Missouri University of Science and Technology

Scholars' Mine
International Conferences on Recent Advances 1991 - Second International Conference on
in Geotechnical Earthquake Engineering and Recent Advances in Geotechnical Earthquake
Soil Dynamics Engineering & Soil Dynamics

15 Mar 1991, 9:00 am - 10:00 am

Piles Under Dynamic Loads

M. Novak
The University of Western Ontario, Canada

Follow this and additional works at: [Link]

Part of the Geotechnical Engineering Commons

Recommended Citation
Novak, M., "Piles Under Dynamic Loads" (1991). International Conferences on Recent Advances in
Geotechnical Earthquake Engineering and Soil Dynamics. 12.
[Link]

This work is licensed under a Creative Commons Attribution-Noncommercial-No Derivative Works 4.0 License.

This Article - Conference proceedings is brought to you for free and open access by Scholars' Mine. It has been
accepted for inclusion in International Conferences on Recent Advances in Geotechnical Earthquake Engineering
and Soil Dynamics by an authorized administrator of Scholars' Mine. This work is protected by U. S. Copyright Law.
Unauthorized use including reproduction for redistribution requires the permission of the copyright holder. For more
information, please contact scholarsmine@[Link].
 Proceedings: Second International Conference on Recent Advances In Geotechnical Earthquake Engineering and Soli Dynamics,
March 11-1.5, 1991 St. Louis, Missouri, Paper No. SOA14

Piles Under Dynamic Loads

M. Novak
Professor of Civil Engineering, The University of Western Ontario,
London, Ontario, Canada

SYNOPSIS The paper deals with some of the more recent developments in pile dynamics. It reviews
the progress in the analysis of single piles and pile groups, field as well as laboratory experi-
ments and soil-pile-structure interaction. The influence of pile-soil interface is discussed and
extensive references are given.

INTRODUCTION approaches in order that we may examine the

differences among them and summarize what can be
Piles have been used for hundreds of years but learned from experiments and field observations.
the last twenty years or so have seen a
remarkable increase in interest in pile dynamics. There are different dynamic loads that can act
There are a few reasons for this: good sites on piles: earthquake forces, wave forces, wind
which do not require piles are getting scarcer forces, machine unbalances etc. Here, the
and thus piling is used more widely; new emphasis is primarily on conditions relevant to
important areas of application have emerged, for earthquake loading. Dealt with are properties
example offshore towers and nuclear powerplants; and behavior of single piles and pile groups,
piles have repeatedly failed in earthquakes or interaction with the cap, pile experiments, pile-
were damaged; and finally, dynamics of shallow structure interaction and a few other topics.
foundations has reached a point of satisfactory The subject of pile dynamics received a
understanding thus shifting research interests to comprehensive treatment in the state-of-the-art
less understood foundation types. The aim of the report by Taj imi ( 1977) , covering developments up
studies is to increase the safety of the piles to 1977, and in a few special volumes, i.e. De
and the structures they support and to better Beer et al. (1977), O'Neill and Dobry (1980),
understand the interaction between the piles and Nogami (1987) and Prakash and Sharma (1990). A
the structures under both critical and number of papers on piles were presented to this
operational conditions. conference. These are listed together at the end
of the References. Among the special areas of
The damage to piles may result from a few causes pile dynamics not considered here are integrity
such as vibration effects, liquefaction, and testing and pile driving. Recent data on these
embankment movements. A comprehensive survey of subjects can be found in Fellenius (1988). So
pile damage during earthquakes in Japan was many papers have been published on pile dynamics
presented by Mizuno (1987) but damage to piles since Tajimi's (1977) state-of-the-art report
also occurred in the Alaska earthquake of 1964, that it is impossible to refer to all of them in
the Mexico City earthquake of 1985 and the Lorna this report of limited scope. The author trusts
Prieta earthquake of 1989. that the readers will understand this.
Pile behavior is, of course, very complex and
this might have lead Terzaghi and Peck (1967) to SINGLE PILES
state that " ... theoretical refinements in dealing
with pile problems ... are completely out of place The earliest systematic, theoretical studies of
and can be safely ignored". dynamic soil-pile interaction are due to Parmelee
et al. (1964), Tajimi (1966), Penzien (1970),
Fortunately, not everybody got discouraged by Novak (1974) and a few others. Parmelee (1964)
this pessimistic evaluation and a number of and Penzien (1970) employed a non-linear discrete
analytical and numerical approaches to the model and a static theory to describe the dynamic
analysis of pile dynamic behavior have been elastic stress and displacement fields. Tajimi
developed. These approaches provided a much (1966) used a linear viscoelastic stratum of the
sounder theoretical basis for pile design than Kelvin-Voigt type to model the soil and in his
the equivalent cantilever concept or other purely analysis of the horizontal response neglected the
empirical methods which dominated the field for vertical component of the soil motion. Novak
decades. Nevertheless, some differences between (1974) assumed linearity and an elastic soil
the various theoretical approaches exist and the layer composed of independent infinitesimally
experiments reported are sometimes inconclusive; thin horizontal layers extending to infinity.
also, some uncertainties are inevitable when
applying an idealized theory to field conditions. The different approaches formulated and the data
Thus, it may be useful to review some of the they yield are briefly discussed below.

2433
Single Piles in Homogeneous Soil to generate a harmonic motion with a unit
The analytical approaches treat the interaction amplitude in' the specified direction, as is
between the pile and soil, schematically depicted schematically depicted in Fig. 1 for the case of
in Fig. 1, in terms of continuum mechanics. The horizontal impedance. The complex stiffnesses
can be expressed in any of the following ways,
i.e.

(1a)

k + iwc (1b)

k(k' + iwc') (1c)

in which K1 and K2 are the real and imaginary

parts of the complex stiffness, respectively
and i = j-1; k = K1 = true stiffness, c = K2 /w
= coefficient of equivalent yiscous damping,
and w = circular_frequency; k = static stiff-
ness and k' = k/k, c' = cjk' = dimensionless
stiffness and damping constants. All the
Figure 1 Schematic of soil-pile interaction parameters in Eqs. 1 depend on frequency w or the
dimensionless frequency a 0 = r 0 w/V1 where ro =
pile radius and V = soil shear wave velocity.
problem is very difficult to solve, even for the An example of the horizontal impedance of
idealistic assumptions of linear elasticity or endbearing piles is shown for two soil/pile
viscoelasticity, homogeneous soils and the pile stiffness ratios in Fig. 2. In this figure, V0
being welded to the soil. Thus, approximate = primary wave velocity in the pile, L = pile
procedures were formulated first. Tajimi's length, and v = Poisson's ratio; D = 2P = soil
(1966) solution of the horizontal response of an material damping with p = soil material damping
endbearing pile in a homogeneous layer, the first ratio and p = ratio of the soil specific mass to
of its kind, neglected the vertical component of pile specific mass. The depressions visible in
the motion. In 1974, Novak formulated a very Fig. 2a practically disappear for higher soil
simple approach based on plane strain soil
reactions, which can be interpreted as dynamic
Winkler medium or a plane strain, complex
transmitting boundary placed directly to the
pile. This boundary is similar to the standard 1. 2 I.R:;;:E:;;AL~P;;:A-;::RT:;--------------.
viscous boundary but is frequency dependent and
complex, i.e. it has a stiffness part in addition
to the damping part. This solution identified 0.8
dimensionless parameters of the problem, yielded
a number of design charts and tables for dynamic
stiffness and damping of piles, and indicated the 0.4
effect of the pile static load on the horizontal
pile characteristics. Material damping was later
included in closed form expressions for the soil
reactions in Novak et al. (1978). The applica-
tion of the same approach to vertical response of "'
Ia (a)
floating piles (Novak, 1977) indicated great it -0.4 '------------------ -..l...J
sensitivity of the pile behavior to tip condition ~
and showed that floating piles generate more ~ 1.6 r-::-:--:-------------,----~
radiation damping but less stiffness than ;'! IMAGINARY PART
endbearing piles. Torsional response was also "' L = 30
examined in this way (Novak & Howell, 1977, 1978) 1.2 r
0
and the importance of material damping for this
vibration mode was demonstrated.
0.8 ~· 0.5
A somewhat more rigorous solution, similar to D= 0.02
that of Tajimi (1966), was formulated by Nogami p= 0.6
and Novak (1976) for the vertical response and 0.4
for the horizontal response by Novak and Nogami
(1977). These approximate solutions offered a
basic insight into the behavior of the soil-pile
system.

Much of the attention is focussed on the pile FREQUENCY

complex dynamic stiffnesses (impedance functions) NATURAL FREQUENCY OF PILE ALONE
because they have a strong influence on the
response of pile supported buildings and
structures. The impedance functions are defined Figure 2 Horizontal impedance of endbearing pile
as amplitudes of harmonic forces (or moments) for two soil/pile stiffness ratios (Novak &
that have to be applied to the pile head in order Nogami, 1977)

2434
material damping as is schematically depicted in continuum were formulated by Sen et al. (1985)
Fig. 3 in which the differences in impedance and Pak and Jennings (1987). All vibration modes
were investigated by Rajapakse and Shah (1987a,b,
1989) . The latter authors evaluated the accuracy
of some of the existing solutions and concluded
NO OAMPING that continuum models based on harmonic line
loads may not be accurate enough and generated an
extensive set of charts for impedances of
floating piles.
The finite element solutions were formulated by
- 1st LAYER RESONANCE Kuhlemeyer (1976, 1979a,b), Blaney et al. (1976),
Wolf and von Arx (1978), Waas and Hartmann
(1981), Sanchez-Salinero (1982) and others.
Boundary element approaches were developed by
Banerjee (1978), Banerjee and Sen (1987) and a
"'E few others. Ready to use charts and formulae
were produced for homogeneous soils by Kuhlemeyer
"'z (1979a,b), Roesset (1980), Dobry et al. (1982),
" Novak and El Sharnouby (1983) and a few others.
""'
0
Thus, a considerable amount of data on piles in
linear, homogeneous media is available. Some
differences in these data exist but from the
practical point of view, they agree reasonably
well.
Since the pile dynamic stiffnesses for low
frequencies are usually quite close to static
stiffnesses, it may be useful to examine the
differences in pile static stiffness. For axial
loading such a comparison is shown in Fig. 4 in
which K' = Kr 0/~Es is dimensionless static
stiffness, K = k is true static stiffness and A~
= pile cross-sectional area: E , E5 = pile ana
FREQUENCY a0 soil Young's modulus, respect'lvely. Fig. 4
indicates that the individual authors' results
differ substantially, particularly for flexible
Figure 3 Impedances of endbearing pile for three piles or very stiff soils,i.e. small EpiE ratio;
cases of soil material damping for endbearing piles even an illogical t:end for
the stiffness to increase with pile length, L,
may be noticed (Fig. 4a). Inaccuracies of this
functions are emphasized for three cases of soil type result primarily from the small number of
material damping, i.e. no material damping, elements used in pile discretization. In the El
hysteretic (frequency independent) rna ter ial Sharnouby and Novak (1990) analysis, fifty
damping and viscous material damping. Hysteretic elements or more were needed to eliminate the
material damping is more realistic. For floating upward trend visible in Fig. 4a.
piles, the role of soil material damping is much
smaller as the layer resonances are absent.
Single Piles in Nonhomogeneous Soil
A few interesting features of the pile impe-
dances follow from the theoretical solutions Comparing the results of experiments with
indicated in Figs. 2 and 3: pile dynamic theoretical predictions repeatedly showed that
stiffness varies little with frequency, except if the soil is assumed to be homogeneous, both
for very heavy piles or very weak soils for which pile stiffness and damping can be grossly
it diminishes with frequency in a parabolic overestimated (e.g. Novak & Grigg, 1976; Novak
manner and can even become negative; for & Sheta, 1982). An example of this is shown in
endbearing piles vibrating below the fundamental Fig. 5 in which the theoretical response was
frequency of the soil layer, the geometric calculated with two constant values of shear wave
damping is absent because no progressive waves velocity: the value Vt, established from a wave
are generated in an elastic medium, just as with propaga~ion experiment (curve A) and a much lower
shallow foundations, leaving soil and pile value, v. = 0.26 Vt backcalculated from measured
material damping as the only sources of energy static deflections (curve B). Such a reduced
dissipation. Apart from this low frequency value yields a better estimate of pile stiffness
region, a fully embedded slender pile, not (resonance frequency) but does not allow a satis-
supporting any additional mass, is usually factory prediction of radiation damping and thus
overdamped and consequently does not exhibit any resonant amplitude. The reasons for the
marked resonance peak in dynamic tests. deficiencies of the theory based on the assump-
tion of soil homogeneity are schematically
More rigorous solutions, not neglecting one depicted in Fig. 6. They are the variation of
component of the motion, followed. For the soil shear modulus with depth, particularly its
horizontal response of an infinitely long pile, reduction toward ground surface which results
Kobori et al. (1977) obtained a solution in the from the diminishing confining pressure, and pile
form of infinite series of multiple integrals. separation from the soil or gapping. Single
More recent analyses based on the solution of piles under horizontal loading, as in Fig. 5, are
the governing equations of a three dimensional particularly sensitive to these two factors.

