Ground Movements A Hidden Source of Loading on

Deep Foundations
Harry G. Poulos, Coffey Geosciences, Sydney, Australia; +61 2 9911 1000; [email protected]

Ground movements can arise from a large number of sources and can have a significant effect on
nearby piles and deep foundations. The loading of the piles by ground movements is a different
mechanism to that arising from direct applied loading to the pile head, and consequently it is not
generally possible to adequately analyze the effects of ground movements simply by applying
some type of equivalent loading to the pile head. The main effects of ground movements are the
development of additional movements, axial forces and bending moments in the piles, and thus the
key design aspects are related to movements and to the structural integrity of the pile. However,
the ultimate geotechnical load carrying capacity is generally not affected by the ground movements
themselves.
This paper will describe an approach to the analysis of ground movement effects on piles, considering
axial and lateral movements separately. Some of the main features of pile response will be discussed
for three specific problems involving ground movements:
1. Piles near and within embankments;
2. Piles near an excavation for a pile cap;
3. Piles subjected to seismic ground motions.

Introduction
There are many circumstances in which pile
foundations may be subjected to loadings
arising from vertical and/or lateral movements Consolidation Expansive soil Piles near tunnelling Installation of
operations adjacent piles
of the surrounding ground. Fig. 1 illustrates
a number of these circumstances. In such
cases, at least two important aspects of pile
foundation design must be considered:
1. the movements of the piles caused by the
Slope instability Piles near an Excavation Construction of
ground movements embankment adjacent building

2. the additional forces and/or bending [Fig. 1] Some Sources of Ground Movements

moments induced in the piles by the

ground movements, and their effect on the This paper summarizes a consistent theoretical
structural integrity of the piles. approach to the analysis of ground movement
effects on piles, for both vertical and horizontal
Problems involving the effects of ground
movements, which falls into the second
movements on piles may be analyzed in at least
category. Two distinct stages are involved in
two ways:
this analysis:
1. Via a complete single analysis (generally
1. Estimation of the free-field soil
numerical) involving modelling of the
movements which would occur if the pile
pile, the soil and the source of the ground
was not present;
movements. This will give a complete
solution for the behaviour of both the soil 2. Calculation of the response of the pile to
and the pile. these computed ground movements.
2. Via a simplified approach involving initial Some specific cases of ground movement are
separation of the soil and the pile (sub- then considered, and in each case, a discussion
structuring) so that the soil movements are is given of the general features of pile behaviour
first computed and then imposed on the revealed by the theory, and typical applications
pile. In this approach, the focus is generally to practical field cases are described.
placed on the behavior of the pile.

DFI JOURNAL Vol. 1 No. 1 November 2007 [37]

 Analysis of Pile Response to Externally allowed for, thus allowing a nonlinear response
Imposed Soil Movements to be obtained.
Axial response The analysis has been implemented via
FORTRAN computer programs, including a
The analysis used for axial response of the pile
proprietary program called ERCAP, and in an
has been described by Poulos and Davis (1980)
alternative approach, via the program PALLAS
and has been used to analyze problems of
(Hull, 1996).
negative friction of piles in consolidating soil,
and of tension and uplift of piles in expansive Group Effects
soil. It employs a simplified form of boundary-
The analysis of ground movement effects on
element analysis, in which the pile is modelled
groups of piles has been reported by several
as an elastic column and the surrounding soil as
authors, for example, Kuwabara and Poulos
an elastic continuum.
(1989), Teh and Wong (1993), Chow et al (1990),
The pile is divided into a series of cylindrical Xu and Poulos (2001). All these authors have
elements. The vertical movement of each found that under purely elastic conditions, group
element is related to the applied load, effects tend to be beneficial to the pile response
the pile-soil interaction stresses, the pile as compared to single isolated piles, i.e. the
compressibility, and the pile tip movement. group effects tend to reduce the pile movement
The vertical movement of each supporting soil and the forces and moments induced in the
element depends on the pile-soil interaction piles. This is especially so for the inner piles
stresses, the modulus or stiffness of the soil, within a group, which, because of the pile-soil-
and also on any free-field movements that may pile interaction are, in effect, shielded from the
be imposed on the pile. To simulate real pile soil movements by the outer piles. Experimental
response more closely, allowance may be made work reported by Chen (1994) indicates that the
for slip at the pile-soil interface, i.e., the pile-soil ultimate lateral pile-soil pressures are affected
interaction stresses cannot exceed the limiting to some extent by grouping and that the group
pile-soil skin friction. effect may either increase or decrease the pile
The above analysis has been implemented via a response, depending on the pile configuration
FORTRAN computer program, PIES (Poulos 1989). and spacings.

The analysis of axial pile response requires a From the viewpoint of design, it is generally
knowledge of the pile modulus, the distribution both convenient and conservative to ignore
of soil modulus and limiting pile-soil skin group effects and analyze a pile as if it were
friction with depth, and the free-field vertical isolated. Thus, in the remainder of the paper,
soil movements. The assessment of the pile-soil attention is concentrated on ground movement
parameters (in particular the soil modulus and effects on single isolated piles.
limiting pile-soil skin friction) has been discussed
by several authors (e.g. Meyerhof 1976; de Cock
Loading Via Ground Movements Versus
and Legrand, 1997; Poulos 1989, 2001). Direct Applied Loading
There is a widespread misconception that
Lateral response the effects of externally imposed ground
Details of the lateral-response analysis have movements on piles can be estimated by the
been given by Poulos and Davis (1980), and it application of equivalent loadings at the pile
also relies on the use of a simplified boundary head. To illustrate the consequences of this
element analysis. In this case, the pile is procedure, the case in Fig. 2 has been analyzed.
modelled as a simple elastic beam, and the soil A single pile in a two-layer soil profile is
as an elastic continuum. The lateral displacement considered, and the pile is subjected to the
of each element of the pile can be related to the following sources of loading:
pile bending stiffness and the horizontal pile-soil An applied vertical load of 1.0 MN
interaction stresses. The lateral displacement of
An applied lateral load of 0.1 MN at the pile head
the corresponding soil elements is related to the
soil modulus or stiffness, the pile-soil interaction Vertical ground movement profile which
stresses, and the free-field horizontal soil decreases from 100 mm at the ground surface
movements. A limiting lateral pile-soil stress can to zero at a depth of 12 m
be specified so that local failure of the soil can be

