Computers and Geotechnics 32 (2005) 304315

www.elsevier.com/locate/compgeo

Coupled eects in stability analysis of pileslope systems

Jinoh Won a, Kwangho You b, Sangseom Jeong

a,*

, Sooil Kim

Department of Civil Engineering, Yonsei University, Seoul 120-749, Republic of Korea

Department of Civil Engineering, The University of Suwon, Hwasung-Si 445-743, Republic of Korea

Received 12 January 2004; received in revised form 11 February 2005; accepted 18 February 2005
Available online 10 May 2005

Abstract
A numerical comparison of predictions by limit equilibrium analysis and 3D numerical analysis is presented for a slopepile system. Special attention is given to the coupled analysis based on the explicit-nite-dierence code, FLAC 3D. To this end, an internal
routine (FISH) was developed to calculate a factor of safety for a pile-reinforced slope according to a shear strength reduction technique. Coupled analyses were performed for stabilizing piles in a slope, in which the pile response and slope stability are considered
simultaneously and subsequently the factors of safety are compared to a solution for a homogeneous slope using an uncoupled analysis (limit equilibrium analysis). Based on a limited parametric study, it is shown that the factor of safety for the slope is less conservative for a coupled analysis than for an uncoupled analysis and thus represents a denitely larger safety factor when the piles are
installed in the middle of the slopes and the pile heads are restrained.

2005 Elsevier Ltd. All rights reserved.
Keywords: Pileslope system; Limit equilibrium analysis; 3D numerical analysis; Shear strength reduction technique; Coupled/uncoupled analysis;
Homogeneous slope; Factor of safety

1. Introduction
The stabilization of slopes by placing passive piles is
one of the innovative slope reinforcement techniques
in recent years. There are numerous empirical and
numerical methods for designing stabilizing piles. They
can generally be classied into two dierent types: (1)
pressure/displacement-based methods [110]; (2) nite
element/nite dierence methods [1115].
The rst type of method is based on the analysis of
passive piles subjected to lateral soil pressure or lateral
soil movements. Generally, the lateral soil pressure on
piles in a row is estimated based on a method proposed
by Ito and Matsui [3]. This model is developed for rigid
piles with innite length and is assumed that the soil is
rigid and perfectly plastic. Thus, this model may not
*

Corresponding author. Tel.: +82 2 2123 2800; fax: +82 2 364 5300.
E-mail address: soj9081@yonsei.ac.kr (S. Jeong).

0266-352X/$ - see front matter
2005 Elsevier Ltd. All rights reserved.
doi:10.1016/j.compgeo.2005.02.006

represent the behavior of actual piles in the eld: this

model does not take into account the actual behavior
of nite exible piles, soil arching and soft soil, etc.
[8,23] On the other hand, the corresponding lateral soil
movements are estimated using either measured inclinometer data or from an analytical result using the nite
element approach, empirical correlations or based on
similar case histories [9,14]. However, a major problem
with the displacement-based methods is the estimation
of free soil displacements, because lateral soil displacements are notoriously dicult to estimate accurately.
Moreover, the rst method for pile-reinforced slopes
often uses limit equilibrium, where soilpile interaction
is not clearly considered and thereby has some degree
of weakness in representing the real pileslope system.
The second type of method has been used to investigate
the pileslope system, which is analyzed as a continuous
elastic or elasto-plastic medium using either niteelement or nite-dierence formulations. This method

J. Won et al. / Computers and Geotechnics 32 (2005) 304315

305

lateral soil movement

D

s
s

unstable layer
passive
portion
stable layer
active
portion

(b) Pile response.

Lx

(a) Pileslope system.

Fig. 1. One-row piles undergoing lateral soil movements.