2435
 5.5

G(Z)

(a) (b)
4.5

(a)
4 ~--~25~---=5~0--~7~5~--~~o~o~~L-I~d~
z

-- -----·-
12 Floating Piles
v = 0.5 ~ - . Ep/E = 1000
..-· .-· 5 Figure 6 Schematic of pile separation and soil
10 /"
modulus reduction towards ground surface
- - - Present ( 50 Elements l
- - - Poulos ( 10 Elements)
Observations of this kind lead to the development
· - · - · - · Salinera ( > 20 Elements l
of approaches better suited for nonhomogeneous
X • X Rajapakse a Shah soils. A significant improvement in the finite

--- --- ---

6
element model was formulated by Roesset and his
Ep/E 5 = 100 co-workers (Blaney et al., 1976; Roesset &
4 Angelides, 1979) who placed the consistent,
frequency dependent boundary, derived by Kausel
et al (1975), directly to the pile or outside the
(b) cylindrical finite element zone around the pile.
z~--~25~----~5~0----~7~5----~~o~o~~
This approach was then used by Krishnan et al.
L/d (1983) and by Gazetas (1984) in their extensive
parametric studies.

Figure 4 Comparison of static axial pile Significant further progress was made by Kaynia
stiffness calculated by different authors for (1982a,b) and Kaynia and Kausel (1982, 1990) who
homogeneous soil: (a) - endbearing piles, (b) - based their solution of piles in generally
floating piles (Present data by El Sharnouby and layered media on the. formulation of displace-
Novak, 1990) ment fields due to uniformly distributed forces
on cylindrical surfaces (so called barrel load).
(This solution will be discussed in more detail
in the paragraph on pile groups.)

Banerjee and Sen ( 1987) presented ~ bounda:y

I
element solution for piles embedded 1n a sem1-
p
THEORETICAL, Vs VARIABLE infinite nonhomogeneous soil in which the soil
'/- II -
r L; fiA THEORETICAL. Vs • Vt
.
"· modulus, E5 , varies linearly with depth, z.
I II Banerjee and Sen's results suggest that, unlike
in layered soils, the frequency variations of the

,,,,
SYMBOL LB-IN
" II
.J 30
0 0.855 I "".'it~
·i:. () ,: 0 '1;
c.
.:>·..
impedance functions, normalized by static stiff-
UJ
0
0.598 ... 0 33 ness, are quite smooth and are affected very
::::>
1- II
0 0.427 little by soil nonhomogeneity. The actual magni-
::::; 0 0.171 tude of the stiffness and damping diminishes with

I ,,
a. II T£5T PILE 2
:::;:
<t 20
II
2r. • 2.41N. (6.1 em I
E5 (0), however.
(/) I • 0.000032 FT 4 ( 27.6 cm 4 1
(!)
w
II A few other methods suitable for linear generally
THEORETICAL, Vs" 0.26 Vt l;r. • 77.9
z-' 8 II layered media use a semi-analytical finite ele-
Q ment approach. These methods treat the wave pro-
II
~ 10 I I
w pagation in the horizontal direction analytically
:::;: I I and in the vertical direction employ finite ele-
0 I \
', ment idealization including auxiliary sublayers.
The pile is modelled by beam elements. One of
0~~~~~~~~~~
o m~m w w 60 the advantages of this approach is that it may
FREQUENCY (CPS) avoid the mathematical ill-conditioning resulting
from the large magnitude of Lame's constant, ~.
for soil Poisson's ratio, v, approaching 0. 5.
Solutions of this type were formulated by Tajimi
Figure 5 Comparison of experimental horizontal and Shimomura ( 1976) , Shimizu et al. ( 1977) ,
response of steel test pile with theoretical Waas and Hartmann (1981, 1984) and Mizuhata and
predictions (Novak and Sheta, 1982) Kusakabe (1984).

2436
An approximate analytical solution based on the
extension of the Novak and Nogami (1977) approach ......
·. G,p , v
was formulated for layered media by Takemiya and
Yamada (1981).
A much simpler and very versatile solution,
particularly well sui ted for high frequencies,
was formulated by Novak and Aboul-Ella (1978a,b)
who extended the plane strain approach to include
layered media and incorporated it in the code
PILAY. This code was used later by Novak and El
Sharnouby (1983) to generate design charts and
tables for parabolic soil profiles, as well as
homogeneous ones. With this approach, and
assuming a parabolic soil profile, with an
allowance for pile separation in the form of a
small free length, very satisfactory agreement
with the theory was obtained as indicated by
curve c in Fig. 5. Roesset et al. (1986) also Figure 1 Cylindrical boundary zone around pile
found the plane strain approach to work very well
for high frequencies. For very low frequencies, order to prevent wave reflections from the
an adjustment to the plane strain soil reaction fictitious interface between the cylindrical zone
is made for the vertical and horizontal and the outer region. These reflections occur
directions as discussed in Novak and El Sharnouby with nonzero weak zone mass, p., and result in
(1983) and implemented in the code PILAY. The undesirable undulations in bot'h stiffness and
plane strain approach works well for high damping of the composite medium. This is
frequencies because, in a layer, elastic waves exemplified in Fig. 8 in which a and ~ are
tend to propagate more and more horizontally as
the frequency increases, like in a wave guide.
The sensitivity of the response to pile separa-
tion and free length shows when evaluating most
experiments. The prediction of the separation
length is difficult and only empirical sugges-
tions can be made at this time. For small ampli-
tudes, 6, El-Marsafawi et al. (1990) observed the
following approximate relationship for pile
separation length, L8 :

Ls
d
260 ~ , 0.001 5 ~ 5 0.005 (2)

{b)
For larger displacements, a large separation ~~--------~--------~----~~~~------~
length may be needed (Han and Novak, 1988) . More 0 2 3
data on the separation effect will be given in Frequency 0; : r0 W/V;
the paragraph on nonlinear response.

As for possible deviations of the theoretical

assumptions from reality, pile deficiencies may Figure 8 Dimensionless vertical impedances of
also have a profound effect. This is shown by Wu composite medium with P; = p and tjr 0 = 1.0 (soil
et al. (1991) who, in their paper to this confe- damping ratio= 0.05)
rence, examine the influence of pile necking
using a combination of the BEM and FEM.
nondimensional stiffness and damping constants
Radial nonhomogeneity of the composite medium respectively. These
undulations can make the solution with P; 1' 0
While the consideration of a free separation actually less suitable for practical applications
length in the analysis may produce the reduction (Novak and Han, 1990). The difficulty with wave
in both pile stiffness and damping often observed reflections can be avoided by providing for a
in experiments, a better measure to this effect, continuous transition of stresses from the inner
or a complementary one, may be to account for zone to the outer region. Such a solution was
soil nonhomogeneity in the radial direction. A explored by Lakshmanan and Minai (1981), Dotson
simple way of doing this is to assume a weak, and Veletsos ( 1990) and Mizuhata and Kusakabe
cylindrical boundary zone around the pile (Fig. (1984). The latter authors found that even with
7). The zone is homogeneous and features a soil the weak zone, the experimental resonance
shear modulus, G., smaller than that of the outer amplitude measured on a 43.2 m long pile was five
zone and a large~ material damping. The purpose times larger than the theoretical value while the
of such a zone is to account in a very approxi- resonance frequency was predicted quite well.
mate way for soil nonlinearity in the region of This is consistent with other observations and
the highest stresses, pile separation, slippage indicative of the need to account for pile
and other deficiencies of the pile-soil separation.
interface. Such a zone was proposed by Novak and
Sheta (1980). In their plane strain solution, Wolf and Weber (1986) conducted a more rigorous
the mass of the boundary zone was neglected in study of the effect of soil tension exclusion,

2437
also assuming the circular cavity in the
unbounded thin lay~ (plane strain) . They found 1.1
that soil separatidn hardly affects horizontal Theory
1.3
stiffness, lS,, but reduces damping, ch, by more Experiment
than 50 per cent (Fig. 9b), a result quite 1.2
mee (kgmm)
similar to that of Novak and Sheta (1980). In ~ 1.1
addition, if shear is eliminated and hence 1.0 + 96
slipping is allowed in the zone of contact, <lJ
"0 0. ~

stiffness is also strongly reduced (Fig. 9c). In ....,:::> 0 171

o. a
Fig. 9, the linear case (a) indicates the 259
analysis with tension allowed. The size of the c..
E
o. 7-

contact area appears to be of 1 i ttle effect. <>: o. G

0. 5
Many other authors studied the interface ....,
behavior. Among the more recent ones are Mamoon "'
<:::
0. 1

(1990) and He (1990). However, when applying the 0

N
0. 3
various plane strain approaches to the interface, ,_ v. 2
the variation with depth is a problem for which 0
I: 0.1
very little guidance is available. 0.0

10 20 30
Linear Shear Slipping FREQUENCY <HZ>

,,,,,,,,.,,, ~
([)(D Figure 10 Theoretical and experimental
horizontal response of concrete pile for three
levels of harmonic excitation (El Marsafawi et
al., 1990)
,,,,,,,,~
QG) Penzien (1970), Matlock et al. (1978, 1980) and
a number of others, feature nonlinear springs,
nonlinear dampers, gaps and Coulomb friction
blocks. The combination of these elements makes
4.28 4.17 2.32 it possible to generate a variety of nonlinear
13.23 6.42 2.60 force-displacement relationships. An example of
the lumped mass model, formulated by Matlock et
(a) (b) (c) al. (1978, 1980) and incorporated in the code
SPASM, is shown in Fig. 11. Models of this type
are very versatile but it is difficult to relate
the characteristics of the discrete elements to
standard geotechnical parameters of soil. To
Figure 9 Effect of elimination of tension in help overcome this difficulty, various nonlinear
separation zone (Wolf & Weber 1986; a 0 =0. 629,
v=0.48)
/t
I I
Recognizing the separation effect and allowing I I 1 li
for it in an approximate way, a reasonable I I T1!
\ 1 Fluid
agreement between the theoretical results and \ \ Domp1nQ
experiments can be obtained. This is exemplified
, 'HI
in Fig. 10 comparing the theoretical and 1I
\
,,,, Reaction

.~..J
experimental responses of a concrete pile 7.5 m
in length and 0. 32 m in dia. The soil was
multilayered and a cylindrical weak zone was
assumed when calculating the response using the
code DYNA3. In this code, the weak zone is
IJ\l: I
~
analyzed as massless but its mass is added to
that of the pile in full or in part. Similar

+
tests and comparisons were reported by Gle
(1981), Woods (1984) and a number of others.