[38] DFI JOURNAL Vol. 1 No. 1 November 2007

 A lateral ground movement profile which also pile. Again, it can be seen that the distribution
decreases from 100 mm at the ground surface of induced bending moment is very different
to zero at 12 m depth. for applied loading and for lateral ground
movement. In the latter case, the maximum
P = 1.0MN
100mm moment occurs well below the pile head, near
the bottom of the zone of ground movement.
The maximum bending moment under the
LAYER 1
E s = 15MPa
combined loadings is also at the latter location,
12
f s = 25kPa since the moment due to the applied loading
is virtually zero where the moment due to the
lateral ground movements is largest.
ASSUMED GROUND
MOVEMENT PROFILES
LAYER 2
8 Es = 120MPa Bending Moment MNm
fs = 100kPa -0.1 0 0.1 0.2 0.3

fs = 1.8MPa 0

0.100 MN Load Only

2
[Fig. 2] Typical Problem Analyzed
4
0.100 mm Lateral
Fig. 3 shows the computed axial load Movement Only
distributions for the applied load acting
6

alone, the vertical ground movement acting 8

0.100 MN Load +
100 mm Lateral
alone, and the applied load and the ground Movement

Depth m
movement acting together. It can be seen that
10

the distributions of axial load in the pile due 12

to applied loading are very different from

those induced by the ground movements. In 14

the latter case, the maximum axial load occurs 16

near the bottom of the upper soil layer which

is subjected to movement. It can also be seen 18

that the addition of the two profiles of axial 20

load gives axial loads which are less than those [Fig. 4] Comparison of Lateral Responses
computed for the combined loading and ground
movement case.
In terms of practical pile design, the above example
Axial Load demonstrates the following important points:
0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 The effects of ground movements can not be
0
simulated accurately by the application of a
2 load to the pile head;
The superposition of axial load distributions
4
due to axial applied loading and vertical
6 ground movements may underestimate the
maximum axial load in the pile;
8
The maximum load in a pile subjected to
lateral ground movements may occur well
Depth m

10

below the pile head. In this particular case,

having the pile reinforced only to resist
12

14 applied lateral loading (for example, in the

upper 6 m or so) will be inadequate to resist
16
Axial Load Only the ground-movement induced moments.
100 mm Soil Movement Only The pile may well fail structurally at a
18
Axial Load + 100 mm Soil Movement considerable depth below the pile head.
Sum of Axial Load+100 mm Soil Movs
20 Thus, it is important to consider the
[Fig. 3] Comparison of Axial Responses possibility of ground movements in pile
design, and to allow for reinforcement to
Fig. 4 shows the corresponding distributions of resist deep-seated moments that may be
moment computed for the lateral response of the induced by these movements.

DFI JOURNAL Vol. 1 No. 1 November 2007 [39]

 Generic Design Charts For Piles depth below the surface, may be adopted.
Subjected To Soil Movements The maximum value may be estimated from
measured ground surface movements or
Vertical Soil Movements via appropriate empirical approximations
Design charts for the settlement and maximum which relate movement to the height of the
axial force in a single end bearing pile on rock retained soil, for example, Peck (1969).
and subjected to vertical soil movements, have 2) For landslides involving relatively large soil
been published by Poulos and Davis (1980), movements (for example, up to about 0.4
While similar charts for a single floating or pile diameters), a uniform soil movement
friction pile have been presented by (Poulos, profile may be adopted.
1989; Poulos and Davis, 1980). Such charts
The above study by Chen and Poulos also
will tend to give upper bound values of both
shows that the elastic design charts can give
pile settlement and induced pile force, because
reasonably good estimations of the lateral pile
there is no limit to the pile-soil shear stress that
response, provided that the ground movements
is developed between the pile and the soil. In
are not very large, for example, less than about
reality, the existence of an ultimate skin friction
30-40% of the pile diameter.
will result in a limit to the axial force and pile
movement that can be generated within a pile.
Piles Near Embankments
The use of elastic solutions therefore tends to
be conservative when applied to practical cases. Introduction
Corrections for pile-soil slip and other practical The construction of embankments on clay
effects are presented by Poulos and Davis (1980) can result in the development of substantial
and Nelson and Miller (1992). immediate and time-dependent vertical and
horizontal movements of the soil beneath and
Lateral Soil Movements adjacent to the embankment. In situations
If the distribution with depth of free-field where bridge abutments adjacent to such
lateral movements can be simplified, it is embankments are supported on piles, these
possible to develop useful design charts to piles may experience significant axial and lateral
enable approximate assessment of the pile head loads which are induced by the soil movements.
deflection and the maximum bending moment The design of abutment piles therefore requires
in the pile. Chen and Poulos (1997, 1999) have a proper consideration of the pile response
presented such charts, both for a pile in soil to the externally imposed soil movements,
subjected to a uniform movement with depth including an assessment of the consequent
(to a depth zs below the surface), and for a soil bending moment and lateral deflection profiles
in which the horizontal movement decreases of the piles.
linearly with depth, from a maximum at the Marche and Lacroix (1972), Heyman (1965), and
surface to zero at a depth zs. These solutions Heyman and Boersma (1965) have presented
assume that the soil remains elastic, and field data of the lateral pile response to
they therefore generally give an upper bound embankment-induced soil movements. Poulos
estimate of the pile moment and deflection. The and Davis (1980) have compared the results
extent of the possible over-estimation increases of a boundary element analysis with this field
with increasing lateral soil movements, due data, and have found fair agreement. Stewart
to the progressive departure from elastic et al (1991, 1992) have conducted centrifuge
conditions which results from the development model tests of piles bridge abutments and
of plastic flow of the soil past the pile.
have developed an empirical procedure for
From a study of several examples, Chen and estimating the lateral deflection and bending
Poulos (2001) have suggested the following moment in a pile or pile group. Poulos (1996)
preliminary guidelines for the determination has compared various methods of analyzing
of soil movements in making theoretical lateral pile behaviour, including some design
predictions via the generic design charts : methods based on estimation of pressures
1) For unstrutted excavations or relatively developed between the pile and the soil. Such
small slope movements, a linear soil methods appear to be generally less reliable
movement profile, with a maximum value than methods based on a proper pile-soil
at the ground surface and zero at a certain interaction analysis.

[40] DFI JOURNAL Vol. 1 No. 1 November 2007

Estimation of Ground Movements The ratio of undrained shear strength of
Settlements beneath and near embankments surface crust and underlying clay (M=sus/su).
can, in principle, be predicted by conventional They identified three general distributions of
methods of settlement analysis, whether they horizontal displacement, which are illustrated
are based on one-dimensional analysis, elastic- in Fig. 5b. These distributions apply at or
based analysis, or numerical analyses, such as near the toe of an embankment, and can be
the finite element method or the program FLAC. expressed by the following equations:
An important requirement for such settlement Curve 1 (Overall mean curve):
predictions is to make allowances for non-
Y = 1.83Z3 4.69Z2 + 2.13 Z +0.73 (1a)
linear soil behaviour, including differences
in compressibility in the normally- and over- Curve 2 (where the compressible layer lies
consolidated states, local yielding or failure several metres below the surface):
within highly-stressed zones, and the effects of Y = -2Z3 + 1.5Z + 0.5 (1b)
consolidation and creep.
Curve 3 (where the soil compressibility is
In contrast, horizontal movements below and reasonably uniform with depth):
near embankments are difficult to predict
Y = 3.42Z3 - 6.37Z2 + 2.14Z + 0.81 (1c)
accurately (e.g. Poulos, 1972) and it is often
more appropriate to make use of empirical where Y = dimensionless horizontal
information on such movements. Bourges and displacement = h / hmax
Mieussens (1979) have evaluated the results of a h = horizontal displacement
number of field observations and recommended hmax = maximum horizontal
a practical design approach which is displacement
summarized here. They have found (see Fig. 5a)
Z = z/D = dimensionless depth.
that the significant parameters of the horizontal
displacement profile are: The magnitude of the maximum horizontal
displacement, hmax, is given as the sum of the
The ratio of layer depth (D) to mean
immediate and consolidation components, i
embankment width (B);
and c respectively. i is related to the position
The ratio of distance (X) of a point from the of the point (X/L) and the safety level F, where:
crest, to the horizontal extent of the slope (Lw);
F = 5.14suav / H (2)
The undrained shear strength su;
where suav = average undrained shear strength in
X the clay layer
L
= unit weight of embankment fill
B