provides coupled solutions in which the pile response and

slope stability are considered simultaneously and thus,
the critical surface invariably changes due to the addition
of piles, even though it is computationally expensive
and requires extensive training because of the threedimensional and nonlinear nature of the problem.
For slopes, the factor of safety F is traditionally
dened as the ratio of the actual soil shear strength to
the minimum shear strength required to prevent failure
[16]. As Duncan [17] points out, F is the factor by which
the soil shear strength must be divided to bring the slope
to the verge of failure. Since it is dened as a shear
strength reduction factor, an obvious way of computing
F with a nite element or nite dierence program is
simply to reduce the soil shear strength until collapse occurs. The resulting factor of safety is the ratio of the
soils actual shear strength to the reduced shear strength
at failure. This shear strength reduction technique was
used as early as 1975 by Zienkiewicz et al. [18], and has
been applied by Naylor [19], Donald and Giam [20],
Matsui and San [21], Ugai and Leshchinsky [22], Cai
and Ugai [23] and You et al. [24], etc.
The shear strength reduction technique is used in this
study. It has a number of advantages over the method of
slices for slope stability analysis. Most importantly, the
critical failure surface is found automatically. Application of the technique has been limited in the past due
to a long computational run-time. But with the increasing speed of the desktop computer, the technique is
becoming a reasonable alternative to the method of
slices, and is being used increasingly in engineering
practice.
In this study, factors of safety obtained with the shear
strength reduction technique were investigated for the
one-row pile groups on the stability of the homogeneous
slope. The case of an uncoupled analysis using limit
equilibrium analysis and subsequently the response of
coupled analysis based on the shear strength reduction
method were performed to illustrate the changes of crit-

ical surface invariably due to addition of piles on the

pileslope stability problem. The coupled eects were
tested against other case studies on the pileslope stability problem (see Figs. 1 and 2).

2. Uncoupled analysis by limit equilibrium method

A comprehensive study of uncoupled analyses has
been reported by Jeong et al. [10]. They report an uncoupled analysis in which the pile response and slope stability are considered separately. Here, the slopepile
stabilization scheme analyzed is shown in Fig. 3. The
conventional Bishop simplied method is employed to
determine the critical circular sliding surface, resisting
moment MR and overturning moment MD. The resisting
moment generated by the pile is then obtained from the
pile shear force and bending moment developed in the
pile at the depth of the sliding surface analyzed. It is assumed that the lateral soil pressure exerted by the sliding
slope on the pile results in the mobilization of shear

Soil pressure
z

passive portion
interface
active portion

Pu

pile

Fig. 2. A pile subjected to lateral soil pressure.

306

J. Won et al. / Computers and Geotechnics 32 (2005) 304315

Based on this, the safety factor of the reinforced slope

with respect to circular sliding is calculated as:
F F i DF

Fig. 3. Forces on stabilizing piles and slope.

forces and bending moment. The plastic state theory

developed by Ito and Matsui [3] is used to estimate the
pressure acting on the pile, q, as follows (Fig. 4):
"






1
D1
D2
p /

exp
q Ac
N / tan / tan
N / tan /
8 4
D2
#

1=2

1=2
2 tan / 2N / N /
1=2

2N / tan /
1
1=2
N / tan / N /
1
!
1=2

1=2
2 tan / 2N / N /

1=2

c D1

2D2 N /
1=2
N / tan / N /
1







cz
D1
D2
p /

A exp
N / tan / tan

D2 ;
N/
8 4
D2
1
where c is the cohesion intercept; D1 is center-to-center
distance between piles; D2 is opening between piles; /
is angle of internal friction of soil; c is unit weight of soil;
z is depth from ground surface; N/ = tan2 [(p/4) + (//2)]
and A D1 D1 =D2

1=2

N /

tan /N /
1

M R V cr  R  cos h
M cr V head  Y head

;
MD
MD

where Fi is the safety factor of unstabilized slope; DF is

increased safety factor of slope reinforced with pile; Mcr
is bending moment developed at critical surface; Vcr is
shear force developed at critical surface; Vhead is shear
force at pile head; R is radius of the sliding surface;
and h is the angle between a line perpendicular to the
pile and the failure surface. A computer program has
been developed using an uncoupled formulation to analyze the pileslope stability problem as described above
(Fig. 5).

3. Coupled analysis by strength reduction method

3.1. Shear strength reduction technique
To calculate the factor of safety of a slope dened in
the shear strength reduction technique, a series of stability analyses are performed with the reduced shear
strength parameters c0trial and /0trial dened as follows
(Fig. 6):
c0trial
/0trial

1 0
c;
F trial

3

arctan

F trial


tan / ;
0

where c 0 , / 0 are real shear strength parameters and Ftrial

is a trial factor of safety. Usually, initial Ftrial is set to be
suciently small so as to guarantee that the system is
stable. Then the value of Ftrial is increased by Finc values
until the slope fails. After the slope fails, the Fstart is
replaced by the previous Flow and Finc is reduced by 1/5.
Then the same procedure is repeated until the Finc is less
than user-specied tolerance (e). Fig. 7 shows the owchart of the routine to calculate a factor of safety. This
iterative procedure is based on the incremental search
method. This nal value of Flow, by denition, is identical to the one in limit equilibrium analysis. It should be
noted, however, that in the nite element and nite difference methods, local equilibrium is satised everywhere, whereas in the limit equilibrium analysis, only
global equilibrium for the sliding mass is considered in
the analysis.
3.2. Explicit nite dierence scheme

Fig. 4. Plastically deforming ground around stabilizing piles [3].