Nonlinear Response of Single Piles

The theories discussed thus far are essentially (b)
linear and thus quite adequate for small
displacements. At large displacements, piles
behave in a nonlinear fashion because of soil (a l
nonlinearity at high strain, pile separation
(gapping), slippage and friction. To incorporate
these factors into a continuum theory is
extremely difficult and therefore, lumped mass Figure 11 (a) Nonlinear lumped mass model of
models are most often used when nonlinear pile, (b) Observed cyclic reaction-deflection
analysis is required. Such models, employed by characteristics (Matlock et al., 1978)

2438
,;oil resistance-deflection relationships known as
p-y curves and t-x curves have been recommended
in the literature. For applications in offshore
structures, the American Petroleum Institute
(1986) specifies the p-y curves for clay as well
as sand making a difference between static
loading and cyclic loading. Extensive data on
the p-y curves and nonlinear pile response were
obtained by Yan (1990) using model piles and the
hydraulic gradient similitude method to reproduce
prototype conditions. An example of Yan's
results is shown in Fig. 12. Notice the
narrowing and partial linearization of the Stlakodown Incrqmczntal Collaps.Q
hysteresis loop with the number of cycles; this
trend increases with depth.

Figure 13 Pile stabilization (shakedown) and

incremental collapse under cyclic loading with
constant amplitude (Swane & Poulos, 1982)

point of view of the applicability of the various

linear theories for dynamic response analysis.
It explains why, with adequate adjustments
particularly for pile separation, such theories
may give reasonable results, as in Fig. 10, even
in cases where overall strong nonlinearity of the
response is clearly manifested.
Under vertical, steady state vibration, a similar
stabilization and partial linearization takes
place. Figure 14 shows the vertical displacement

1.5 , - - - - - - - - - - - - - - - - - - - - - - ,
Pile Deflection - y (mm)
• e8 •" s14
~ } Measured
o e. zs
- - - Colcuio!ed
E
Figure 12 Example of p-y curve under cyclic E - - - Bockbone (]
loading (Yan, 1990) 1.0
c
E
Cyclic loading is defined as repetitive loading e
0

with very low frequency so that no significant ~

inertia forces and radiation damping arise. It 0
provides basic insight into the material ~ 0.5
degradation due to soil plasticity and mechanic Q:;
degradation due to gapping associated with large >
displacements. Many studies were devoted to this /
/
subject but only a few may be mentioned here. 0.1 /
Trochanis et al. ( 1988) found theoretically a
dramatic decrease in pile stiffness due to 0 10 20 30
gapping. Morrison and Reese ( 1988) conducted Frequency ( Hz )
extensive full scale investigation of piles and
pile groups. To this conference, Purkayastha and
Dey (1991) report on their experimental study of
the degradation of vertical stiffness.
Summarizing their observations, Swan and Poulos Figure 14 Vertical pile response measured and
(1982) postulate that during cyclic lateral backcalculated for three levels of excitation
loading the two forms of degradation lead to the intensity (Han & Novak, 1988)
increase in pile deflection and bending stresses;
but if this degradation stabilizes, the pile is
said to "shakedown" to a state of permanent amplitudes measured on a 3.38 m long test pile
strains and residual stresses and will react with increasing intensity of harmonic excitation.
elastically to any further cyclic loading of the As the excitation forces grow, the resonance
same intensity. When the pile does not stabilize frequencies are markedly reduced, indicating
into an elastic or inelastic response, the pile strong nonlinearity. To the response curves,
deflections continue to increase and incremental backbone curves, n, can be constructed and from
collapse may result. The two situations are them the pile restoring force-displacement
depicted in Fig. 13. relationships are established (Fig. 15). It
appears that each response curve has its own
The shakedown phenomenon is favourable from the backbone curve and corresponding stiffness

2439
 15,------------------------.

107k 0.6
XX
0. 28
(lbf/ft) 0.4
7. 0.010
0.2,
oo+- --,-~~~~~~
o 200 400 600 800 1000
HORIZONTAL FORCE (kips)

1.0
3.5-r-------------,
1.5
Vertical Oisplocemenr ( mm)
3.0
r66cxx 2.5
Figure 15 Pile restoring forces vs displacement
corresponding to response curves from Fig. 14 <rbttttlz.o
(Han & Novak, 1988)
1.5
1.0
characteristic. The individual stiffness
characteristics feature strong overall softening 0.5 d=4ft, t=2.5in., N=IO cycles
with excitation intensity (8) but relatively O.O F== horizontal force at top of pile (kips)
modest nonlinearity.
0.0 0.5 1.0 1.5 20 2.5 30
The nonlinear pile stiffness characteristics were FREQUENCY f (Hz)
investigated for both horizontal and vertical
dynamic response by Angelides and Roesset (1980)
using toroidal finite elements in the region Figure 16 Variations of pile horizontal
surrounding the soil and the consistent boundary stiffness, kxx' with force and pile equivalent
matrix. Even neglecting slippage and gapping, damping, cxx' with force and frequency due to soil
they demonstrate the dramatic reduction in pile nonlinearity (Angelides & Roesset, 1980)
horizontal stiffness and equivalent damping with
increasing pile force (Fig. 16). The p-y curves,
also used for comparison, give lower stiffness
because they account for gapping and a high num-
ber of load cycles, N, while only 10 load cycles
were applied by Angelides and Roesset. The I
effect of a stable gap on soil resistance to pile I
steady-state vibration is schematically depicted I +
in Fig. 17. The reduction of the equivalent \
\.
linear stiffness and the necking of the loop are '
I

evident. Progressive degradation occurs under - - ~- -"'"-''"""""""""'".::...:..i:z:L~'-::..o~

incrementally increasing loads when the hystere- (a J
sis loops exhibit different shapes for sands and
clays. This is exemplified by Fig. 18 showing
the force-displacement relationship obtained by
Kishida et al. (1985) on their model piles expos- tR
ed to horizontal loads. In clay, the gap indica- Friction
/L...--'..:.:.::.r::::...-.-/ ~ / -
/
ted in Fig. 17 may expand with each cycle giving I ___. - / --._Hysteretic 6
rise to the characteristic elongated loops with It"' Radiation
reduced radiation damping. /r-;:;:: ,_..:-- _/_-
( b) _ ;;/ ...- 1 Damping
If a generally nonlinear and particularly /

transient response rather than the steady-state

response is to be investigated, time-domain
analysis is called for. The lumped mass models,
such as the one in Fig. 11, are readily amenable
to such analysis. Another type of time-domain
analysis, extending the dynamic Winkler model to Figure 17 Schematic of (a) pile under steady-
allow for non1inearities, was formulated by state vibration in stable gap, and (b) corres-
Nogami and Konagai (1986), Nogami, Konagai and ponding soil reaction, R, vs pile displacement,
Otani (1988), and Mitwally and Novak (1988). u, for stable cycle
Under axial vibration, much of the nonlinearity
is due to slip and friction. One model allowing
for slip at the pile surface, nonlinearity near
the pile and infinity of the outer zone is
depicted in Fig. 19. One of the advantages of
this model is that its properties are specified

2440
 engineering and are being presented to this
conference.

Linear Behavior of Pile Groups

Under static loads, pile interaction increases
group settlement, redistributes the loads on
individual piles and reduces bearing capacity,
unless this reduction is counteracted by
densification of the soil within the group due to
pile driving. The first suggestion of this kind
of effects probably can be attributed to
Sooysmith (1896). The investigation of static
group effects was put on a rational basis,
relying on continuum mechanics, by Poulos (1968,
-80 1971, 1979) and Butterfield and Banerjee (1971).
(a) Sand (b) Clay Extensive data on static group effects are
available in Poulos and Davis (1980), Butterfield
and Douglas (1981), El Sharnouby and Novak (1985,
Figure 18 Force-displacement relationship for 1986, 1990) and elsewhere. The static data are
incrementally increasing horizontal load (Kishida useful even to those interested in dynamics
et al., 1985) because at low frequencies, and particularly
below the fundamental frequency of a stratum
(Fig. 2), the dynamic stiffness is usually quite
close to the static stiffness.
Dynamic investigations of pile groups are more
PHe Elemeo"\ recent. The techniques employed are extensions
of the approaches used for single piles and most
of them are limited to linear interaction with no
allowance for gapping. The methods rely on the
availability of Green's functions with which the
load transfer from the pile surface to soil can
Reactions be calculated. These loading conditions,
representing one of the basic differences between
various approaches, range from point loads to
I line loads, ring loads, disk loads and finally to
cylindrical (barrel) loads; for the pile base,

-~X
disk loads are the rule. Applying this loading
to individual segments into which the pile is
(a ) discretized, the soil dynamic displacement field
SoH Reaction• is established, yielding the soil dynamic
at Pile Tip ;;;; flexibility matrix; inverting the latter, soil
stiffness matrix is obtained. In this process,
the presence of the pile cavities outside the
loaded segment is usually ignored, which implies
that wave scattering among the piles is not
accounted for, and the soil displacements are
calculated either for the pile axes or as
averages of the circumferential values. A
typical model for this analysis is shown in Fig.
20. Then the soil stiffness matrix is combined
with the pile structural stiffness and the soil-
pile system can be analyzed for any type of
Figure 19 Pile model for vertical vibration excitation. Different authors proposed various
allowing for slip, nonlinearity and infinity of refinements or simplifications to this procedure.
the outer zone (Mitwally & Novak, 1988)
The first theoretical analysis of pile-soil-pile
interaction was conducted by Wolf and von Arx
in terms of the standard geotechnical parameters. (1978) who employed an axisymmetric finite
element formulation to establish the dynamic
displacement field due to ring loads. Waas and
PILE GROUPS Hartmann ( 1981, 1984) formulated an efficient
semi-analytical method which uses ring loads and
Piles are usually used in groups and if they are is well suited for layered media, properly
not very widely spaced they interact with each accounting for the far field; the layers ought to
other generating phenomena known as pile-soil- be thin. Kaynia (1982a,b, 1988) and Kaynia and
pile interaction or group effects. These effects Kausel (1982, 1990) further improved the accuracy
have attracted much interest in recent years. A by combining the cylindrical loads, actually a
number of papers on the subject have appeared, a boundary element formulation, with the consistent
few exhaustive Ph.D. dissertations were written stiffness matrix of layered media to account for
(e.g. Kaynia, 1982a; Ostadan, 1983, Mamoon, 1990, the far field. A very similar approach is
Hassini, 1990) and many contributions have been employed in the paper to this conference by
made to the world conferences on earthquake Kobori et al. (1991) who use the cylindrical