O r H
s
H = embankment height.
(C u) s
Pile From measured data, Fig. 6 plots the
D
dimensionless maximum immediate horizontal
z Cu

displacement ( , where = imax / D) against F

Z for three values of X/Lw. As would be expected,
(a) Definition of Embankment Geometry
increases as F decreases, and becomes
relatively large as failure is approached (i.e. F
y = /max
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 approaches 1.0).
0
0.1 The maximum consolidation lateral
0.2 displacement at any time, cmax,t, is correlated
0.3 with the consolidation settlement Sc of the
0.4 centre of the embankment, as follows:
0.5
0.6 2 cmax,t = 0.16 Sc = 0.16 (St - Si) (3)
1
0.7
3 where St = total settlement at time t
0.8
0.9 Si = immediate settlement.
1.0
Thus, at any time t, the maximum horizontal
Z/d

(b) Horizontal Movement Distributions with Depth

displacement hmax can be approximated as:

[Fig. 5] Pile Near Embankment and Soil Movement Distributions hmax = .D + 0.16(St Si) (4)

DFI JOURNAL Vol. 1 No. 1 November 2007 [41]

 of clay layer, embankment height, pile size, and
delayed installation of the pile. The latter is
2
particularly important, and it is found that, if
the installation of the pile can be delayed until
after embankment construction is completed,
1.5 1 X/L < 2 the maximum bending moments in the pile can
2 X/L < 3
Movement Factor %

3 X/L < 4 be reduced to 10-15% of the values which would

X/L = 0 otherwise occur.
1 As mentioned previously, group effects are
generally beneficial when piles are subjected to
soil movements, leading to a shielding effect
0.5
and reduced moments and shear forces in the
piles, and it is therefore generally conservative
X/L= 1 to consider a single pile for design purposes.
X/L= 2
X/L= 3
800
0 Measured (Stewart et al, 1991)
0.8 1.1 1.5 2 2.5 3 3.5 4 4.5
Single pile
Factor of Safety F

Maximum Bending Moment (kNm)

Group-front row
[Fig. 6 ] Dimensionless Immediate Horizontal Movement 600 Computed
Factor l (Bourges & Mieussens, 1979) Single pile
Group pile

Some Characteristics of Pile Behaviour 400

Poulos (1994) has undertaken a study of the

behaviour of piles near and through embankments
200
on clay, using a two-stage procedure:
1. assessment of the horizontal soil movements
due to the embankment, using the empirical 0
0 50 100 150
approaches developed by Bourges and Embankment Surcharge (kPa)
Mieussens (1979), described above;
[Fig. 7] Comparison Between Measured and Theoretical
2. analysis of the response of piles to these Maximum Moments Deep Clay Layer
movement, using the program ERCAP to
compute lateral response and PIES for Bending Moment (kNm)
-500 0 500
axial response. 0
Measured (Stewart et al, 1991)
Comparisons were made between the analysis Single pile
results and the results of centrifuge tests 5
Group-front row
Computed
reported by Stewart et al (1992). Fig. 7 compares Single pile
the relationship between maximum horizontal 10
Group pile
Depth (m)

movement of a pile and embankment height.

The centrifuge tests show that, beyond a certain
15
embankment height (which corresponds to an
average pressure of about 3suav), the rate of
pile displacement with increasing embankment 20

height accelerates. This behaviour is reproduced Embankment height = 8m

quite well by the theoretical analyses. Fig. 25

8 compares the measured and computed [Fig. 8] Comparison Between Measured and Computed
distributions of bending moment along a Moment Distributions Deep Clay Layer
pile, and shows fair agreement in both the
magnitude and distribution of moment. Typical Design Charts
A study of the importance of a number of For the idealized problem shown in Fig. 9(a),
parameters was also made by Poulos (1994), a series of design charts has been developed
and the factors which have the most significant for a range of values of embankment height,
influence on pile response were found to be: soil layer thickness, soil strength and pile size.
pile position relative to the embankment toe, For a typical case of a vertical pile located at
undrained shear strength of the clay, thickness the crest of the embankment, Fig. 9(b) to 9(d)

[42] DFI JOURNAL Vol. 1 No. 1 November 2007

shows the maximum positive and negative limiting soil pressure of about 2pp (where pp
moments and the pile head deflection for a is the Rankine passive pressure) (see Chen &
400 mm square precast concrete pile, for an Poulos, 1997). The measured free-field lateral
embankment height of 8 m. It is assumed that soil movements, shown in Fig. 10(a), generally
the pile head is restrained from rotation and decrease with depth, after reaching a maximum
that the pile is installed after construction of the at a relatively shallow elevation.
embankment has been completed. Such charts Chen & Poulos (1997) have shown that a full
have the potential to provide a more rational analysis via the computer program PALLAS
design approach for piles within embankments. can give estimations of pile bending moments
Successful use of such charts demands and deflections very close to those measured,
judicious selection of the necessary geotechnical using the measured soil movement profile
parameters and appropriate modelling of the shown in Fig. 10(a). The pile bending moment
real problem to reduce it to the idealised cases and deflection profiles estimated using
for which the design charts are derived. PALLAS are shown in Fig. 11, together with
Fixed head, free to translate those measured, and a fairly good agreement
Pile diameter = 400mm
1:1.5 slope 120 between the estimated and the measured
he Embankment
values can be observed.
100 h s (m)
Pile Top Deflection hs (mm)

hs Compressible
clay soil 80 = 1.6 t/m 3 Sheet pile wall
Fill = 32

16m
c = 0.5 t/m 2 Free support of

11m
the Pile Head
60
Firm bearing stratrum 24m 25m 8.3m
Row of Test Piles
40
Pile locations: (a) Lay-out of the test
D A Toe Soil horizontal
B Midslope Cone resistance (kg/cm2 ) displacement (mm)
20 c
C Crest 25 0 80 160 240 320 t/m3 t/m 2 0 -20 -40 -60
15 0
D Corner medium 2.0 35 0.5
dense
5 fine
C B A 0 loose 2.0 30 0.2
0 100 200 sand
90cm medium 2.0 35 0.3
Undrained Shear Strength c u (kPa) dense
10
(a) Problem Definition (b) Pile Head Deflection Depth (m) sandy
loam
28m

dense sand
20 with shells
Maximum Negative Moment M NS (kNm)

Maximum Positive Moment M PS (kNm)

Boom clay
-100 h s (m) -100 h s (m) dense
clayey sand
30
Asschian clay
(b) Data of the test on a steel tube pile (d= 90cm)

Soil horizontal
Cone resistance (kg/cm 2 ) displacement (mm)
c
-50 -50 0 80 160 240 320 t/m3 t/m2 0 -20-40-60
0 dense 2.0 35 0.5
25 fine loose 2.0 30 0.2
5
medium 2.0 35 0.3
90cm sand dense
15
15 10
23.20m

25 sandy
Depth (m)