The response of a slopepile system is analyzed by

using a three-dimensional explicit-nite dierence

J. Won et al. / Computers and Geotechnics 32 (2005) 304315

307

start

Input shape of slope, soil property and pore water

pressure
Assume safety factor of slope and critical surface
by using simplified Bishop methods
Input flexible rigidity (EI), diameter, length and
boundary condition of stabilizing pile
Input position of stabilizing pile in slope
and center-to-center spacing

Calculate ultimate pressure (Pu)

Modify flexible
rigidity (EI) and
diameter

Input soil pressure(Ito & Matsuis pressure)

Analysis the behavior of stabilizing pile

based on pressure-based method
Modify
position and
spacing

Calculate displacement, bending moment, shear

force and soil reaction force

No

Compare with allowable

displacement and allowable
bending moment
Yes
Calculate bending moment and shear force on
critical surface
Calculate safety factor of stabilized slope

Compare with safety factor

of un-stabilized slope

No

Yes
Determine optimized flexible rigidity (EI), diameter,
position and spacing of stabilized pile

end

Fig. 5. Flow chart of computer program.

Fig. 6. A relationship between the actual strength and a strength

reduced by a trial factor of safety.

approach. The mesh consists of three-dimensional eightnoded solid elements and is assumed to be resting on a
rigid layer, and the vertical boundaries at the left- and

right-hand sides are assumed to be on rollers to allow

movement of soil layers.
The pile element is assumed to remain elastic at all
times, while the surrounding soil is idealized as a
MohrCoulomb elasto-plastic material. This model
was selected from among the soil models in the library
of FLAC 3D [25], the commercial explicit nite-dierence package used for this work. Factors of safety are
computed using FLAC 3D. In order to consider the
eect of an interface between a pile and soil, shell
elements, which satisfy the shear-yield constitutive
model, are used. The shear stiness of the shell element
is assumed to be isotropic and is inferred by the following equation:

308

J. Won et al. / Computers and Geotechnics 32 (2005) 304315

4. Validation and application of the coupled model

Fig. 7. Flowchart of the calculation routine for factor of safety.

K s aG=l;

where l is the minimum length of the shell elements and

G is the shear modulus of the soil adjacent to the shell
element. It is assumed that a = 20 is large enough to
make the initial slope of the load displacement relationship closely resemble the elastic analytical solution [23].
The normal stiness for the interface is taken as a very
high value under the reality that a pile and the surrounding soil do not overlap at the interface.
For a given element shape function, the set of algebraic equations solved by FLAC is identical to that
solved with nite element methods. In FLAC, however,
this set of equations is solved using dynamic relaxation
[26], an explicit, time-marching procedure in which the
full dynamic equations of motion are integrated step
by step. Static solutions are obtained by including
damping terms that gradually remove kinetic energy
from the system.
The convergence criterion for FLAC is the ratio
dened to be the maximum unbalanced force magnitude
for all the gridpoints in the model divided by the average
applied force magnitude for all the gridpoints. If a model
is in equilibrium, this ratio should be close to zero. For
this study, a simulation is considered to converge to
equilibrium when the ratio becomes less than 10
5.

