Escolar Documentos
Profissional Documentos
Cultura Documentos
ABSTRACT
This paper describes a numerical approach for the design of drilled shafts to
reinforce slopes. The approach uses the program FLAC and the strength reduction
method to predict the Factor of Safety of the improved slope. The method
successfully predicted the maximum demands (shear force and bending moments)
necessary for the structural design of stabilizing piles. The method also successfully
predicted the critical mode of failure and slope stability Factors of Safety for piles
with a plastic bending moment.
INTRODUCTION
In urban environments, drilled shafts are often used by geotechnical engineers
to improve slope stability and stabilize existing landslides (e.g., Smethurst and Powrie,
2007). Permanent drilled shafts used in landslide stabilization must resist the large
lateral loads imposed by the landslide. Typical shafts are closely spaced and have
large diameters (1 to 3 meters). Because of the large bending moments to which they
are subjected, these drilled shafts are generally reinforced with large steel I-beams
instead of reinforcement cages and the design often incorporates several rows of
tiebacks that extend below the landslide basal plane. Depending on the geographical
area these stabilizing drilled shafts are often known as stabilizing piles, shear pins,
caissons, or soldier piles.
DESIGN METHODS
The design methods used by geotechnical engineers for the design of
stabilizing piles vary widely. To prevent movement of the stabilized slope and upper
building pads, plastic flow of the soil between the shear pins must be prevented. Thus
in practice geotechnical designs typically involve closely spaced drilled shafts with
typical center-to-center spacing on the order of 2 to 3 shaft diameters. In frictional
materials, a spacing of 3 shaft diameters is generally considered adequate to ensure
full passive load transfer (Broms, 1964), prevent plastic flow (Smethurst and Powrie,
2007), and transfer the maximum lateral loads to the stabilizing piles by arching. In
rare situations where a large spacing is specified, lateral deformations associated with
soil flow between the piles should be anticipated and load transfer by arching must be
920
921
922
slip surface, the stiffness of the pile, and the resisting forces of the pile, are
compatible. Poulos (1995) proposed a relatively similar method that involves
three steps: (a) evaluating the total shear forces needed to bring the slope to a
target Factor of Safety, (b) evaluating the maximum resistance that each pile
can provide (using, e.g. Viggiani 1981), and (c) selecting the type and number
of piles. The method uses a proprietary program called ERCAP to model the
interaction of the slope and stabilizing piles. Due to the complexity of the
Reese and Poulos methods, designers rarely use them in practice.
Numerical Solutions Methods: Designs based on numerical methods, i.e.,
using the Finite Element Method (FEM) or Finite Difference Method (FDM),
have previously been designed and implemented. Most designs use numerical
modeling to predict the likely bending moments and deflections of the
stabilizing piles and do not predict the Factor of Safety (FOS) of the improved
slope. Thus these designs do not necessarily satisfy the requirements of
customary building codes. As a consequence, reviewing government agencies
will generally request that the geotechnical engineer demonstrates, in addition
to the FEM/FDM model, that the design satisfies the minimum required
calculated Factors of Safety using conventional slope stability methods.
L
30 deg.
H=13.7m
S
Soil:
c = 23.94 kPa
= 10 deg.
= 19.63 kN/m3
Fu
D=3.0m
Bedrock:
c = 23.94 kPa
= 40 deg.
= 19.63 kN/m3
Stabilizing Pile
cd =
tan(d ) =
922
c
R
(1)
tan( )
R
(2)
923
The highest value of the strength reduction factor that satisfies equilibrium
and is kinematically acceptable is the Factor of Safety, FOS, of the slope:
Rmax = FOS
(3)
METHODOLOGY
To evaluate the capabilities of the strength reduction method for the analysis
of a slope stabilized with drilled shafts, we adopted the soil strength, pile stiffness and
slope geometry, similar to a previously published study (Hassiotis et al., 1997). The
characteristics of the slope and stabilizing piles are shown in Figure 1.
We used the program FLAC (Itasca, 2008) to model the soil and piles, and
programmed automated routines in FLAC to slowly lower the shear strength of the
soil in accordance with equations 1 and 2. Since a row of closely spaced stabilizing
piles will act as a wall due to arching, the piles were modeled using structural beam
elements. FLACs pile elements were also used in a limited number of runs without
discernable geotechnical difference. Frictional interfaces connected the beam
elements to the soil nodes on both the upslope and downslope sidesBeam elements
are directly attached to the soil nodes on both the upslope and downslope sides, hence
full friction between pile and soil is assumed to be mobilized. An elastic modulus of
the stabilizing piles E = 200 GPa, a cross-sectional area A = 1 m2/m, and moment of
inertia I = 1 m4/m were specified for elastic piles.