5 loam
0 0
0 100 200 0 100 200 medium to dense
sand
Undrained Shear Strength cu (kPa) Undrained Shear Strength c u (kPa) 20
Boom clay
(c) Maximum Negative Moment (d) Maximum Positive Moment

[Fig. 9] Design Charts for Pile at Embankment Crest 4 m 30

(c) Data of test on a reinforced concrete pile pile (d= 60cm)
Embankment Height
Fig. 10 Test at Zelzate (De Beer & Wallays, 1972)
Application to Case Study
De Beer and Wallays (1972) reported a field test The soil movement profile has also been
in Belgium that aimed to study the influence simplified to a linear profile for two cases,
of embankment construction on adjacent pile to assess whether simplified soil movement
foundations. Measured results were presented estimates can give adequate predictions of pile
for a steel pipe pile and a reinforced concrete response. One case has a surface displacement
pile. The steel pipe pile was 28 m in length, (s0) of 20 mm, while the other has a so value
0.9 m in diameter, and 1.5 cm in wall thickness, of 40 mm, with both cases having a zero value
while the reinforced concrete pile was 23.2 m occurring at a depth of about 18 m. It has
in length and 0.6 m in diameter. The pile heads been found that the measured profiles are
were restrained from lateral displacement. The encompassed by those estimated for the above
soil deposit consisted mainly of sand, with a two so values.
Youngs modulus Es of about 30 MPa and the

DFI JOURNAL Vol. 1 No. 1 November 2007 [43]

 Bending moment (kN.m) Pile deflection (mm) of ground movements. The distributions of
-1500 -1000 -500 0 500 -5 0 5 10 15 20 25
0 0 movement with depth are difficult to estimate
5 5
without some form of analysis, as they depend
on wall flexibility and excavation support
10 10
conditions, but it may sometimes be adequate
Depth (m)

Depth (m)
15 15 to assume a linear distribution with depth
(Chen and Poulos, 2001).
20 20
Common design practice employs two-
dimensional analyses, and near the centre
25 25

30 30 of an excavation, two dimensional analyses

predicted measured predicted measured
(a) (b) can give reasonable soil movement estimates
Case 1 - Steel Tube
(for example, Yong et al, 1996). Thus, in
Bending moment (kN.m) Pile deflection (mm) the following examples, a two-dimensional
-400 -300 -200 -100 0 100 200 -5 0 5 10 15 20 25
0 0 analysis, employing the computer program
FLAC, has been used to estimate the ground
movements due to excavation for a pile cap.
5 5

10 10 The case examined is shown in Fig. 12, and

Depth (m)

Depth (m)

involves an excavation in medium-soft clay for

15 15
a 3 m deep pile cap, 10 m in width, with no
20 20
lateral support provided for the excavation.
Figs. 13 a-d show typical distributions of the
25 25 vertical and lateral movements with depth, at
predicted measured predicted measured
(c) (d) various distances from the excavation. Two
Case 2 - Reinforced Concrete Pile
different values of the surface pressure are
[Fig. 11] Calculated and Measured Pile Responses considered, 0 kPa (a green-field situation)
and 50 kPa, a typical situation that may arise
beneath an existing building. It can be seen
Piles Near a Pile Cap Excavation
that, as would be expected, the movements for
Introduction
the 50 kPa surface pressure are considerably
When an excavation is carried out for a new larger than those for zero pressure, and that
pile cap in the vicinity of existing piles, there the movements tend to decrease with increasing
is often little or no support provided for the distance from the excavation. It is further
excavation, since pile cap thicknesses are assumed that the excavation is carried out
typically 1-3 m and the excavation is therefore relatively rapidly, and that no drop in the level
relatively shallow. However, under conditions of the water table arises from the excavation.
in which the ground is highly stressed (for
example, within the footprint of an existing
building), even such modest excavations
deserve careful consideration as ground
movements will inevitably be generated by the
excavation process. In addition, it is possible
that dewatering may also be necessary, in
which case additional ground movements (both
vertical and lateral) will be generated by the
process of groundwater lowering.

Ground Movements
It is now common for the ground movements
around excavations to be estimated via detailed
numerical analyses such as the finite element [Fig. 12] Pile Cap Excavation near Existing Pile
method. When numerical analyses cannot be
carried out, it is possible to use approximations Pile Response to Ground Movements
developed by Clough and his co-workers to For the case as shown in Fig. 12, Figs. 14 and 15
estimate vertical and horizontal distributions summarize the computed maximum bending

[44] DFI JOURNAL Vol. 1 No. 1 November 2007

moment and shear in an adjacent pile, as a Movement (mm)
0 1 2 3 4 5 6 7
function of the distance from the excavation 0

and the surface pressure. It will be seen that -2

X=6 m
X=10 m
the induced maximum bending moment is very X=15 m
large when the pile is close to the excavation. -4 X=20 m
Indeed, for a 0.6 m diameter reinforced concrete

Depth (m)
-6
pile with 1% reinforcement, carrying a working
axial load of 800 kN (corresponding to a factor -8
of safety of about 2), the maximum design
moment capacity is about 0.56 kNm. Thus, -10

Fig. 14 implies that piles within about 10 m of -12

Excavation Depth = 2.5 m Zero
Surface Pressure
the axis of the excavation could have induced
moments that exceed the design capacity of the [Fig. 13(c)] Computed Horizontal Ground Movements Due to
pile, if the surface pressure is 50 kPa. Cap Excavation p0 =0

Movement (mm)
0 5 10 15 20 25 30 35
0
Movement (mm)
00 1 2 3 4 2

-2 4

Depth (m)
6
-4 x=6 m
8 x=10 m
x=20 m
-6
Depth (m)

10

12
-8
14
-10 X=6 m
X=10 m [Fig. 13(d)] Computed Horizontal Ground Movements Due to
Excavation Depth Cap Excavation p0 = 50 kPa
-12 X=15 m
= 2.5 m Zero
Surface Pressure X=20 m
1
-14
Maximum Induced Moment (MNm)

[Fig. 13(a)] Computed Vertical Ground Movements Due to 0.8

p=0
Cap Excavation p0 =0
p=50 kPa
0.6

0.4
Movement (mm)
-5 0 5 10 15 20
0 0.2

2
0
4 0 5 10 15 20 25
Distance of Pile Axis from Excavation Axis (m)
6
Depth (m)

x=6 m [Fig. 14] Computed Pile Moment Due to Cap Excavation

8 x=10 m
0.7
x=20 m
10 0.6

12 p=0
0.5
Maximum Shear (MN)

p=50 kPa
14 0.4
[Fig. 13(b)] Computed Vertical Ground Movements Due to
0.3
Cap Excavation p0 =50 kPa
0.2

0.1

0
0 5 10 15 20 25
Distance of Pile Axis from Excavation Axis (m)

[Fig. 15] Computed Pile Shear Due to Cap Excavation

DFI JOURNAL Vol. 1 No. 1 November 2007 [45]