The present coupled method is based on a shear

strength reduction technique using the explicit nite difference code, FLAC. The validation of the present coupled model was done by the comparison with others
coupled analysis results.
Cai et al. [23] performed a numerical analysis to
investigate the eect of stabilizing piles on the stability
of a slope. They performed a coupled analysis based
on a three-dimensional nite element method with an
elasto-plastic constitutive model and the shear strength
reduction technique. The numerical results by their coupled analysis were compared with those obtained by the
present method. However, it should be noted that the
nal calculated factor of safety by nite dierence methods depends highly on the size of element unlike nite
element methods in which a shape function can be used
within the elements. In general, the ner the size of elements is, the more precise the result is.
An idealized slope with a height of 10 m and a gradient of 1V:1.5H and a ground thickness of 10 m is analyzed with a three-dimensional nite element mesh, as
shown in Fig. 8. A steel tube pile with an outer diameter
(D) of 0.8 m was used. The piles are treated as a linear
elastic solid material and are installed in the middle of
the slope with Lx = 7.5 m, and the center-to-center spacing s = 3D. The piles are embedded and xed into the
bedrock or a stable layer. The material properties for
prediction purposes were selected based on Cai et al.s
assumptions, as shown in Table 1. The safety factor of
a slope stabilized with piles was compared for two dierent pile head conditions (free and xed) and two dierent pile Youngs modulus values (60 and 200 GPa).
When the slope is not reinforced with piles, the Cai
et al.s shear strength nite element method, the nite

Fig. 8. Model slope and nite element mesh [23].

J. Won et al. / Computers and Geotechnics 32 (2005) 304315

Table 1
Material properties and geometries [23]
Soil
Unit weight (kN/m3)
Plastic (MohrCoulomb)
Cohesion (kPa)
Friction angle ()
Dilation angle ()

10.0
20
0

Elastic
Elastic modulus (kPa)
Poissons ratio

2.0 105
0.25

Steel pile
Unit weight (kN/m3)
Elastic modulus (kPa)
Poissons ratio
Diameter (m)

78.5
2.0 108, 6.0 107
0.2
0.8

Interface
Elastic modulus (kPa)
Poissons ratio
Cohesion (kPa)
Friction angle ()
Dilation angle ()

2.0 105
0.25
10.0
20
0

20.0

dierence code, FLAC and Bishops simplied method

gave safety factors of 1.14, 1.15 and 1.13, respectively.
Based on this, it is obvious that the failure mechanism,
indicated by the nodal velocities in the shear strength
reduction techniques, agree well with the critical slip
circle given by Bishops simplied method.
On the other hand, the safety factor of a slope stabilized with piles for two dierent pile head conditions and
two youngs modulus values, were summarized in Table
2. Here, the critical depth was taken as the level of the
slip surface associated with the lowest factor of safety
although the FLAC could not predict a clear slip surface
like the limit equilibrium method. By comparing the
magnitude and distribution of the nodal velocities inside
the slope (Fig. 9(a)), especially at the pile position, the
nite dierence code, FLAC shows slightly higher values
in terms of predicted safety factors, compared with the
results by the shear strength reduction nite element
method. This is because there is a dierence in mesh
renement in the region surrounding the piles between
the two methods, even though in this study, a relatively
ne mesh was used near the pilesoil interface and became coarser further from the pile. It can be said that
these compare fairly well with each other with respect
to the calculated safety factors. However, the depth of
Table 2
Comparison of numerical methods on safety factor
Youngs modulus
of piles (GPa)

FLAC 3D

Cai and Ugai [23]

Bishop

Free

Fixed

Free

Fixed

Free

60
200

1.46
1.55

1.56
1.56

1.36
1.55

1.55
1.56

1.49
1.49

309

the slip surfaces predicted by the FLAC 3D analysis

are deeper than those located by Bishops simplied
method, where the reaction force of the piles is determined by ItoMatsuis equation (Fig. 9(b)). In addition,
the slip surface by the FLAC is divided into two dierent segments around the pile element, compared with
the unique single line by Bishops simplied method.
Thus, the depth and distribution of the slip surface implies that Bishops simplied method cannot indicate
the true failure mechanism for the slope reinforced with
piles.
Bishops simplied method can not incorporate the
pile head conditions well on the calculation of the safety
factor due to the limitation of ItoMatsuis equation,
which is derived for rigid piles. In this respect, Table 2
shows that Bishops simplied method can obtain a
smaller value of the safety factor, compared with the
xed head pile of FLAC 3D. The reason for this is that
because of the larger soil pressure, followed in order of
xed and ItoMatsuis pressure, the safety factor predicted by the FLAC 3D is denitely larger than that obtained by Bishops simplied method, as shown in Fig.
10 and in Table 2.