Stabilizing Pile
Soil
Bedrock
After bringing the original slope to static equilibrium, the drilled shafts were
installed in the Finite Difference Model (Figure 2). Then very slowly, i.e., in
thousands of steps, the strength of the soil and bedrock were decreased by a reduction
factor, R. The procedure was programmed into FLAC and the reduction factor, R, was
increased until the predicted maximum displacement, max, exceeded 2 meters (e.g.,
Figure 3). The procedure was repeated for different pile locations and two different
values of the angle of dilation were used in our analyses, = 0 and = . In
addition, the piles were modeled as both linear elastic (without a maximum bending
moment), and with a maximum plastic moment, Mp = 2,000 kN.m/m.
923
924
The procedure worked flawlessly and the strength reduction technique easily
predicted the Factor of Safety, FOS, for all the slope configurations. Results from a
FLAC run with a row of elastic stabilizing piles located at S = 10.45 m from the toe
of the slope (S/L = 0.44) are shown in Figures 3, 4 and 5b. Figure 3 shows that as R
approaches Rmax = 1.68, displacements become very large, which means that the slope
is yielding. Figure 4 shows that the relationship between displacements and bending
moments is roughly bilinear, and that as the slope starts to yield the bending moment
remains approximately constant (Mmax = 3750 kN.m/m). Hence, one of the main
advantages of this procedure is that it provides the maximum structural demands in
addition to the Factors of Safety of the slope.
5
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
Figure 3. Maximum displacement vs. strength reduction factor for an elastic pile.
5000
4500
4000
3500
3000
L=23.73m
S/L=0.44
2500
S=10.45m
2000
1500
1000
500
0
0
0.5
1.5
Another advantage of the strength reduction technique is that the critical mode
of failure can be easily determined. As the reduction factor, R, approaches the Factor
of Safety of the slope, the predicted displacements become increasingly larger and a
zones with large shear strain increments become apparent. The zone with maximum
shear strain increments is the basal plane of the critical failure surface, and defines the
critical mode of failure of the slope. Three different modes of failure were predicted
924
925
in our analyses depending on the location of the stabilizing piles and the limiting
bending moment, Mp:
Slope failure below the stabilizing piles (Figure 5a).
Slope failure above the stabilizing piles (Figure 5b). It should be noted that
even when the failure involves slippage above the pile, bending moments are
predicted because of the plastic yielding of soil (due to strength reduction).
Development of a plastic hinge in the stabilizing piles and failure of the slope
through the stabilizing shear pins (Figure 5c). This mode of failure only
appeared when a small enough plastic moment was specified.
Figure 5. Bending moment, displacement vectors and shear strains for three
different locations of the stabilizing pile and/or pile plastic moment, Mp.
Plastic Moment
926
the maximum Factor of Safety. For elastic stabilizing piles (without maximum plastic
moment), the maximum Factor of Safety is achieved by placing stabilizing piles at
S/L = 0.51. With a limiting bending moment, Mp = 2000 kN.m/m, the optimum
location of the stabilizing piles changes and moves to S/L = 0.4. This change is
important because it shows that the methodology presented here can be used to
optimize the location of stabilizing piles in slopes. As shown in Figure 6 the Factors
of Safety predicted by the strength reduction method compares well with those
obtained by the conventional equilibrium method using simplified Bishops method.
2
FLAC: Elastic Pile, = 0
1.9
1.8
1.7
1.6
1.5
1.4
1.3
1.2
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
1.1
1.2
1.3
1.4
5000
Plastic Pile, = 0
Plastic Pile, =
4000
3000
2000
1000
0
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.1
1.2
1.3
1.4
As R approached Rmax = FOS, the solutions obtained with FLAC are both
statically and kinematically admissible. Nevertheless, since most theoretical plasticity
solutions in geotechnical engineering have been obtained using associated models,
where the dilation angle is = (Schofield and Wroth, 1968), it was considered
worthwhile to study the effect of the dilation angle on the FOS and maximum
bending moments. As can be seen in Figures 6 and 7, the effect of is minor and can
926
927
This paper describes a numerical approach for the design of drilled shafts to
enhance the Factor of Safety of a slope using the strength reduction technique. Unlike
conventional slope stability methods which require assumptions such as failure
geometry, load distribution on the pile and distance to the point of fixity, the
numerical modeling scheme requires no aforementioned assumptions and is able to
predict the structural demand (bending moment and shear forces) of the pile directly.
With the incorporation of strength reduction method, the Factor of Safety of an
improved slope can be found easily. In addition, the method can be used to find the
optimum location of stabilizing piles and predict the critical mode of failure. In our
analyses, the angle of dilation did not appear to affect the results. In conclusion, the
numerical approach with the use of strength reduction technique is straightforward,
requires few hypotheses and is easy to implement.
REFERENCES
927
GEOTECHNICAL
SPECIAL
PUBLICATION
NO.
208
EARTH RETENTION
CONFERENCE 3
PROCEEDINGS OF THE 2010 EARTH RETENTION CONFERENCE
SPONSORED BY
EDITED BY
2010023353