2441
 Basic features of dynamic group effects
A few main features of the dynamic group effects
emerge from the theoretical solutions: both
S 'I stiffness and damping . are strongly frequency
.I
;: 01
I Disp,acements dependent, can be either reduced or increased due
to pile-soil-pil e interaction, may exhibit very
/
'I I (J/ \ 'I 'I /
sharp peaks and are affected even for very large
. I • /, \ I I pile spacings. Some of these features can be
lnfini te seen in the example of a 4x4 group whose
Ring I tJ \1 • I
Soil normalized dynamic stiffness is displayed for
Loads Llli I I 1 ll I Laye rs different spacings in Fig. 22. The normalization
or I I I I I_ _t_.j is done using the product (number of piles x
Cylindrical lill I I I I I single pile stiffness) , and yields a ratio
Loads I I I I I
I
G
2.5 kxx
N ks
XX
(o0 =0)
)))/))})J)))))/?/7/)/))//) 2.0
Halfspace or Rigid Base
1.5

Figure 20 Typical model for pile group analysis

loads for the pile and disk loads for the base as
depicted in Fig. 21. Also presented to this
conference is a paper by Baba (1991) who
formulates a three-dimensi onal analysis of
endbearing piles.
The thin layer method was used by Taj imi and
Shimomura (1976), Shimizu et al. (1977), Masuda 0.2 0.4 0.6 0.8 1.0
et al. (1986) and a few others. Boundary element Oo
solutions, employing Green's functions of
generally layered media, were formulated by
Banerjee and Sen (1987), Banerjee et al. (1987), G
Mamoon et al. (1988, 1990, 1990a, 1990b) and Cxx
Mamoon ( 1990) , who examined a number of cases 6.0
including pile batter and pile cap interaction. Nk s
XX
(o 0 =0)
Simpler solutions based on the dynamic Winkler
medium were developed by Nogami (1980, 1985) and 5.0
Sheta and Novak (1982). The advantages of the
latter approach are that it makes it possible to
include the weak zone (Sheta and Novak, 1982)
nonlinearity (Otani et al., 1991).

@f fi
l~
Qx '
y 11 1.0

My§ Qx~ 0.2 0.4

ao
0.6 0.8 1.0

Pz
.
.
6$J -·cc·l,
'::,;t,,,.,,.. My

Figure 22 Normalized dynamic stiffness and

damping of 4x4 pile group for different spacing
Cyl i ndri ca 1 Load Disk Load ratios, s/d (Kaynia & Kausel, 1982; homogeneous
halfspace, a 0 = wd/V 5 , L/d = 15, Ep!E 5 = 1000,
p 5 /Pp = 0.7)

Figure 21 Cylindrical loads for pile surface and

disk loads for base as used by Kobori et al.
( 1991)

2442
expressing group efficiency. As can be seen in nonlinearity is needed, however, at least in the
Fig. 22, the group properties and their variation form of the weak zone and a pile free length. A
in frequency depend strongly on the spacing few observations on nonlinear analysis will be
ratio, sjd, with the peaks shifting according to made later herein.
this ratio. This is so because pile interaction
depe~ds on the ratio of the wave length to
spac1ng. It has been questioned whether the
sharp peaks in stiffness would not be suppressed
due to soil nonlinearity and interface
deficiencies discussed above. A group solution
including the weak zone around the piles dulls
the peaks, but does not eliminate them, as is
depicted in Fig. 23. On the other hand, soil

'"
..Q ..
Q

80
16
2
14 N 70
"
~ 60
.E
~ 10 ~50
(j,
~"' 6 >- 40
U) 0c:
6 "& 30
0 0

a:"' 4
E
- 20
---Nonhomogeneous
----Homogeneous

0.25 0.5 0.75

Frequency o 0 • Rw I V5 (b) Frequency o 0 = Rw /Vs
5

Figure 23 Effect of weak zone on vertical

dynamic stiffness of 2x2 group of four concrete
floating piles (s/d = 4, d = 2 ft = 0. 61 m,
parabolic soil profile; Sheta & Novak, 1982) Figure 24 Horizontal dynamic stiffness of 3x3
pile group in homogeneous and nonhomogeneous soil
(Kaynia, 1988; sjd = 5, L/d = 20, E5 /Ep = 0.01)
nonhomogeneity can make the peaks either more
pronounced, as shown in Fig. 24, or duller as is
argued by Gazetas and Makris (1991), depending on
conditions such as frequency and spacing. 20
P(tl =me ew 2 cos wt
With the pile-soil-pile interaction theories
~
being so complex, it is of importance to examine
IE
~ 16
how the theories perform when compared with

•
E • w Static
experiments. Figure 25 shows one such comparison . Interaction
based on a group of four closely spaced model N
'
<{ 12
piles tested in the field and evaluated using the
plane strain theory for soil reactions with a "
1J
10~ ~10
No Interaction
~ ~ol
weak zone. The responses evaluated ignoring ii 20 G Gm=G
E
interaction or assuming static interaction are "' e
completely inadequate. The dynamic interaction ::l Dynamic Interaction
theory gives far better results. On a test group c0" theory
of 102 model piles 1. 06 m in length similarly ·~
c: 4 symbol rb in
encouraging results were obtained (Novak and El •E ~
0 ~24010

Sharnouby, 1984}. For six full scale piles very i3 3·40180

good results were also obtained but the weak zone
and separation had to be included for a satisfac- 0 10 50 60
tory match (Fig. 26}. Successful experiments
conducted on a group of 56 full scale piles were
reported by Masuda et a 1. ( 19 8 6) . In their
report to this conference, Kobori et al. (1991)
also found the theory to be of sufficient Figure 25 Vertical response of 2x2 group of
applicability. Thus, it may be concluded that closely spaced piles: theory vs. experiment
the linear theory works quite well as long as the (Sheta & Novak, 1982; L = 3.4 m, d = 60.3 mm)
experiments do not deviate too much from the
theoretical assumptions, as might be expected.
Often, a correction for separation, gapping and

2443
......
Co
::>•
1-N
H
.. •
•t
': .,, FLOATING

•
--'

.'.....
IL

~= '
__,-
<
1-
··!· REAL
·-- IMAGINARY
L/d=15
Ep/E5 =1000
z y
~::: "*:ha&i•
H.
a::-
0
.
X
en
eno
w"!
.......
z .. Theory
0
H
en t• 0 96 kgmm
z~ +• --'
~0 t• Exp. 171 kgmm ~
i'
A
H 0::-0.2
0
+ 259 kgmm
2
0
0

[Link] 15.00 20.00 25.00 30.00 35.00 40. -0.4 f----+---+---+-----+- ---+---+----l
FREQUENCY (HZ) 0.0 0.2 0.4 0.6 0.8 1.0
a 0 =wd/Vs

Figure 26 Horizontal theoretical and Figure 27 Vertical dynamic interaction factor

experimental response in Y-direction for group of for different spacings vs. dimensionless
six concrete piles 7.50 m long, 0.32 m in frequency (Kaynia & Kausel, 1982)
diameter (El Marsafawi et al., 1990)

Interaction Factors luxFx (8=0.0) ENDBEARING

0.6 r - - - : : - - - - - ; - : - - - - - - - - - - - - - - - ,
The pile group analyses discussed above differ in s/d
accuracy and computing effort but for all of them REAL L/d=15
the computing requirements are quite severe, (/) 0.4 IMAG
f- Ep/E 5 =1000
particularly for larger groups. Therefore, a::
<{
Kaynia and Kausel (1982) formulated the concept a.. 0.2
of dynamic interaction factors being an extension >-
c::
<{
of the widely used static interaction factor z
0.0
approach. In this approach, only two piles are l5
<{
considered at a time and the group properties are 2:
obtained by superposition. .;o -0.2
---'

- ..
', 2
Dynamic interaction factors are dimensionless, ~ -------
frequency dependent complex numbers, defined as c:: -0.4

_ Dynamic displacement of pile 2

aij - Static displacement of pile 1 (3) -0.6
0.0 0.2 0.4 0.6 0.8 1.0
a 0 =wd/Vs
in which the displacement of pile 2 is caused by
a unit harmonic load on pile 1 and the static
displacement of pile 1 is established for an
isolated pile. The displacement is either Figure 28 Horizontal dynamic interaction factors
translation or rotation. Examples of the real for two endbearing piles in line
and imaginary parts of the interaction factors,
calculated using the Kaynia and Kausel ( 1982)
method, are plotted for homogeneous soil, using circumvented if the interaction factors are
the authors' notations, in Figs. 27 to 29. The expressed in terms of amplitude, Ia!, and phase,
interaction factors are oscillatory in character, ¢, i.e.
i.e. negative as well as positive. Negative a = a 1 +ia 2
values of the imaginary part indicate a possible
increase in group damping characterized by group (4)
efficiency greater than unity. A complete set of
interaction factors is available for floating
!a lei¢
piles, homogeneous soil and a limited selection As an example, the interaction factors from Fig.
of parameters in Kaynia and Kausel (1982) and for 27 are presented in this form in Fig. 30.
vertical vibration in linearly nonhomogeneous
soil in Banerjee (1987). correcting the available interaction factors for
pile length, endbearing and other effects a very
The interaction factors, such as those shown in efficient approximate procedure for' group
Figs. 27 to 29, are commonly displayed in terms analysis is obtained. For example, the vertical
of their real and imaginary parts. This is a or horizontal dynamic stiffness of a group with
usual form but it makes interpolation for a rigid cap becomes
intermediate spacings difficult, especially at
higher frequencies. This difficulty can be (5)

2444
 The interaction factor approach would be
luxFx (1J=1f/2) ENDBEARING mathematicall y accurate if the interaction
0.6 factors as well as the single pile properties
s/d -REAL L/d=15 were calculated with all piles present in the
system, which is not normally done. Neverthe-
(f) 0.4 . -- IMAG Ep/E5 =1000 less, the results may be quite adequate for most
1-
0::
<(
applications. Kaynia and Kausel (1982) found the
a.. 0.2 accuracy of the interaction factor approach to be
~
<(
quite sufficient for a homogeneous medium; for a
z nonhomogeneou s medium, Kaynia (1988) observed the
0.0 --.
~ ~ \ approach to be less accurate. Judging from
static pile group behavior more significant
~ ~ ~: .. -- .. - ?.. --- -
.:g -0.2 "'--- . ' errors, overestimatin g the interaction effects,
...J
-- - ' ' 2--------- may occur in the vertical response of endbearing
L5 • piles (Fig. 31) .
0:: -0.4