5. Comparison with other coupled analysis

5.1. Model slope
Hassiotis et al. [8] proposed a methodology for the
design of slopes reinforced with a single row of piles.
To estimate the pressure acting on the piles, they used
the theory developed by Ito and Matsui [3] and proposed a stability number by the friction circle method
to take into account the critical slip surface changes
due to the addition of piles. This is a pressure-based coupled analysis. The slope in Fig. 11 has a height of
13.7 m, a slope angle of 30, and is made of a homogeneous material with cohesion 23.94 kN/m2, friction angle 10 and unit weight 19.63 kN/m3. The water table
was not considered here. It was found that the safety
factor of the slope (without the pile reinforcement)
was about 1.08.
The slope is simulated by the FLAC 3D. Two symmetrical boundaries are used, so that the problem analyzed really consists of a row of piles with planes of
symmetry through the pile centerline and through the
soil midway between the piles. The actual size of the
mesh is related to the pile length; the lower rigid boundary has been placed at a depth equal to pile lengths and
the side boundary has been extended laterally to
rm = 2.5L(1
t) [27]. It is found that this size was sucient for the analysis of one-row pile groups. The material properties used for prediction purposes were
described in Table 3. A numerical comparison of predictions by the two analyses is presented below.

310

J. Won et al. / Computers and Geotechnics 32 (2005) 304315

(a) Nodal velocity vectors by FLAC 3D and critical slip circle by Bishops simplied method.

(b) Comparison of the depth and shape of critical slip surface.

Fig. 9. Comparative results between shear strength reduction and Bishops simplied method.

5.2. Eect of pile bending stiness

The eect of the bending stiness is investigated by
changing only the equivalent Youngs modulus, Ep, of
the piles. The piles are installed with Lx = 12.2 m, and
the center-to-center spacings of 2.5D, 3.0D, 3.5D, and
4.0D. As shown in Fig. 12(a), the safety factor of a slope
stabilized with piles for dierent bending stiness values
shows that the pile head conditions have more inuence
on the safety factor of the slope when the piles are more
exible (Ep = 1.43 GPa). However, for piles with larger
Youngs modulus (Ep = 25 GPa), the safety factor is
almost the same, regardless of the pile head conditions.
This is because the pressure on the free headed pile with
smaller Youngs modulus (Ep = 1.43 GPa) are consider-

ably smaller than that of xed head piles, whereas the

piles with larger Youngs modulus (Ep = 25 GPa) have
almost identical pressure distributions, regardless of
the pile head conditions, as shown in Fig. 12(b) and
(c). In addition, the pressure on the piles is almost the
same for the two bending stiness values when the pile
head is xed.
In this study, the value of EpIp was taken as constant by assuming the pile elastic. However, the cracking of the pile itself may occur in the loading with a
signicant reduction in EpIp. To understand the true
behavior, yielding of the pile is considered by taking
into account the compressive strength of the concrete.
Fig. 13 shows the typical reduction distribution of the
EpIp as the bending moment is increased; therefore, a

J. Won et al. / Computers and Geotechnics 32 (2005) 304315

Displacement (cm)
3

Displacement (cm)
12

15

Depth (m)

Depth (m)

311

10

12

12

Flac 3D (free)
Flac 3D (fixed)
Bishop (Ito-Matsui [3])

Flac 3D (free)
Flac 3D (fixed)
Bishop (Ito-Matsui [3])
15

15

Soil pressure (kPa)

Soil pressure (kPa)

0

40

80

120

160

200

Depth (m)

Depth (m)

40

80

120

160

200

12

12
Flac 3D (free)
Flac 3D (fixed)
Bishop (Ito-Matsui [3])

15

15

Flac 3D (free)
Flac 3D (fixed)
Bishop (Ito-Matsui[3])

(b) Pile modulus, Ep = 200 GPa.

(a) Pile modulus, Ep = 60 GPa.

Fig. 10. Pile behavior characteristics by FLAC 3D.

Table 3
Material properties and geometries [8]
D

16.7m

3.0m

1:1.7

Lx
10.0m

L=23.7m

20.0m

53.7m

Fig. 11. Model slope and element mesh [8].

s/2

Soil
Unit weight (kN/m3)
Plastic (MohrCoulomb)
Cohesion (kPa)
Friction angle ()
Dilation angle ()

23.94
10
0

Elastic
Elastic modulus (kPa)
Poissons ratio

4.79 103
0.35

Concrete pile
Unit weight (kN/m3)
Poissons ratio
Diameter (m)
Elastic modulus (kPa)
Compressive strength of the concrete (kPa)
Yield strength of the re-bar (kPa)

23.0
0.2
0.62
2.5 107
2.74 104
4.14 105

19.63

1.43 106
89.63

312

J. Won et al. / Computers and Geotechnics 32 (2005) 304315

2

safety factor

1.8

1.6

1.4

Ep=1.43Gpa
Flac 3D (free)
Flac 3D (fixed)

Ep=25Gpa

1.2

Flac 3D (free)
Flac 3D (fixed)
Hassiotis et al. [8]
Bishop

1
2

2.5

(a)

3.5

4.5

s/D
Soil Pressure (kPa)
0

100

200

300

Fig. 13. Bending stiness, EpIp as a function of bending moment for a

pile.