 Fig. 16 summarizes the computed additional measured for different depths of excavation. The
movement of an existing pile adjacent to the measured free-field soil movements at different
excavation. In this case, if there is zero surface distances from the wall and corresponding to
pressure, the adjacent pile tends to move an excavation depth of 4.5 m were found to
upwards slightly because of the excavation, decrease almost linearly with depth, from a
but it settles if the surface pressure is 50 kPa. maximum value at the surface.
In the latter case, the additional axial force The generic elastic design charts (Chen and
induced in the pile by the vertical ground Poulos, 1997) were used to calculate the pile
settlement is small, even if the pile is relatively response, with the soil movements simplified
close to the excavation. to linear profiles. The estimated results are
Thus, it would appear that the issue that may shown in Fig. 17, together with those measured.
cause most concern is the induced bending It can be seen that, in this case, the elastic
moment and shear in the pile due to the lateral generic design charts using the simplified soil
ground movements. movements generally give fairly good estimates
of the pile response.
2.5
300
Additonal Vertical Movement (mm)

2 Estimated by design charts

Pile-1
Measured

Maximum Bending Moment (kN.m)

250
Measured
1.5
p=0 200

Test PC
1
p=50 kPa

Test PC
150

Pile-2
0.5

Test PC
100

Test PC

Test PC
0
50
-0.5
0
0 2 4 6 8 10
-1 Distance from Retaining Wall (m)
0 5 10 15 20 25
Distance of Pile Axis from Excavation Axis (m)
40
[Fig. 16] Computed Vertical Movement of Pile due to Cap Estimated by design charts
Exacavation Measured
Maximum Deflection (mm)

30
Test PC

Comparisons Between Computed and

Measured Pile Response in Centrifuge Tests
Test PC

20
Test PC

There do not appear to be any measurements of

Test PC5
Test PC

pile response to cap excavation-induced ground 10

movements. However, Leung et al (2000) have

presented results from centrifuge model tests 0
0 2 4 6 8 10
on a single pile adjacent to unstrutted deep Distance from Retaining Wall (m)

excavations in dense sand. The model pile was [Fig. 17] Measured and Computed Pile Moments due to
fabricated from a hollow square aluminum tube Excavation
and instrumented with 10 pairs of strain gauges
protected by a thin layer of epoxy. The model Response of Piles to Seismically-induced
pile simulated a prototype concrete bored Ground Motions
pile of 0.63 m in diameter, and 12.5 m in total
embedded length. The retaining wall supporting Introduction
the excavation had an embedment depth of Many pile failures during earthquakes have
8 m. The Youngs modulus of the sand, Es, was occurred due to the inadequacy of the pile to
estimated to increase linearly with depth, z (in withstand large induced moments and shears
metres), and expressed approximately as Es = (Mizuno 1987, 1996). The pile designer (who
Nhz = 6z MPa. often is not the person who performs structural
Several tests were carried out in which the pile dynamic analysis) needs to know approximately
was located at different distances from the what the maximum seismically-induced internal
retaining wall. The free-field soil movements, moment and shear of the pile would be. This
pile bending moments and deflections were problem is similar in principle to the problems

[46] DFI JOURNAL Vol. 1 No. 1 November 2007

of statically-induced ground movements 4. The maximum values of the displacements
discussed above, but there are additional along the pile obtained in step 3 are treated
complexities that must be recognized. In as a static soil movement profile, although
particular, there are two sources of loading of the displacement at each point may have
the pile by the ground movements: occurred at different times.
1. Inertial loading at the pile head, caused by 5. The surface motion obtained in step 3 is
the lateral forces imposed on the structure used in an ordinary spectral analysis of a
by the earthquake and which are then single degree of freedom system whose
imposed on the piles; period is equal to the period obtained in step
2. Kinematic loading along the length of the 2. The spectral acceleration is calculated.
pile, caused by the lateral (and to a lesser 6. The lateral force to be applied to the pile is
extent, the vertical) ground movements obtained from multiplication of the spectral
developed by the earthquake. acceleration obtained in step 5 and the
Traditionally, many foundation designers have cap-mass or the mass of the single degree
considered only the effects of inertial loading, but of freedom model of the superstructure
kinematic loading can also be very important. calculated in step one.
A relatively simple pseudo-static approach for 7. A static analysis, in which the pile is subjected
analyzing this problem, taking into account to the simultaneous application of a lateral
both inertial and kinematic effects, has been force at its head equal to the force obtained
developed by Tabesh and Poulos (2001) and is in step 6, and a soil movement profile formed
described below. in step 4, is performed and the maximum pile
moment and shear are obtained.
Pseudo-Static Approach
Verification of The Pseudo-Static Method
The pseudo-static methodology developed by Tabesh
and Poulos (2001) involves the following steps: In order to examine the performance of the
1. The superstructure is modelled by a single proposed pseudo-static methodology, Tabesh
degree of freedom system whose natural and Poulos (2001) considered a soil mass
frequency is equal to the fundamental consisting of two layers. Various ratios of
frequency of the superstructure. The layer stiffness were considered, and a range
simplest way is to reduce the superstructure of pile diameters was analysed. The Newcastle
to a cap-mass. The designer should be careful 1994 earthquake was used as the excitation
about such a crude approximation; the mere source. Eighty different pile-soil configurations
eccentricity of the superstructure mass may were considered for which the envelopes
have a profound effect on the response. of the positive and negative moment and
shear along the pile were obtained via a more
2. The natural lateral period T of the pile head
complete dynamic analysis, and the shear and
is estimated through published formulae
moment distributions along the pile were also
and charts, or in the case of a pile-cap-mass
calculated from the proposed pseudo-static
system, via the approximate relationship:
analysis with the maximum computed free-field
T = 2 (Cap Mass/ Kx) 0.5
(5) soil movements as input. The cap-mass (and
where Kx = lateral pile head stiffness. hence the vertical applied load) was assumed
to be zero. Without any exception, excellent
3. A free-field site response analysis is
agreement was obtained between the dynamic
performed to obtain both the time history of
analysis and the pseudo-static methodology.
the motion at the surface and the maximum
displacement of the soil mass along the These comparisons suggest that, regardless
pile. In this analysis the well-known SHAKE of the soil non-homogeneity, the static
(Schnabel et al (1972)) program, or similar methodology gives very good results for the
computer codes such as the ERLS program maximum values of the moment and shear
used herein, may be used. As the moment along the pile. Thus, when the response of
and shear depend on the curvature of the pile is dictated by the free-field ground
the pile, the points whose maximum movements, the internal response of the pile
displacements are to be obtained must be can be easily estimated by a very simple static
closely spaced, especially near the surface. analysis. When the effects of cap-mass were

DFI JOURNAL Vol. 1 No. 1 November 2007 [47]

 taken into account, it was observed that, while 1. If kinematic effects are ignored, and only
in many cases the agreement between the inertial (lateral load) effects are considered,
proposed pseudo-static method and dynamic the maximum moment at the pile head can
analysis was close, while in some others the be seriously under-estimated.
pseudo-static approach overestimated the
2. If only inertial effects are considered,
maximum moment and shear.
the moment at depths in excess of about
Significance of Inertial and Kinematic Effects 7m becomes insignificant, but with the
kinematic effects incorporated, there is
The influence of inertial effects (via vertical loading
a significant moment between depths of
and/or cap-mass) on the seismic response of pile about 7 to 10m, i.e. in the vicinity of the
foundations depends on the frequency content of interface between the softer upper layer and
the earthquake and the natural period of the pile- the stronger lower layer.
soil-cap-mass system. Mylonakis et al (1997) have
identified the following characteristics: The importance of considering both kinematic
as well as inertial effects is clearly emphasized
1. Inertial bending can be significant, in this example.
especially in the upper part of the
piles, when the dominant period of the DEPTH c u Go Do fs py P
(m)
earthquake is similar to the fundamental 0
kPa MPa kPa kPa MPa

period of the soil-pile-structure system. 0.14

SOFT-
2. Kinematic bending can be significant when MEDIUM 35 35 .005 35
5 CLAY 0.315
the dominant period of the soil motions are
similar to the natural period of the soil strata.
10
3. The three most likely areas of damage of
a pile are the pile head, interfaces between STIFF- d=0.75m
layers of different stiffness, and the pile 15 HARD 180 180 .005 100 1.08
CLAY
toe. Pile head damage is most likely in
homogeneous strata while damage at strata
20
interfaces is most likely when there is a
marked stiffness contrast between the layers. SHALE
BEDROCK 2000 2000 .005 - -
The kinematic bending strains at the pile toe 25

may be significant when the toe is restrained.