-0.6
0.0
-· 0.2 0.4 0.6 0.8 1.0
0.6

a 0 =wd/V5
"'
':>::"'-' 0.5

Figure 29 Horizontal dynamic interaction factors

for two endbearing piles with 90° incidence angle Q
~ 0.4
a:
>-
u
z 0.3
0.8 0 ~
~
I.J..
-2 I.J..
w
w0.4 -4 a..
::)
0 w 0
5
11.
~-6
D.
ll:
<.:)

~0.2 -8 0.1
DIRECT ANALYSIS
INTERACTION FACTORS

0.00.0 - 12o.o 0 1 2 3 4 56
0.1 0.2 0.3 0.4 0.5 0.1 0.2 0.3 0.4 0.5 7 8 9 10 II
~ ~ FLOATING - - - Eb;E - - END-BEARING
s

Figure 31 Comparison of vertical static group

Figure 30 Vertical interaction factors from Fig. efficiency using direct analysis and interaction
25 in terms of amplitude and phase factors (El Sharnouby & Novak, 1985)

A remarkably simple approximate method for dyna-

in which k is static stiffness of a single pile mic interaction factor evaluation was proposed by
and , are the elements of the inverted matrix Dobry and Gazetas (1988) and extended for non-
[a], listing all the complex interaction factors homogeneous soils by Makris, Gazetas and Fan
between any two piles in the group. (For all the (1989) and Gazetas and Makris (1991).
vibration modes, the corresponding formulae can For
homogeneous soils, these authors assume that the
be found in Novak and Mitwally, 1990). When the displacement field around a vibrating pile and
loads on all the piles in a group are the same, thus also the displacement of the neighbouring
as in a doubly symmetrical group of four piles, pile (the interaction factor) is governed by the
a simple formula for the group stiffness applies, law of cylindrical wave propagation. Then, e.g.,
i.e. the vertical interaction factor is simply

n k ro 1/2 s
KG (6) a ~ (--) exp(-pw--) exp(-iwvs ) (7)
n v s ~ s
f' + l: aij
j=2 where p = soil hysteretic damping ratio. In
their comparisons with the more rigorous
solutions for floating piles, the authors
in which f' = k/K is the ratio of the single pile obtained a very reasonable, although not quite
static stiffness to its complex, dynamic perfect, agreement. For endbearing piles in a
stiffness n is the number of piles and the homogeneous stratum (Figs. 28 and 29), the
interacti~n factors refer to one reference pile. frequency variations of the interaction factors
Eq. 6 is often used as approximate even for more are apparently too irregular to allow a simple
general pile configuration s. description by a formula such as Eq. 7 and the
same may be true for markedly stratified media.

2445
Nonlinear Analysis of Pile Groups
Nonlinear dynamic analysis of pile groups is very o Sfc! = 3 {3 = a• I st LOAD
difficult and this may be the main reason why it "' S;d = 6 {3 = o• Ist LOAD
received much less attention than the linear
analysis. Akiyoshi and Fuchida (1982) formulated
b;_j I 6
+ 5~c~ = 3 {3 = o· RELOAD
an approximate solution for vertically vibrating :z <:~·­
------1 x S;d = 6 {3 = a· RELOAD

~i[~~ -.--~--~
endbearing piles considering imperfect adhesion oo
between the pile and the soil modelled by a >-
u
friction type interface. They found that slip w
-'
occurs near the ground surface and proceeds to LL
w
the bottom of the soil layer as the applied force 0 <-.J
increases. Nogami and Konagai (1987) developed 0
w
0

a group analysis assuming also that in the ~-.~

vertical vibration, response nonlinearity stems _J

-<:
~4-- '+- + +
mainly from slippage at the soil-pile interface; :L:
a:
..,.'"'I.__~ --~-.L_ ~ . . . -....., . . . _ -----:-- ___!:
......
they represented the soil using the dynamic (')
zo ::t- ___
Winkler model. They found that this nonlinearity
reduces the wave interference effects, making the --------)("---~~
stiffness less frequency dependent, and under >:;_- -x- X 0/d
transient loading produces residual skin friction .02 .04 .06 .OB .10 .12 .14
and residual axial force in the pile. Then, =to~~~~~-T~~-r~~~~~TL~~~-T~~
3 S 10 II 12 13 It
Nogami et al. (1988) and Otani et al. (1991)
extended the concept of the dynamic Winkler DEFLECTION OF LOADED PILE IN NM
medium further to include horizontal response,
slippage, gapping and inelastic soil behavior
being able to generate a variety of degrading Figure 33 Horizontal static interaction factors
hysteresis loops. for first loading and reloading (Janes & Novak,
1989)
The opinion is sometimes expressed that under
large displacements most of the action occurs
right around the pile and consequently, pile- factor diminishes with increasing deflection,
soil-pile interaction is not very significant. dropping to about one half of the original value
Some insight into this can be obtained from at the deflection of about 3.5 per cent of the
static experiments. Figures 32 and 33 show the diameter, and then levels off. This reduction
results of field tests conducted on free-headed varies with spacing and the angle of incidence.
test piles being steel pipes 0.1016 m in outer If the pile is unloaded and reloaded, the
diameter and 3. 05 m in length. The soil was interaction factors for small displacements
stratified, mainly silty sand changing to gravel. become much smaller than the original ones,
apparent~y due to gaps generated by the preceding
Figure 32 shows two curves, one depicting the
nonlinear response of the loaded pile and the large dlsplacements, and then asymptotically
other showing the interaction factor (normalized approach the values from the first loading. This
deflection), a, vs deflection. The interaction behavior obviously makes the analysis of
irregular transient response very difficult.
Much more research in this area is needed. The
preliminary conclusion is that under large
displacements pile-soil-pile interaction is
reduced but not eliminated.
0 I{)
(j
0 1:!1 ~ q
c,a~ [Link] ---1--~-- e:j--- OTHER FACTORS AFFECTING PILE BEHAVIOR
~
m ::JI
= w
(/)z
" z There are a few other factors that affect pile
g• I ~ (j~ response, among them pile batter, soil-pile-cap
6 I- u
z a w interaction and soil liquefaction. These are
_J
z LL briefly discussed in this section.
row
0~
0
oo
<
0
_j
0
~- .-o 0
w Pile Batter
_j
< 0
f-3--l ~
N
:J Pile batter is often used to increase the
0: g O<t
w~
>- ~
horizontal stiffness of the group. For machine
-co foundations and other structures this is
_jg o LOADED PILE P-Y CURVE 0:
0
z sometimes useful. However, under earthquake
0
0
A ADJACENT PILE INTERACTION loading, pile batter may not always be beneficial
becau~e it restri~ts t?e pile's ability to sway
N

and y1eld, result1ng 1n greater seismic forces

a~d pos~ible da~age. to the piles and the cap.
6 7 10
L1ttle 1nformat1on 1s available on the dynamic
effects of the batter. As a very approximate
DEFLECTION IN MM practical approach, the pile can be analyzed
first as_if it were vertical and the stiffness
matrix [~] o?tained in this way taken as valid
Figure 32 Lateral load vs deflection and static for ~he 1ncl1ned element coordinates; then, this
interaction factor a vs deflection (Janes & matr1x can be transformed into global coordi-
Novak, 1989) nates, being horizontal and vertical, to give the

2446
 battered pile stiffness matrix in these coordi- 43 per cent. Also, a slight shift in the peak
nates, [K]. More details on this are given by can be noticed. The batter effect on the
Novak (1980). For static conditions, Poulos imaginary parts is similar but at frequencies
(1980) employs a similar technique. higher than o. 6 the imaginary parts of the
He recom- impedances are increased.
mends the evaluation of interaction between two For the horizontal
battered piles such as that of two vertical piles response, the data available are not sufficient
whose distance is equal to the separation to make a general conclusion.
measured on the inclined piles at L/3 from the
top.
Soil-Pile-Cap Interaction
One of the few dynamic solutions of pile groups In most situations piles have caps and soil-pile-
with batter was produced by Mamoon (1990) using cap interaction may occur. The cap influence
an approximate analytical formulation, denoted as depends not only on the size and embedment of the
Method I. This procedure involves the construc- cap but also on the quality of its contact with
tion of an integral representation for the soil the soil. Considering the behavior of actual
domain modelled as an elastic halfspace. An soils under static and dynamic loading, it may be
example of Mamoon•s results is shown in Fig. 34 speculated that this contact will be well
maintained in stiff clays and dense sands; but in
loose to moderately dense sands the cap base may
5.11 separate from the soil and in soft clays the
contact in the cap base as well as along the cap
4.1 sides can be lost; finally, the separation of the
v. == 0.40
/3. == 0.05 base is more likely to occur for endbearing
3.11 METHOD I piles.
KAYNIA
[Link] The few dynamic analyses that have been reported
invariably presume full contact and perfec~
~t= 1.11 elasticity and thus their results should be
applied with some allowance for the actual soil
behavior.
.II --- Banerjee and Sen (1987) observed a
rather small effect of the cap on the vertical
impedances of single piles and groups of two and
-I.B four floating piles respectively. This might be
(a) Vertical Piles a valid conclusion for the rather stiff piles
they analyzed (Ep!E. = 10000) . For more flexible
piles the cap may cause a more significant
3.11r-----------------------------------------~ increase in pile impedances as can be deduced
- · - Sfd=2 from static analysis. Figure 35 illustrates this
--S/d=S point.
----- S/d 10 =
z.a
/--......__............
,'
1.7

I - ~

~
1.6

__ _,' /
-·-·-·---·-·-·--·-.
1.5
l/d-1G

,...it:: ...
.B
-iol-

(b) Inclined Piles

1.l
-I.B .'::BB;:---------'.':;;zs;----------'-.sa;;;:-- --------':.75;;--------:-'[Link]

I 2

Figure 34 Normalized real part of vertical 11

impedance of 3X3 pile group for (a) - vertical
piles and (b) - piles with 15° incline (L/d = 15,
Ep!E. = 1000, p 0 /Pp = 0.7) (Mamoon, 1990)

comparing the normalized vertical stiffness (real

part) of a 3x3 group of vertical piles with that Figure 35 Ratio of vertical static stiffness of
of a similar group featuring piles with a 15° single pile with cap, K , to stiffness of pile
batter in one plane. (Notice the vertical scale without cap, K, for different cap diameters, d 0
is not the same for both cases.) Kaynia 1 s (E.,!Es = 1000, v = 0.5; Liu & Novak, 1990)
solution of the vertical group is displayed for
comparison. The normalization is by the static
stiffness of a single vertical pile multiplied by An extensive theoretical study of the dynamic cap
n. The comparison of cases (a) and (b) suggests effects was conducted by Mamoon (1990). He
that for the separation s/d = 5 and higher included cap inertia in his analysis but ignored
frequencies, the inclination of the piles causes the shear stresses in the mat base, even for the
a significant reduction in the real part of the horizontal response. An example of Mamoon 1 s
impedance. For the peak, this reduction is about