It is important to mention that in the numerical

results obtained by the FLAC analysis, the safety factor
of slopes reinforced with xed head piles is larger than
that with free head piles when the piles are more exible
(Ep = 1.43 GPa). Therefore, a restrained pile head
(xed) is recommended to stabilize the slope. For a xed
pile head condition, the safety factor predicted by
uncoupled analysis (e.g., Bishops simplied method) is
excessively conservative.

Depth (m)

4
free head
2.5D
3.0D
3.5D
4.0D

fixed head
2.5D
3.0D
3.5D
4.0D

10

5.3. Eect of pile spacing (s/D)

Ep =1.43 GPa

(b)

Soil Pressure (kPa)

80

120

160

200

240

280

Depth (m)

free head
2.5D
3.0D
3.5D
4.0D
fixed Head
2.5D
3.0D
3.5D
4.0D

10

(c)

Ep =25 GPa

Fig. 12. Eect of pile bending stiness and soil pressure. (b) Ep =
1.43 GPa; (c) Ep = 25 GPa.

modication in EpIp may be needed for accurate computations, especially if deection will control the
loading.

When piles with an equivalent Youngs modulus,

Ep = 25 GPa are installed with the horizontal distance
between the slope toe and the pile position, Lx of 7.6,
12.2, and 17 m, the eect of pile spacing on the safety
factor is shown in Fig. 14 for two dierent pile head conditions; free and xed. As expected, the safety factor increases signicantly as the pile spacing decreases. Here,
spacing equal to or larger than 2.5 diameters were
selected because the ratios less than 2.5 are not practical.
When a center to center spacing-to-diameter ratio is 2.5,
the safety factor of the slope approaches its maximum
value. This is explained by the fact that the lateral soil
movement between the piles is resisted more and more
by the piles as the spacing becomes closer and closer.
Fig. 14 also shows coupled eects in the safety factor
on pile spacings. The present method (FLAC analysis)
shows that the pile head conditions have more inuence
on the safety factor of the slope. The safety factor of a
slope reinforced with xed head piles, obtained by the
present method is a quite similar rate change but is
signicantly higher than that obtained the pressurebased coupled analysis proposed by Hassiotis et al. and
Bishops simplied method. The dierence in the safety

J. Won et al. / Computers and Geotechnics 32 (2005) 304315

2

2
Flac 3D (free)
Flac 3D (fixed)
Hassiotis et al. [8]
Bishop

1.8

safety factor

1.8

safety factor

313

1.6

1.4

1.2

1.6

1.4

Flac 3D (free)
Flac 3D (fixed)
Hassiotis et al. [8]
Bishop

1.2

1
2

2.5

3.5

4.5

2.5

s/D

3.5

4.5

s/D

(a) Lx = 7.6 m (Lx/L = 0.32).

(b) Lx = 12.2 m (Lx/L = 0.51).

2

safety factor

1.8

1.6

1.4

1.2

Flac 3D (free)
Flac 3D (fixed)
Bishop

1
2

2.5

3.5

4.5

s/D

(c) Lx = 17 m (Lx/L = 0.72).

Fig. 14. Eect of pile spacings on safety factor (Ep = 25 GPa).