[Fig. 18] Example Analyzed
To facilitate an understanding of the relative
importance of inertial and kinematic effects, Moment (MNm)
-0.4 -0.3 -0.2 -0.1 0 0.1 0.2 0.3
analyses have been performed on the fixed head 0

single pile shown in Fig. 18. The analysis has 2

been carried out via the pseudo-static approach

4
described above, so that the results provide an
envelope of maximum bending moments and 6

shears along the pile. It is assumed that the site

Depth (m)

is subjected to the 1994 Northridge earthquake

10
with a maximum bedrock acceleration of 0.2 g.
Three cases have been considered: 12

A pile with no vertical load/cap mass; 14

A pile with a lateral inertial load of 0.2 MN 16

A pile with the same lateral inertial load as Kinematic effects only
18

in the second case, but where the kinematic Inertial Effects Only
20

ground movements are not included in (H=0.2 MN)

the analysis. Kinematic + Inertial
22

Effects
Fig. 19 shows the computed distributions of
bending moment along the pile. Two key points [Fig. 19] Effects of Kinematic and Inertial Loading of Pile on
Moment Distribution
emerge from this figure:

[48] DFI JOURNAL Vol. 1 No. 1 November 2007

Design Charts 6
Ep=10000 MPa, FS=2, L=20 m
Tabesh and Poulos (2006) have attempted to 5
Es=25 MPa
provide a simple means of making preliminary

Mom. (MN-m)
4
Es=50 MPa
estimates of maximum bending moment and
3 Es=100 MPa
shear in single piles embedded in homogeneous
clay layers and subjected to seismic excitation 2

(Fig. 20). To develop these design charts, a time 1

domain method has been employed in which
0
the earthquake motion has been input, and 0 0.5 1 1.5
the moment, shear and relative displacement d (m)
of the pile obtained at all time steps during
earthquake. The maximum values at any time 2.5
Ep=10000 MPa, FS=2, L=20 m
during the earthquake have then been extracted
2
and used for the design charts.

Shear (MN)
1.5
P
1

0.5
Clay layer modulus=Es
0
0 0.5 1 1.5
H L d (m)
Pile modulus=Ep
[Fig. 21] Typical Design Charts for Maximum Moment and
Shear Induced in a Pile.

near Tokyo. The seismic observations at this

bridge and one of its pile foundations were
Earthquake excitation conducted between 1981 and 1985 by Shimizu
[Fig. 20] Idealized Case for Design Charts Corporation. During this period, 14 earthquakes
hit the region. Among them, the twelfth
Eight earthquakes were selected to develop earthquake induced the largest peak horizontal
the design charts, covering a wide range surface acceleration which was 0.11 g.
of prominent earthquake frequencies. The The bridge was supported by 11 piers and is
piles were assumed to have a fixed head (i.e. 484.8 m long and 10.75 m wide. The girder is
zero head rotation), and to be embedded continuous from pier 5 to pier 8. Piers 5, 7, and 8
in homogeneous clay layers with Youngs are equipped with movable bearings, and pier 6 is
modulus values of 25, 50 and 100 MPa. The pile of the fixed-shoe type. Fig. 22 shows the details of
diameters covered the range of values between the bridge and the pile foundation of pier 6 where
0.2 to 1.5 metres in 0.1 m increments. Two pile the strain metres were installed. The soil profile
modulus values, 10000 MPa and 30000 MPa shown in Fig. 22 was obtained from a borehole
were considered, being representative of timber near pier 6. The top soil layers were extremely
and concrete respectively. soft with SPT-N values being almost zero. The
Fig. 21 shows typical design charts for the shear wave velocity of the top layers was between
maximum moment and shear developed in a 40-100 m/s. The tips of the piles were embedded
20 m long pile with an axial load equivalent to a in a stiff layer of clay with a shear-wave velocity
factor of safety of 2. As would be expected, both of 400 m/s. The length of the vertical piles was
the moment and shear increase with increasing 22 m, with 2 m in the stiff clay.
pile diameter d. The piles were steel pipes with a diameter of
0.6 m; the thickness of the vertical piles was 9
Case Study mm, and the wall thickness of the battered piles
The pseudo-static methodology is used to was 12 mm. A total of 32 wall strain metres were
estimate the maximum moment developed in installed on one vertical and one battered piles
the Ohba-Ohashi bridge in Japan. This bridge is at four different elevations (sa1-sa4, sb1-sb4 in
located in Fujisawa city, Kanagawa prefecture, Fig. 22). A total of 11 units of accelerometers of

DFI JOURNAL Vol. 1 No. 1 November 2007 [49]

 servo type were installed, with one unit (GS1) at
the ground surface, four units (GB1-GB4) at the
bearing substratum, three units (BS1-BS3) at the
footings, two units (BR1 and BR3) at the piers
and one unit (BR2) at the girder.

Organic
silt

Clay

1.5m

Battered Pile
1.5m

Vertical Pile [Fig. 23] Pile Model used in Analysis

Vertical pile with strain gauges

12m

With these assumptions the pseudo-static

Battered pile with strain gauges
approach was used to obtain the response of
the pile to the combination of the following
disturbances:
a soil movement profile formed from the
12m
maximum free-field soil movement obtained
[Fig. 22] Ohba-Ohashi Bridge Pile Foundation.
at each pile element,

For this analysis, the eighth earthquake was a head force equal to F= ms. Aspec in which
considered and the pseudo-static methodology ms = mass of superstructure and Aspec is the
was used to estimate the maximum moment spectral acceleration obtained for a period
developed in the vertical instrumented pile. equal to that of the pile and based on the
Group effects were ignored. In the Ohba- surface motion,
Ohashi bridge measurements the free-field a moment equal to M=F.e in which e is
displacements along the length of the pile were the eccentricity.
not measured and only the base and surface The profile of the moment along the pile
motions were monitored. A free-field analysis obtained from the pseudo-static method, along
was therefore required for which ERLS program with the maximum moments measured at 4
was employed and the motion monitored at locations along the pile are shown in Fig. 24.
GB1 was used as the input motion. The soil
was modelled as a system of 7 horizontal
layers, and the mass of the superstructure was
concentrated in two points, as shown in Fig. 23.
The maximum value of the free-field response
was obtained at 48 points corresponding to
the centre of 48 pile elements. The spectral
acceleration corresponding to the pile natural
period and based on the surface motion was
calculated to be 0.092 m/s2. All piles were
assumed to carry equal loads. The pier was
very stiff and was considered to be rigid. The
[Fig. 24] Comparison of Calculated and Measured Moments
eccentricity was calculated to be 16.3 m. Along a Vertical Pile in the Ohba-Ohashi Bridge Foundation.