2447
results is shown in Fig. 36. The principle Effect of Soil Liquefaction on Pile Behavior
observation is that for some conditions, cap Piles are often used in loose saturated sands and
inertia can reduce or even eliminate the sharp silts. If such deposits liquefy due to increased
peaks in the impedances, typical of pile groups pore water pressure during earthquakes, the piles
without caps. lose much of their lateral and vertical support
which can result in a substantial increase in
bending moments, loss of stability and failure.
2-Br--------------------------------------------, Damage of this type occurred in the Niigata and

-z.a
·-=-=-="'-:::::_- -----~
Alaska earthquakes of 1964 and elsewhere.
Relatively few studies were devoted to this

ol;:>
-6.a
- - S/d = 2, B/d = 6.5
----- S/d = 5, B/d = 12.5
."\\ ./I "\ important subject, e.g. Finn and Martin (1980),
Matlock et al. (1981) and Yoshikawa and Arano
(1988). To this conference Nomura et al. (1991)
present their theoretical and experimental study
of pile behavior during liquefaction.
:.;1~ -11!.9
- · - S/d = 10, B/d = 22.5
.\ Their
theory employs a one-dimensional effective stress
\.I analysis and Ramberg-Osgood' s
relationship for soil.
stress-strain
For the piles, a lumped
mass model similar to that of Penzien et al.
(1964) is used. The experiments were conducted
(a) Real Part in a 4. o m long container on a large shaking
-Ia. a '-.-=il-------'-.-=z-------'-.-:.-------'-.-=s-------'-.-::a:------~~.il table. The authors achieved excellent agreement
between the theory and experiments with regard to
8.8,--------------------------------------------,
...
ground motions, pore water pressure and pile
r-- B---j response. The differences between the behavior

.. .. ..
__L_
S.B of flexible piles and rigid piles were
~ demonstrated and the one order of magnitude
s
•.a ___._
__________ I ·"""'\
------::.::-~-::--..
increase in both ground and pile motions due to
liquefaction was documented .

of
0~
3.B

2.B /.///
./-~/ \\ I
r-----~
SOIL-PILE-STRUCTURE INTERACTION
/" ,,"'' . . Once the properties of the pile foundation are
I.B
/. ~-~~­ established, they can be incorporated into the
.a -- -- ---
. examination of pile-structure interaction just as
with other types of foundations. A number of
(b) Imaginary Part studies have been devoted to this subject. As
-1.2'--::-------'--::-------~------~~------~~----~
.a
there is a difference between direct excitation
.2 .• .s .a 1.a
of the structure by external loads and excitation
by seismic motions of the ground, these two cases
will be discussed separately.
Figure 36 Normalized vertical impedances of 3x3
groups with caps for different spacings and cap
sizes (Mamoon, 1990; L/d = 15; EJ~. = 1000, p 5 fpp Pile-Structure Interaction Under External Loads
= 0.7, v 5 = 0.4, ~. = 0.05, cap fh1ckness = 3d) Typical examples of direct external loads are
unbalanced forces acting on machine foundations,
wind forces on buildings and wave forces on
An approximate practical approach to cap offshore towers. In such cases, the pile
interaction is employed by Kobori et al. (1991) foundation impedances can be superimposed on the
in their paper to this conference. To analyze structural system matrices to give the governing
the response of a group with an embedded cap, equations of the pile-structure system in the
these authors superimpose three partial solutions standard form, i.e.
to the entire problem as indicated in Fig. 37,
and add side soil springs to account for the cap
embedment. In the comparison of their analysis (m) {U} + [c] {li} + [k] {U} = (P(t)} (8)
with experiments they get fair agreement.
in which [m], [c] and [k] are the mass, damping
and stiffness matrices incorporating the
structure and foundation properties and, in some
cases, other factors such as hydrodynamic
effects, aerodynamic damping etc.; {u} and {P(t)}
are the displacement vector and loading vector
Wilhoul a gap wan a pap Soil columns respectively. Two examples of structural
response to external loading are given here, both
with the aim of illustrating the effects of pile-
Figure 37 soil-pile interaction.
Approximate approach to cap
interaction based on superposition of three
partial solutions (Kobori et al., 1991) Figure 38 shows the horizontal and rocking
components of the response of a compressor
foundation to harmonic unbalanced forces. The
foundation is a concrete block 4.88 x 3.05 m in
plan supported by eight endbearing wood piles

2448
"'-' lnlerocflon Consider~

...
ON
:::>-
•••• "'e G.o
"''Q
lnttroctlon

.J
a..
I:
<o
••••
11
Wove
Spectrum

~ 4.5

...
(f)
(f) &
"'-'
.J
z ll:
0
I;
;;:;"'
z "'No J 3,0
"'-'
{:.
I: Interaction
0
E
.J'"'
~
<
f-
z
0
""'z .
0.
V>
J,S

N
a: ...
""
(_)
00
0 a:..:.
:I:

N
"'e s . o , _ - - - - - - - - - - - - - - - - ,
"''g
lnteroctlon Considered
~ 4.0 lnlerocllon Neglected
75 !50 225 60 120 180 2+0 300 c
0 ~ ~~;c"trum U_•_3_0_m-/s'j
r-1 102 E
FREQUENCY w .,~
FREQUENCY w
0: 3.0 ,, ·.
ft\
..
~

Effect of dynamic pile-soil-pile

¥ .\. 101 g
Figure 38
interaction on harmonic response of machine ~
_ 2.0
' I
~~
foundation 0
I '. 100 ll.
e l\ . · ·..• _...-Wove Spectrum ...------ E
c \ ·. ·~
~ 1.0 I '·., 1 I
with a minimum spacing ratio sjd = 5.4. As can
be seen from Fig. 38, dynamic pile-soil-pile ~ \ '···~... 10- ~
interaction reduces the resonant amplitudes quite
substantially but shifts the resonance
~ 0 L_!.__ _ '_~-="""--"~"-t-'~·,,··,·_··____,__ _.J I0- 2 ~
/!. 0 (JJl 2 3
frequencies only slightly.
Frequency W( ,-II

The second example involves the response of a

pile supported offshore tower to random wave
forces. The tower is a steel template structure,
122 m high, supported by eight steel piles. (For Figure 39 Effect of pile-soil-pile interaction
details see Novak and Mitwally, 1990). The on power spectra of steel offshore tower response
response of the tower to wave forces was analyzed to wave forces for two wind velocities (Novak &
in terms of random vibration twice, i.e. Mitwally, 1990)
considering pile-soil-pile interaction and
neglecting it. The power spectra of the tower
response are shown in Fig. 39. The waves are expected to act on the structur~ and the piles in
wind driven. For a medium wind velocity of 22 a seismic event. such [Link] can be done
mjs, the response spectrum features two peaks: experimentally or theoretically.
one is centred around the fundamental frequency
of the tower, w1 , while the other coincides with The experimental investigations are most often
the peak of the wave spectrum. The first one is conducted on models using shaking table tes~s,
dramatically reduced due to pile-soil-pile less often in a centrifuge. The tests [Link]
interaction because it is of resonant type and as careful scaling and special design of the test
such is sensitive to the increase in damping this bin boundaries which should prev~nt wave
interaction causes. The second peak occurs well reflections (the box effect) . [Link] table
below the fundamental tower frequency, is tests of pile supported structures were reported
quasistatic and indicates response amplification by Mizuno et al. (1984), Nomura et al: (1991) and
due to increased flexibility. At the higher wind a few others; pile scaling was [Link] by ~ana
velocity of 30 mjs, the dominant frequency of the et al. (1986) and the modeling of. free-:-[Link]
wave spectrum is very low, most of the response conditions in centrifuge tests was [Link].g~ted
is quasistatic and is increased due to increased by cheney et al. ( 1990) . Earthquake observat~ons
pile foundation flexibility (reduced stiffness). on a large scale model featuring 7.5 m long [Link]
In addition to this stiffness reduction, gapping were made by Kobori et al. (1991) ·
was observed to temporarily reduce tower natural
frequencies during heavy storms. For design purposes, the theoretical ana~ysis of
pile-structure interaction is more [Link] and
is conducted much more often. Adequate for
Pile-Structure Interaction Under Seismic routine designs is a simple procedure ?ased on
Loading substructuring and the following [Link]: the
The evaluation of soil-pile-structure interaction input ground motion is given for the level of
is needed in order to establish the forces pile heads and is not affected by the presence of

2449
the piles and their cap1 soil-pile interaction incidence was examined by Mamoon and Banerjee
analysis is conducted separately to yield the {1990a), Mamoon and Ahmad (1990) and Ahmad
pile foundation impedances1 and, the seismic (1991).
response is obtained from Eq. 8 using standard
analysis, even response spectra. For shear For pile groups, kinematic interaction can be
buildings all the matrices in Eq. 8 may be more significant. Waas and Hartmann {1984)
rearranged to take on the form that is common to examined a single pile and a large group of 356
shallow foundations (see, e.g., Novak and El piles and concluded that while a single pile
Hifnawy, 1984). This type of analysis, known as follows the earthquake motion of the soil with
inertial interaction analysis, usually indicates little deviation, a large group of stiff piles in
that the pile foundation flexibility and soft soil shows a response significantly diffe-
dissipative properties result in the reduction of rent from the free-field motion. Significant
the seismic forces as well as the base shear and kinematic interaction effects were also observed
an increase in the relative building response, for a similar pile group by Wolf and von Arx
just as in the case of shallow foundations (Novak (1982) who considered horizontally traveling
and El Hifnawy, 1984). waves. Thus, for important projects such as
nuclear power plants, a complete analysis
The assumption of the input ground motion not including kinematic interaction may be desirable.
being affected by the presence of the piles is Such a complete response analysis of a pile-
based on the ideas that the dominant seismic wave supported structure, in which the kinematic
lengths are much larger than the pile diameter, interaction is evaluated beforehand to give the
and given the bending flexibility of slender ground motion for the inertial interaction
piles, the piles will follow the horizontal calculation, is schematically indicated in Fig.
motion of the ground. A more comprehensive 41 with M representing the mass of the structure
examination of these assumptions involves and a 0 input bedrock acceleration. Analysis of
consideration of the wave scattering effect, this type was conducted by Waas and Hartmann
known as kinematic interaction. (Unfortunately, {1984), Hadjian et al. (1990), Kobori et al.
there is not a unique definition of this term.) (1991) and others.
A few researchers examined this phenomenon.
Gazetas (1984) conducted an extensive parametric 0 Ehl

TI_
study of the response of single endbearing piles
exposed to harmonic shear waves propagating
upward from the bedrock. He defined the 1 .

kinematic interaction factor as

... . .. . .
v
v

( 10)
Iu
'l.
.
•a
"o a
• v

0
2 4 6 8 10
in which uP, u 0 are the absolute values of the FB
horizontal displacements, relative to the
bedrock, of the embedded pile head and the ground
surface motion in the absence of the piles,
respectively, and u• is the pile head rotation Figure 40 Kinematic interaction factor for
absolute value. The magnitude of I~ depends on parabolic soil profile vs dimensionless frequency
the soil profile, the stiffness rat~o Ep!E., the parameter F8 1 E8 = E5 (z=d) 1 (Gazetas, 1984)
slenderness ratio L/d and the frequency ratio
f/f\ where f = wave frequency and f 1 = fundamental
hor~zontal frequency of the soil layer being for The two step response analysis shown in Fig. 41
a parabolic soil profile equal to 0.56 V5 /L. indicates that pile stresses also come from two
When there is no kinematic interaction Iu = 1. sources, i.e. pile deflection due to ground
Synthesizing his numerical results, Gazetas found motion and inertial interaction. One limitation
it possible to express the kinematic interaction of the accuracy of most kinematic interaction
factors for each soil profile in terms of a studies is that they assume soil linearity. It
dimensionless frequency parameter. For the is well known that for strong earthquakes linear
parabolic soil profile this parameter becomes site response analysis can yield unrealistic
displacements and stresses.