factor between the coupled and uncoupled analyses can

be explained by the pressure acting on the piles presented earlier: the larger the pressure on the piles, the
larger the reaction force to the sliding body supplied
by the piles, and the higher the safety factor of the slope
reinforced with piles.
5.4. Eect of pile positions
Fig. 15(a) shows the safety factor as a function of
the relative position of the pile row with s/D of 2.5
on the slope. Here, the pile positions in the slope are
shown with a dimensionless ratio of the horizontal distance between the slope toe and the pile position, Lx,
to the horizontal distance between the slope toe and
slope shoulder, L. The coupled FLAC results, obtained
with the shear strength reduction technique, show that
the improvement of the safety factor of slopes reinforced with piles is the largest when the piles are in-

stalled in the middle of the slopes, irrespective of pile

head conditions. However, Hassiotiss coupled solution
shows that the piles should be placed slightly closer to
the top of the slope for the largest safety factor. This is
the same as the results of the Bishops method. The
reason for this is that when the piles are placed in
the middle portions of the slopes, the shear strength
of the soilpile interface is suciently mobilized by
the fact that the pressure acting on the piles is larger
than that on the piles in the upper portions of the
slopes (Fig. 15(b)).
Fig. 16 shows coupling eects in the safety factor
both on pile positions and on pile spacings. The safety
factors of slopes analyzed by coupled analyses are larger
than those by uncoupled analysis, as pile spacing decreases. This clearly demonstrates that the coupled eect
exists between piles and soil so that the critical slip surface can change due to the addition of piles. It is noted,
therefore, that the uncoupled analysis, which can only

314

J. Won et al. / Computers and Geotechnics 32 (2005) 304315

(a) 2.4

consider a xed failure surface, should be limited in its

application.

Flac 3D (free)
Flac 3D (fixed)
Bishop
Hassiotis et al. [8]

2.2

Safety factor

6. Conclusions
1.8

1.6

1.4

1.2

1
0

0.2

0.51

0.4

0.72

0.6

0.8

Lx/L
Soil Pressure (kPa)

(b)
0

50

100

150

200

250

300

Depth (m)

Lx/L=0.51
Flac 3D (free)
Flac 3D (fixed)
Bishop (Ito-Matsui [3])

Lx/L=0.72
Flac 3D (free)
Flac 3D (fixed)
Bishop (Ito-Matsui [3])

10

Fig. 15. Eect of pile positions on safety factor (s/D = 2.5) and soil
pressure (Ep = 25 GPa).
2.2
2.5D (Flac 3D)
3.0D
3.5D
4.0D

Safety factor

1.8

1.6

1.4
2.5D (Bishop)
3.0D
3.5D
4.0D

1.2

1
0.2

0.3

0.4

0.5

0.6

0.7

0.8

Lx/L

Fig. 16. Eects of pile positions and pile spacings on safety factor.

In this study, a coupled analysis of slopes stabilized

with a row of piles has been presented and discussed
based on an analytical study and a numerical study.
The numerical results are compared with those obtained
by the limit equilibrium method for slope stability analysis. A limited study of numerical analysis was carried
out to examine the pileslope coupling eect on relative
pile position and dierent pile spacings. The numerical
results have clearly demonstrated the important coupling eect of stabilizing piles in a slope with dierent
head conditions and pile bending stiness. From the
ndings of this study, the following conclusions are
drawn:
(1) A coupled eect has been identied between piles
and soil, so that the critical slip surface invariably
changes due to the addition of piles. It is noted,
therefore, that the uncoupled analysis, which can
only consider a xed failure surface, should be limited in its application. Hassiotis et al.s coupled
analysis based on the modied friction circle
method is a relatively eective for a pileslope system. However, their approach is intermediate in
theoretical accuracy between coupled FLAC analysis and uncoupled analysis (Bishops simplied
method).
(2) Through comparative studies, it has been found
that the pile head conditions and the bending
stiness inuence the safety factor of the slopes.
If the piles are more exible, the safety factor of
the slope is signicantly smaller than that of a
slope reinforced with xed head piles. However,
for piles with larger Youngs modulus
(Ep = 25 GPa), the safety factor is almost the
same, regardless of the pile head conditions. As
a result, for xed pile head condition, the prediction in the factor of safety in slope is much more
conservative for an uncoupled analysis than for a
coupled analysis.
(3) The numerical results show that the pressure acting on the piles is the largest when the piles are
placed in the middle portion of the slope. Therefore, the piles should be installed in the middle
of slopes and restrained in the pile head, when
the stability of a slope is required to be improved
optimally. A restrained head condition can be
obtained by connecting the pile heads with a buried beam which is xed by the tie-rods or tension
anchors.