[50] DFI JOURNAL Vol. 1 No. 1 November 2007

In this figure the envelope of the positive and 1) piles near an embankment.
negative moment along the pile obtained from 2) piles near a pile cap excavation.
an independent dynamic analysis (SEPAP) is also
3) piles in ground subjected to seismic action.
shown (Tabesh and Poulos, 2001).
Some of the features of behaviour have been
The proposed static methodology gives good
discussed, and some typical design charts
estimates of the measured values, despite the
have been presented to assist in the practical
fact that the Ohba-Ohashi bridge was very
estimation of the forces, moments and
complicated and was over-simplified for the
displacements induced in the piles by ground
analysis. Similar results were obtained for several
movements.
other measured earthquakes (Tabesh, 1997).
Examples of comparisons between theoretical
It is believed that the pseudo-static analysis
and measured behaviour are described, and
can be used by practicing engineers to obtain a
reasonable estimate of the maximum moment these generally show a reasonable measure
and shear which is likely to be developed in a pile of agreement.
when it is subjected to earthquake excitation. Attention has been concentrated on single
This analysis is likely to give good results in piles, but it has been found that, in general, the
many practical cases, but it may overestimate the consideration of a single pile is conservative from
maximum moment and shear in certain other a design viewpoint, as interaction among piles in
cases, especially when the period of the pile is a group subjected to ground movements usually
close to the natural period of the soil mass in has a beneficial effect and reduces the induced
which case significant interaction may occur deflections, forces and moments as compared
between the pile and soil. One reason is that the with a single isolated pile. In addition, in contrast
maximum free-field effects and maximum inertial to the case of piles subjected to direct head
effects have been assumed to act simultaneously, loading, elastic analyses tend to give conservative
which does not occur in a dynamic analysis. estimates of pile responses to ground movements
More importantly, the assumption behind using because there is no limit to the pile-soil pressures
spectral acceleration is that the cap-mass is developed by the moving soil.
excited by the surface motion. In reality, the cap-
mass is excited not through the soil surface, but Acknowledgements
via the pile head whose motion is different from Much of the work described in this paper was
the surface motion. The pile head and surface supported by grants from the Australian Research
motions are very close for a homogeneous soil Council. The author gratefully acknowledges the
mass, but when the soil is strongly layered, the major contributions of a number of researchers
pile head motion is often less severe. who worked on these projects, including Drs. L.T.
Chen, N. Loganathan, C.Y. Lee, K.R. Xu, A. Tabesh,
Conclusions T.S. Hull and J.C. Small.
There are several circumstances in which
ground movements may influence significantly References
the behaviour of piles. This aspect of pile 1. Amirsoleymani, T. (1991). Elimination of
behaviour has often been overlooked or excessive differential settlement by different
not recognized, often leading to excessive methods. Proc. 9th Asian Conf. Soil Mech.
foundation deformations and possible Found. Eng., Bangkok, 2: 351-354.
structural damage of the foundation system.
2. Bourges, F. and Mieussens, C. (1979).
This paper has described a consistent procedure
Dplacements lateraux proximit des
for analyzing the response of piles to ground
remblais sur sols compressibles mthode
movements, via simplified boundary element
de prvision. Bull. Liason Lab. P. et Ch, Vol.
analyses. For such analyses, it is necessary to be
101, 73-100.
able to predict the distribution of the free-field
ground movements, and then to use these in the 3. Chen, L.T. (1994). The effect of lateral soil
analyses together with the pile-soil parameters movements on pile foundations. PhD Thesis,
which are required for the normal analysis of The University of Sydney, Australia.
piles subjected to applied head loadings. 4. Chen, L. and Poulos, H.G. (1997). Piles
Three specific examples of piles subjected to subjected to lateral soil movements. Jnl.
ground movements have been considered: Geot. & Geoenv. Eng., ASCE, 123 (9), 802-811.

DFI JOURNAL Vol. 1 No. 1 November 2007 [51]

 5. Chen, L.T., and Poulos, H.G., (1996). Some surcharges on the soil around the piles.
aspects of pile response near an excavation. Proc. 5th European Conference on Soil
Proc. of 7th Australia and New Zealand Mechanics and Foundation Engineering, Vol.
Conference in Geomechanics, pp. 604-609. 1, 325-332.
6. Chen, L.T. and Poulos, H.G. (1997). Piles 17. De Cock, F. and Legrand, C. (1997). Design
subjected to lateral soil movements. Jnl. of of axially loaded piles European practice.
Geotechnical and Geoenv. Eng., ASCE, Vol. A.A. Balkema, Rotterdam.
123, No. 9, pp. 802-811. 18. Finno, R. J., Laurence, S.A., Allauh, N.F.
7. Chen, L.T. and Poulos, H.G. (1999). Design and Harahap, I.S. (1991). Analysis of
charts for analysis of piles subjected to performance of pile groups adjacent to deep
lateral soil movements. Proceedings of 8th excavation. Jnl. Geot. Eng., ASCE, Vol. 117,
Australia and New Zealand Conference in No. 6, 934-955.
Geomechanics, pp. 367 373, Edited by 19. Goh, A.T., Teh, C.I. and Wong, K.S.
Vitharana and Colman, Hobart, Australia. (1996). The response of vertical piles to
8. Chen, L.T., Poulos, H.G. and Loganathan, N. ground movements from adjacent braced
(1999). Pile responses caused by tunnelling. excavations. Proc. 12th Southeast Asian
Jnl. of Geot. and Geoenv. Eng., ASCE, Vol. Geot. Conf., Kuala Lumpur, 1: 403-409.
125, No. 3, pp. 207 215. 20. Heyman, L. (1965). Measurement of the
9. Chen, L.T., Poulos, H.G. and Loganathan, influence of lateral earth pressure on pile
N. (2000). Approximate design charts for foundation. Proc. Sixth Int. Conf. Soil Mechs.
piles adjacent to tunneling operations, Found. Eng., Montreal, Vol. 2, 257-260.
GeoEng2000 - International Conference 21. Heyman, L. and Boersma, L. (1961). Bending
on Geotechnical & Geological Engineering, moments in piles due to lateral earth
Melbourne, Australia, CD volume. pressure. Proc. Fifth Int. Conf. Soil Mechs.
10. Chen, L.T. and Poulos, H.G. (2001). Foun. Eng., Paris, Vol. 2, 425-429.
Approximation of lateral soil movements 22. Hull, T.S. (1996). PALLAS Users Guide.
for analyzing lateral pile response. Proc. Centre for Geotechnical Research, University
20th Annual Seminar of Geot. Divn., H.K. Inst. of Sydney.
Engrs, 14-23.
23. Kuwabara F. and Poulos, H.G. (1989).
11. Chow, Y.K., Chin, J.T. and Lee, S.L. (1990). Downdrag forces in a group of piles. Jnl.
Negative skin friction on pile groups. Int. Jnl. Geot. Eng., ASCE, 115 (6), 806-818.
Num. Anal. Meths. in Geomechs., 14 (1), 75-91.
24. Lee, C. Y., Poulos, H.G. and Hull, T.S. (1991).
12. Chow, Y.K. (1996). Performance of piles Effect of seafloor instability on offshore pile
used to stabilize landslides. Proc. 12th foundations. Can. Geot. Jnl., Vol. 28, No. 5,
Southeast Asian Geot. Conf., Kuala Lumpur, 729-737.
1:417-422.
25. Lee, R.G., Turner, A.J., and Whitworth, L.J.
13. Chu, Y.K. (1994). A failure case study of (1994). Deformations caused by tunneling
island method excavation in soft clay. Proc. beneath a piled structure. Proc. XIII
Int. Conf. Des. and Constrn. Deep Founds., International Conference on Soil Mechanics
Orlando, 3: 1216-1230. and Foundation Engineering, University
14. Clough, G W & Schmidt, B (1981), Press, London, 873-878.
Excavations and tunneling. In Soft Clay 26. Leung, C.F., Chow, Y.K. and Shen, R.F. (2000).
Engineering (ed. E W Brand and R P Brenner), Behaviour of pile subject to excavation-induced
Chap 8. Amsterdam: Elsevier. soil movement. Jnl.of Geot. and Geoenv.
15. Clough, G.W., Smith, E.M. and SAweeney, B.P. Engineering, ASCE, Vol. 126, pp. 947-954.
(1989). Movement control of excavation 27. Loganathan N & Poulos H G (1998),
support systems by interative design. Analytical prediction for tunneling-induced
Foundation Engineering: Current Principles ground movements in clays, Jnl. Geotech.
and Practices, Ed. F.H. Kulhawy et al, ASCE, Eng., ASCE, 124, No. 9, 846-856.
2: 869-884.
28. Loganathan, N. and Poulos, H.G. (1999).
16. De Beer, E.E., and Wallays, M. (1972). Tunnelling-induced ground deformations
Forces induced in piles by unsymmetrical

