.6

1.7

1.8

1.9

R: Reduction Factor
65

Slope stabilization using piles


at mid-slope (Mp=1MN.m/m)

Video 5

Shown:
- shear strains
- bending moment
- displacements

66

33

Slope stabilization after piles have


been moved downslope
(Mp=1MN.m/m)

Shown:
- shear strains
- bending moment
- displacements

Video 6

67

Mmax:Maximum Bending Moment (kN.m/m)

When M < Mp a maximum/reasonable


design load is obtained
5000
4500
4000
3500
3000

L=23.73m

S/L=0.44

2500

S=10.45m
2000
1500
1000

Elastic pile

500
0
0

0.5

1.5

max: Maximum slope displacement (m)

2
68

34

Critical failure mode is clear

Soil
Bedrock

69

70

35

71

72

36

73

Numerical Modeling
Advantages & disadvantages
No hypothesis of the critical failure mode is
needed
Great graphics => critical failure mode is obvious
Convergence problems can be problematic!
Available software for method of characteristics
provides adequacy factors and not Factors of
Safety (Eurocode 7)
Sophisticated reviewer needed
74

37

75

What is the critical


failure mechanism?

76

38

Method of characteristics

77

FLAC (SRM)

Video 7

Shown:
- shear strains
- bending moment
- tieback tension

78

39

Critical mode of failure (3 blocks)


and location of plastic hinges

79

Fish functions & design optimization


2
FLAC: Elastic Pile, = 0

1.9

FLAC: Elastic Pile, =

FOS: Factor of Safety

1.8

FLAC: Plastic Pile, = 0

1.7

FLAC: Plastic Pile, =

1.6

Bishop: Elastic Pile, = 0

Optimum for
plastic pile

1.5
1.4
1.3
L

1.2
S

1.1
1
0.9
0.8
0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

S/L: Horizontal Distance Ratio

1.1

1.2

1.3

1.4
80

40

81

L
S

800

Unbalanced Force (kN/m)

700
600
500
400
300
Moment method (XSTABL & Bishop)

200
Force method (SLIDE & Bishop+Spencer)

100

FLAC (Shear force) Structural shear force

(for comparison only)

0
0.35

0.45

0.55

0.65

0.75

s/L
82

41

Moments
Additionally in
some
analyses we
forced a
triangular
distribution

For Bending Moments


springs were
incorporated (p-y)
83

L
S

Maximum Bending Moment (kN.m/m)

4500
4000
Moment Method
(XSTABL + Bishop +
triangular distribution)

3500
3000

Force method (Slide +


Spencer + concentrated
load)

2500
2000

Force method (Slide +


Spencer + triangular
distribution)

1500
1000

FLAC (Elastic pile)

500
0
0.35

0.40

0.45

0.50

0.55

0.60

0.65

0.70

s/L

84

42

Results of Limited Comparison


Maximum difference in Unbalanced Forces ~10%
Maximum difference in Bending moments ~20%
(with forced triangular distribution)
Generally good
agreement between FLAC
and conventional Force
and Moment methods
when adopting a
reasonable triangular
distribution and p-y
method used

85

86

43

SRM for slope stabilization


Numerical Modeling in combination with the
Strength Reduction Method is a powerful and
versatile tool to design complex slope
stabilizations
In provides not only FOS that are meaningful to
the Geotechnical Engineer but can incorporate
plastic moments, tieback tensile loads, etc.
Hence, it can eliminate the problem of duplication
of FOS (geotechnical + structural)
87

The main advantages of using Numerical


Modeling vs. Limit Equilibrium are:
No failure hypothesis (or guess) is necessary with SRM
Critical failure mechanism is generally evident and even

complex failure mechanisms become apparent


Numerous runs can be automatically performed for
sensitivity analyses (e.g., with varying pile location, plastic
moment Mp or a tieback force Tmax, soil properties)
Models can be used during construction for deterministic
analyses, if instrumentation is available. Additionally,
benchmarks can be established in advance.
When used in addition to Limit Equilibrium it leads to less
uncertainty in the analysis, i.e., higher confidence.
88

44

89

45

You might also like