J. Won et al. / Computers and Geotechnics 32 (2005) 304315

References
[1] De Beer EE, Wallays M. Forces induced in piles by unsymmetrical surcharges on the soil round the piles. Conf Soil Mech
Found Eng 1972;1:32532.
[2] Tschebotario GP. Lateral pressure of clayey soils on structures.
In: Proceedings of the 8th ICSMFE Specialty Session 5, Moscow,
vol. 4(3); 1973. p. 22780.
[3] Ito T, Matsui T. Methods to estimate lateral force acting on
stabilizing piles. Soils Found 1975;15(4):4359.
[4] Baguelin F, Frank R, Said YH. Theoretical study of lateral
reaction mechanism of piles. Geotechnique 1977;27(3):40534.
[5] Bourges F, Frank R, Mieussens C. Calcul des eorts et des
deplacements engendres par des poussees laterals de sol sur les
pieux. Note Technigue. Paris: Laboratoire Central des Ponts et
Chausees; 1980.
[6] Springman SM. Lateral loading on piles due to simulated
embankment construction. PhD thesis, University of Cambridge;
1989.
[7] Poulos HG. Design of reinforcing piles to increase slope stability.
Can Geotech J 1995;32:80818.
[8] Hassiotis S, Chameau JL, Gunaratne M. Design method for
stabilization of slopes with piles. J Geotech Geoenviron Eng,
ASCE 1997;123(4):31423.
[9] Chen LT, Poulos HG. Piles subjected to lateral soil movements. J
Geotech Geoenviron Eng, ASCE 1997;123(9):80211.
[10] Jeong S, Kim B, Won J, Lee J. Uncoupled analysis of stabilizing
piles in weathered slopes. Comput Geotech 2003;30:67182.
[11] Rowe RK, Poulos HG. A method for predicting the eect of piles
on slope behavior. In: Proceedings of the 3rd ICONMIG,
Aachen, vol. 3; 1979. p. 107385.
[12] Oakland MW, Chameau JLA. Finite-element analysis of drilled
piers used for slope stabilization. Laterally Loaded Foundation.
American Society for Testing and Materials; 1984. p. 18293.
[13] Bransby MF, Springman SM. 3-D nite element modelling of pile
groups adjacent to surcharge loads. Comput Geotech 1996;
19(4):30124.

315

[14] Goh ATC, The CI, Wong KS. Analysis of piles subjected to
embankment induced lateral soil movements. J Geotech Geoenviron Eng, ASCE 1997;123(4):31223.
[15] Poulos HG, Chen LT. Pile response due to excavation-induced
lateral soil movement. J Geotech Geoenviron Eng, ASCE
1997;123(2):949.
[16] Bishop AW. The use of slip circle in the stability analysis of
slopes. Geotechnique 1955;5:717.
[17] Duncan JM. State of the art: limit equilibrium and nite-element
analysis of slopes. J Geotech Eng, ASCE 1996;122(7):57796.
[18] Zienkiewicz OC, Humpheson C, Lewis RW. Associated and nonassociated visco-plasticity and plasticity in soil mechanics. Geotechnique 1975;25(4):67189.
[19] Naylor DJ. Finite element and slope stability. Num Meth
Geomech. In: Proceedings of the NATO Advanced Study
Institute, Lisbon, Portugal; 1981. p. 22944.
[20] Donald IB, Giam SK. Application of the nodal displacement
method to slope stablilty analysis. In: Proceedings of the 5th
AustraliaNew Zealand conference on geomechanics, Sydney,
Australia; 1988. p. 45660.
[21] Matsui T, San KC. Finite element slope stability analysis by shear
strength reduction technique. Soils Found 1992;32(1):5970.
[22] Ugai K, Leshchinsky D. Three-dimensional limit equilibrium and
nite element analysis: a comparison of result. Soils Found
1995;35(4):17.
[23] Cai F, Ugai K. Numerical analysis of the stability of a slope
reinforced with piles. Soils Found, Jpn Geotech Soc
2000;40(1):7384.
[24] You KH, Park YJ, Dawson EM. Stability analysis of jointed/
weathered rock slopes using the HoekBrown failure criterion.
Geosyst Eng 2000;3(3):907.
[25] FLAC, Fast Lagrangian analysis of continua, version 3.3, Itasca
Consulting Group; 1995.
[26] Otter JRH, Cassell AC, Hobbs RE. Dynamic relaxation. Proc
Inst Civil Eng 35:63356.
[27] Randolph MF, Wroth CP. Analysis of deformation of vertically
loaded piles. J Geotech Geoenviron Eng 1978;104(12):146588.