(11)

In terms of this parameter the kinematic

interaction factor for translation assumes the 17
form plotted in Fig. 40. As can be seen, for
small f/f 1 , E../E 8 and d/L, the kinematic
interaction facfor is close to unity1 for large
values of these ratios it drops to about 0.5. In
the studies conducted by other authors this drop
can be even more pronounced. This suggests that
the error resulting from the omission of kinema-
tic interaction is either negligible or is on the
-
ao
Complete Solution
MkZ.

-
z~

Do
.z

Kinematic Interaction Inertial Interaction

conservative side. Only for the homogeneous soil

profile, slight amplification of Iu may occur at Figure 41 Schematic of seismic response analysis
low frequencies. The effect of the angle of including kinematic interaction

2450
One more complication may occur if the piles are Angel ides, D. and Roesset, J .M. (1980).
not adequately connected to the cap or if this "Nonlinear Dynamic Stiffness of Piles", MIT
connection fails in a severe earthquake. Then Res. Rep. R80-13, Dept. of Civil Eng.,
the cap may· uplift, as indicated in Fig. 42, Cambridge, Massachusetts.
modifying the seismic forces on the building and
substantially increasing the forces on the Aubry, D. and Chapel, F. (1981). "3-D Dynamic
peripheral piles that maintain the connection. Analysis of Groups of Piles and comparisons With
These piles can become overloaded and may fail. Experiments", SMIRT, Paris, 9 p.
Uplift of the tip of an endbearing pile, which
was not socketed, from the bearing stratum may Banerjee, P.K. (1978). "Analysis of Axially and
have similar but less severe results. More data Laterally Loaded Pile Groups", Chapter 9 in
on the uplift effects can be found in El Hifnawy Developments in Soil Mechanics, Ed. C.R. Scott,
and Novak (1986, 1987). Appl. Sci. Pub., London, 317-346.
Banerjee, P.K. and Mamoon, S.M. (1990). "A
Fundamental Solution Due to a Periodic Point
Force in the Interior of an Elastic Half Space",
Earthq. Eng. Struct. Dyn., vol. 19, no. 1, 91-
105.
Banerjee, P.K. and Sen, R. (1987). "Dynamic
Behaviour of Axially and Laterally Loaded Piles
CAP I and Pile Groups", Chapter 3 in Dynamic Behavior
UPLIFT c!:::::--- I of Foundations and Buried Structures
(Developments in Soil Mech. Found. Eng., vol.
·· .. 'I ·. rf.·. 3), Ed. P.K. Banerjee and R. Butterfield,
Elsevier App. Sc., London, 95-133.
Banerjee, P.K., Sen, R. and Davies, T.G. (1987).
"Static and Dynamic Analyses of Axially and
(ol FIXED HEAD ( bl FIXED HEAD (c) FREE HEAD
END BEARING PILES
Laterally Loaded Piles and Pile Groups", Chapter
END BEARING PILES
SOCKETED PILES
7 in Geotechnical Modeling and Applications
(Dean Alexander Vesic Memorial Volume), Ed. S.M.
Sayed, Gulf Pub. Co., Houston, p. 322.
Figure 42 Uplift of pile cap and pile tip under
seismic loading Blaney, G.W., Kausel, E. and Roesset, J.M.
(1976). "Dynamic Stiffness of Piles", Proc.,
2nd Int. conf. on Num. Methods in Geomech. ,
CONCLUSIONS Blacksburg, Virginia, 1001-1012.
Considerable progress has been made in the Blaney, G.W., Muster, G.L. and O'Neill, M.W.
development of dynamic analyses of single piles (1987). "Vertical Vibration Test of a Full-
and pile groups, experimental techniques for Scale Pile Group", Geotech. Special Pub., ASCE,
laboratory and field pile investigations and no. 11, 149-165.
understanding of pile behavior. Further research
is needed, particularly into soil-pile interface Butterfield, R. and Banerjee, P.K. (1971). "The
behavior, nonlinear pile-soil-pile interaction Elastic Analysis of Compressible Piles and Pile
and the interaction between the piles and their Groups", Geotechnique, vol. 21, 43-60.
caps, both surface and embedded.
Butterfield, R. and Douglas, R.A. (1981).
"Flexibility Coefficients for the Design of
ACKNOWLEDGEMENTS Piles and Pile Groups", Construction Industry
Res. and Inf. Assoc., 6 Storey's Gate, London,
The writer appreciates the cooperation of all his England, Technical Notes, 108.
associates who worked with him over the years and
is indebted to A. Kaynia for making his program Cheney, J.A., Brown, R.K., Dhat, N.R. and Hor,
for pile group analysis available to him. O.Y.Z. (1990). "Modeling Free-Field Conditions
in Centrifuge Models", J. Geotech. Eng., vol.
116, no. 9, 1347-1367.
REFERENCES
Crouse, C.B. and Cheang, L. (1987). "Dynamic
Akiyoshi, T. and Fuchida, K. (1982). "An Testing and Analysis of Pile-Group Foundations",
Approximate Solution of Vertical Vibration of Geotech. Special Pub., ASCE, 1987, No. 11, 79-
End-Bearing Piles With Frictional Interface", 98.
Proc. JSCE, no. 324, August, 31-40 (in
Japanese). Davies, T.G., Sen, R. and Banerjee, P.K. (1985).
"Dynamic Behavior of Pile Groups in Inhomo-
American Petroleum Institute (1986). geneous Soil", J. of Geotech. Eng., Dec., vol.
"Recommended Practice for Planning, Designing 111, no. 12, 1365-1379.
and Constructing Fixed Offshore Platforms" (16th
edition), Dallas, TX. De Beer, E. et al. (Eds.) (1977). Proc. of
Specialty Session 10, "The Effects of Horizontal
Loads on Piles Due to Surcharge or Seismic
Effects", IX ICSMFE, Tokyo, July 14, p. 200.

2451
Dobry, R. and Gazetas, G. (1988). "Simple Method Gle, D.R. (1981). "The Dynamic Lateral Response
for Dynamic Stiffness and Damping of Floating of Deep Foundation", Ph.D. Thesis, Univ. of
Pile Groups", Geotechnique, vol. 38, no. 4, 557- Michigan, Ann Arbor.
574.
Hadjian, A.H., Falgren, R.B. and Lau, L. (1990).
Dobry, R. , Vincente, E., 0' Rourke, M. J. and "Imperial County Services Building Revisited: A
Roesset, J .M. {1982). "Horizontal Stiffness and Reevaluation With Pile-soil-Structure Inter-
Damping of single Piles", J. Geotech. Eng. Div., action", Proc. 4th u.s. Nat. Conf. on Earthq.
ASCE, vol. 108, no. GT3, 439-459. Eng., Palm Springs, Cal., vol. 3, 835-844.

Dotson, K.W. and Veletsos, A.S. {1990). "Verti- Han, Y. and Novak. (1988). "Dynamic Behaviour of
cal and Torsional Impedances for Radially Single Piles Under Strong Harmonic Excitation",
Inhomogeneous Viscoelastic Soil Layers", Soil canadian Geotech. J., vol. 25, no. 3, 523-534.
Dyn. & Earthq. Eng., May, vol. 9, no. 3, 110-
119. Hassini, s. (1990). "Static and Dynamic
Behaviour of Pile Groups", Ph.D. Thesis, Civil
El Hifnawy, L. and Novak, M. (1986). "Uplift in Eng., Univ. of Michigan, Ann Arbor, Michigan.
Seismic Response of Pile Supported Buildings",
J. Earthq. Eng. & Struct. Dyn., vol. 14, August, Hassini, s. and Woods, R.D. (1989). "Dynamic
573-593. Experiments With Model Pile Foundations", Proc.
12th ICSMFE, Rio de Janeiro, vol. 2, 1135-1138.
El Hifnawy, L. and Novak, M. (1987). "Seismic
Response of Buildings With Pile Uplift", Proc. He, Y.A. (1990). "The Mixed Boundary Problem of
5th canadian conf. on Earthq. Eng. , ottawa, Soil-Pile Interaction", Soil Dyn. & Earthq.
July, 181-190. Eng., vol. 9, no. 1, 20-24.

El-Marsafawi, H., Han, Y. and Novak, M. (1990). Hewitt, C.M. (1988). "Cyclic Response of
"Dynamic Experiments on Two Pile Groups", Res. Offshore Pile Groups", Ph.D. Thesis, School of
Rep. GEOT-20-90, Dept. Civil Eng., U. Western Civil and Mining Eng., Univ. of Sydney, June,
Ontario, August. 254 p.