[52] DFI JOURNAL Vol. 1 No. 1 November 2007

 and their effects on adjacent piles. Proc. 10th 42. Poulos, H.G. (1989). Pile behaviour-Theory and
Aust. Tunnelling Conf., Melbourne, 241-250. Application. Geotechnique, 39 (3), 365-415.
29. Loganathan, N. et al. (2000). Centrifuge 43. Poulos, H.G. (1994). Analysis and design
model testing of tunnelling-induced ground of piles through embankments. Proc. Int.
and pile deformation. Geotechnique, 50(3): Conf. Des. and Constrn. of Deep Foundns.,
283-294. Orlando, 3, 1403-1421.
30. Loganathan, N. et al (2001). Ground and 44. Poulos, H.G. (1996). A comparison of some
pile-group responses to tunnelling. Soils methods for the design of piles through
and Foundations, 41(1): 57-67. embankments. Proc. 12th Southeast Asian
31. Loganathan, N. and Poulos, H.G. (2002). Geot. Conf. , Kuala Lumpur, 1: 447-452.
Centrifuge modelling: tunnelling-induced 45. Poulos, H.G. (1997). Failure of a building
ground movements and pile behaviour. Proc. supported on piles. Proc. Int. Conf. on
ITA Conf., Sydney, Australia, CD volume. Foundation Failures, Singapore, IES, 53-66.
32. Mana, A. and Clough, G.W. (1981). 46. Poulos, H.G. (2001). Pile foundations. Ch.
Prediction of movement of braced cuts. 10 of Geotechnical and Geoenvironmental
Jnl. Geot. Divn., ASCE, 107: 759-778. Engineering Handbook, Ed. R.K. Rowe,
33. Marche, R. and Lacroix, Y. (1972). Stabilit Kluwer Press.
des cules de ponts etablis sur des pieux 47. Poulos, H.G., and Chen, L.T. (1996a). Pile
traversant une couche molle. Can. Geot. Jnl. response due to excavationinduced lateral
Vol. 9, 1-14. soil movement. Jnl. Geot.Eng., ASCE, Vol.
34. Mair, R J, Gunn, M J & OReilly, M P (1983), 123, No. 2, pp. 94-99.
Ground movements around shallow tunnels 48. Poulos, H.G., and Chen, L.T. (1996b). Pile
in soft clays, Proc. 10th Int. Conf. Soil Mech. response due to unsupported excavation
Found. Eng., 323-328. induced lateral soil movement. Can. Geot.
35. Mair, R J, Tailor, R N & Bracegirdle, A (1993), Jnl., Vol. 33, pp. 670-677.
Subsurface settlement profiles above 49. Poulos, H.G. and Chen, L. (1997). Pile
tunnels in clays, Geotechnique 43, 315-320. response due to excavation-induced lateral
36. Meyerhof, G.G. (1976). Bearing capacity and soil movement. Jnl. Geot. and Geoenv. Eng.,
settlement of pile foundations. Jnl. Geot. ASCE, 123 (2), 94-99.
Eng. Div., ASCE, 102 (GT3), 195-228. 50. Poulos, H.G. and Davis, E.H. (1980). Pile
37. Morton, J D & King, K H (1979). Effect of foundation analysis and design. John Wiley,
tunnelling on the bearing capacity and New York.
settlement of piled foundations. Proc. 51. Stewart, D.P., Jewell, R. J. and Randolph,
Tunnelling 79, (ed. M J Jones), pp. 57-58. M.F. (1992). Bridge abutments on soft clay
London:IMM experimental data and simple design
38. Nelson, J.D. and Miller, D.J. (1992). methods. Proc. 6th Aust-New Zealand Conf.
Expansive soils Problems and practice On Geomechanics, Christchurch, 199-204.
in foundation and pavement engineering. 52. Stewart, D.P., Jewell, R.J. and Randolph,
Wiley Interscience, New York. M.F. (1991). Embankment loading of piled
39. Peck, R.B. (1969). Deep excavations and bridge abutments on soft ground. Int. Conf.
tunneling in soft ground. State-of-the- On Geotechnical Engineering for Coastal
art Report, Proc. 7th Int. Conf. Soil Mechs. Development, Yokohama, JSSMFE, 741-746.
Found. Eng., Mexico City, 3: 225-290. 53. Yong, K.Y., Lee, F.H. and Liu, X.K. (1996).
40. Poulos, H.G. (1972). Difficulties in Three dimensional finite element analysis of
prediction of horizontal deformations of deep excavations in marine clay. Proc. 12th
foundations. Jnl. Soil Mechs. Founds. Divn., Southeast Asian Geot. Conf., Kuala Lumpur,
ASCE, 98(SM8): 843-848. 1: 435-440.

41. Poulos, H.G. (1982). Developments in the 54. Xu, K.R. and Poulos, H.G. (2001). 3-D
analysis of static and cyclic lateral response elastic analysis of vertical piles subjected
of piles. Proc. 4th Int. Conf. Num. Meth. to passive loadings. Computers and
Geomechs., Edmonton, 3, 117-1135. Geotechnics, 28: 349-375

DFI JOURNAL Vol. 1 No. 1 November 2007 [53]