El Sharnouby, B. and Novak, M. {1984). "Dynamic Hijikata, K., Miura, K., Miyamoto, Y. and Moroi,
Experiments With Group of Piles", J. Geotech. T. (1990). "Dynamic Characteristics of Pile-
Eng., ASCE, June, vol. 110, no. 6, 719-737. Groups Foundation (in Japanese)", Proc. of Japan
Conf. of Struc. Div. in Archit., No. 408, 89-
El Sharnouby, B. and Novak, M. (1985). "Static 98.
and Low Frequency Response of Pile Groups",
Canadian Geotech. J., vol. 22, no. 1, 79-94. Janes, M. and Novak, M. {1989). "Pile Group
Response to Large Lateral Loads", Proc. 42nd
El Sharnouby, B. and Novak, M. (1986). "Flexi- Canadian Geotech. Conf. , Winnipeg, Manitoba,
bility Coefficients and Interaction Factors for Oct. 23-25, 216-222.
Pile Group Analysis", Canadian Geotech. J., vol.
23, 441-450. Kana, D.O., Boyce, L. and Blaney, G.W. (1986).
"Development of a Scale Model for Dynamic
El Sharnouby, B. and Novak, M. (1990). "Stiff- Interaction of a Pile in Clay", J. Energy
ness Constants and Interaction Factors for Resources Technology, vol. 108, 254-261.
Vertical Response of Pile Groups", Canadian
Geotech. J., (in press). Kausel, E., Roesset, J .M. and Waas, G. (1975).
"Dynamic Analysis of Footings on Layered Media",
Fellenius, B.H. (Ed.) (1988). Proc. Third Int. J. Eng. Mech. Div., ASCE, vol. 101, no. EMS,
Conf. on Application of Stress-Wave Theory to 679-693.
Piles. Ottawa, Canada, p. 926.
Kaynia, A.M. (1982a). "Dynamic Stiffness and
Finn, W.D.L. and Martin, G.R. (1980). "Offshore Seismic Response of Pile Groups", Ph.D. Thesis,
Pile Foundations in Sand Under Earthquake MIT, Cambridge, MA.
Loading", Appl. ocean Res., vol. 2, no. 2, 81-
84. Kaynia, A.M. (1982b). "Dynamic stiffness and
Seismic Response of Pile Groups", MIT Res. Rep.
Flores-Berrones and Whitman, R.V. (1982). R82-03, Cambridge, MA.
"Seismic Response of End-Bearing Piles", J. of
Geotech. Eng. Div., ASCE, vol. 108, no. 4, 554- Kaynia, A.M. (1988). "Characteristics of the
569. Dynamic Response of Pile Groups in Homogeneous
and Nonhomogeneous Media", Proc. 9th World Conf.
Gazetas, G. (1984). "Seismic Response of End- on Earthq. Eng., Tokyo-Kyoto, Japan, Aug. 2-9,
Bearing Single Piles", Soil Dyn. and Earthq. vol. III, 575-580.
Eng., vol. 3, no. 2, 82-93.
Kaynia, A.M. and Kausel, E. (1982). "Dynamic
Gazetas, G. and Dobry, R. (1984). "Horizontal Behavior of Pile Groups", 2nd Int. Conf. on Num.
Response of Piles in Layered Soils", J. Geotech. Methods in Offshore Piling, Austin, TX, 509-532.
Eng. Div. ASCE, vol. 101, no. 1, 20-40.
Kaynia, A.M. and Kausel, E. (1990). "Dynamics of
Gazetas, G. and Makris, N. (1991). "Dynamic Piles and Pile Groups in Layered Soil Media",
Pile-Soil-Pile Interaction, I: Analysis of Axial Int. J. Soil Dyn. & Earthq. Eng., in print.
Vibration", J. Earthq. Eng. & Struct. Dyn., vol.
20, no. 2 (in press).

2452
Kishida, H., Suzuki, Y. and Nakai, s. (1985). Mamoon, S.M., Kaynia, A.M. and Banerjee, P.K.
"Behavior of a Pile Under Horizontal Cyclic (1990). "On Frequency Domain Dynamic Analysis
Loading", Proc. XIIth ICSMFE, San Francisco, of Piles and Pile Groups", J. Eng. Mech. Di v. ,
vol. 11, 1413-1416. ASCE, vol. 116, no. 10, to appear.

Kobori, T., Minai, R. and Baba, K. (1977). Masuda, K., Saseki, F., Urao, K., Veno, K. and
"Dynamic Behaviour of a Laterally Loaded Pile", Miyamoto, Y. (1986). "Simulation Analysis of
Proc. of Specialty Session 10, 9th ICSMFE, Forced Vibration Test of Actual Pile Foundation
Tokyo, July 14, 175-180. by Thin Layer Method", Proc. Annual Meeting of
Architectural Inst. of Japan.
Kobori, T., Minai, R. and Baba, K. (1981).
"Dynamic Behavior of a Pile Under Earthquake Matlock, H. and Foo, S.H.C. (1980). "Axial
Type Loading", Proc. Int. Conf. on Recent Analysis of Piles Using a Hysteretic Degrading
Advances in Geotech. Earthq. Eng. and Soil Dyn. Soil Model", Proc. Int. Symp. Numer. Methods
U. of Missouri-Rolla, vol. 2, 795-800. ' Offshore Piling, Institute of Civil Engineers,
London, 127-133.
Krishnan, R., Gazetas, G. and Veliz, A. (1983).
"Static and Dynamic Lateral Deflection of Piles Matlock, H., Foe, H.C. and Bryant, L.M. (1978).
in Non-Homogeneous Soil stratum" I Geotechnique I "Simulation of Lateral Pile Behaviour Under
vol. 33, no. 3, 307-325. Earthquake Motion", proc. Am. Soc. Civ. Engrs.
Specialty Conf. on Earthq. Eng. & Soil Dyn.,
Kuhlemeyer, R.L. (1976). "Static and Dynamic Pasadena, CA, II, 600-619.
Laterally Loaded Piles", Res. Rep. No. CE76-9,
Dept. of Civil Eng., U. of Calgary, Canada. Matlock, H., Martin, G.R., Lam, I.P. and Tsai,
c.F. (1981). "Soil-Pile Interaction in
Kuhlemeyer, R.L. (1979a). "Vertical Vibration of Liquefiable Cohesionless Soils During Earthquake
Piles", J. Geotech. Eng. Div., ASCE, vol. 105, Loading", Proc. 1st Int. Conf. Recent Adv. in
273-287. Geot. Earthq. Eng. & Soil Dyn., St. Louis,
Missouri.
Kuhlemeyer, R. L. ( 1979b). "Static and Dynamic
Laterally Loaded Floating Piles", J. Geotech. Mitwally, H. and Novak, M. (1988). "Pile Driving
Eng. Div., ASCE, val. 105, 289-304. Analysis Using Shaft Models and FEM'', Proc. 3rd
Int. Conf. on Appl. of Stress-Wave Theory to
Lakshmanan, N. and Minai, R. (1981). "Dynamic Piles, Ottawa, Canada, 455-466.
Soil Reactions in Radially Non-Homogeneous Soil
Media", Bul. Disaster Prevention Res. Inst. , Mitwally, H. and Novak, M. (1987). "Response of
Kyoto University, Vol. 31, 79-114. Offshore Towers With Pile Interaction", J. of
Eng. Mech., July, vol. 113, no. 7, 1065-1084.
Liu, W. and Novak, M. (1990). "Pile-Soil static
Interaction Analysis by Finite and Infinite Mizuhata, K. and Kusakabe, K. (1984). "Comparison
Elements", Res. Rep. GEOT-19-90, Dept. Civil of Experimental and Analytical Results of
Eng., Univ. West. Ontario, London, Canada. Vibration of a Full scale Pile", Proc. 8th WCEE,
San Francisco, vol. III, 633-640.
Makris, N., Gazetas, G. and Fan, K. (1989).
"Analytical Results for Pile-Soil-Pile Mizuno, H. (1987). "Pile Damage During Earth-
Interaction in Vertical Harmonic Motion", ERCAD quakes in Japan (1923-1983)", Dynamic Response
Conf., Berlin. of Pile Foundations, ASCE, Geotech. Special
Publ. No. 11 (Edited by T. Nogami) 53-78.
1

Mamoon, S.M. (1990). "Dynamic and Seismic

Behavior of Deep Foundations", Ph.D. Thesis, Mizuno, H., Iiba 1 M. and Kitagawa, Y. (1984).
Fac. of Grad. School, State Univ. of New York at "Shaking Table Testing of Seismic Building-Pile-
Buffalo, N.Y. Two-Layered-Soil Interaction", SWCEE, San
Francisco, vol. III, 649-656.
Mamoon, S.M. and Ahmad, S. (1990). "Response of
Single Piles to Obliquely Incident SH, sv and p Morrison, c.s. and Reese, L.C. (1988). "A
Waves", J. of Geotech. Eng. Div., ASCE, val. Lateral-Load Test of a Full-Scale Pile Group in
116, no. 2, 186-204. Sand", Geot. Eng. Rep. GR86-1, Geotech. Eng.
Center, Univ. of Texas at Austin, TX.
Mamoon, S.M. and Banerjee, P.K. (1990a).
"Response of Piles and Pile Groups to Traveling Nogami, T. (1980). "Dynamic Stiffness and Damping
SH-Waves", J. of Earthq. Eng. and struct. Dyn., of Pile Groups in Inhomogeneous Soil" 1 Proc.
vol. 19, no. 4, 597-610. Session Dyn. Response Pile Found.: Anal.
Aspects ASCE, 31-52.
Mamoon, S.M. and Banerjee, P.K. (1990b).
"Dynamic Behavior of Piles and Pile Groups with Nogami, T. (1985). "Flexural Responses of
Cap Interaction", J. of Geotech. Eng. Div., Grouped Piles Under Dynamic Loading", Earthq.
ASCE, to appear. Eng. and struct. Dyn., val. 13, 321-336.

Mamoon, S.M., Banerjee, P.K. and Ahmad, s. Nogami, T. (Ed.) (1987). "Dynamic Response of
( 1988) • "Seismic Response of Pile Foundations", Pile Foundations Experiment, Analysis and
Tech. Rep. NCEER-88-0034, Nat. Center for Observation", ASCE, Geot. Spec. Publication,
Earthq. Eng. Res., SUNY, Buffalo. No. 11, 186 p.

2453
Nogami, T. and Konagai, K. (1987). "Dynamic Novak, M. and Mi twally, H. ( 1990) . "Random
response of vertically loaded nonlinear pile Response of Offshore Towers With Pile-Soil-Pile
foundations", J. Geotech. Eng., vol. 113, no. Interaction", J. Offshore Mech. and Arctic Eng. ,
2, 147-160. Feb., vol. 112, 35-41.

Nogami, T. and Konagai, K. (1988). "Time Domain Novak, M. and Nogami, T. (1977). "Soil Pile
Flexural Response of Dynamically Loaded Single Interaction in Horizontal Vibration", Earthq.
Piles", J. Eng. Mech. Div., ASCE, vol. 114, no. Eng. Struct. Dyn., vol. 5, 263-282.
9, 1512-1525.
Novak, M. and Sheta, M. (1980). "Approximate
Nogami, T. and Novak, M. (1976). "Soil-Pile Approach to Contact Problems of Piles", Proc.
Interaction in Vertical Vibration", J. Earthq. Geotech. Eng. Div. ASCE National Convention
Eng. & struct. Dyn., vol. 4, 277-293. "Dynamic Response of Pile Foundations:
Analytical Aspects", Florida, Oct. 30, 53-79.
Nogami, T., Konagai, K. and Otani, J. (1988).
"Non Linear Pile Foundation Model for Time- Novak, M. and Sheta, M. (1982). "Dynamic
Domain Dynamic Response Analysis", 9th WCEE, Response of Piles and Pile Groups", Proc. 2nd
Tokyo, vol. III, 593-598. Int. Conf. on Numerical Methods in Offshore
Piling, University of Texas at Austin, Texas,
Novak, M. (1974). "Dynamic stiffness and Damping April, 489-507.
of Piles", Canadian Geotech. J., vol. 11, No.
4, 574-598. Novak, M., Nogami, T. and Aboul-Ella, F. (1978).
"Dynamic Soil Reaction for Plane Strain Case",
Novak, M. (1977). "Vertical Vibration of J. Eng. Mech. Div., ASCE, vol. 104, no. 4, 953-
Floating Piles", J. Eng. Mech. Div. ASCE, Feb., 959.
vol. 103, no. EMl, 153-168.
O'Neill, M.W. and Dobry, R. (Eds.) (1980).
Novak, M. (1980). "Soil-Pile Interaction Under "Dynamic Response of Pile Foundations:
Dynamic Loads", Proc. Num. Methods in Offshore Analytical Aspects", ASCE, 112 p.
Piling, ICE, London, 59-68.
Ostadan, F. (1983). "Dynamic Analysis of Soil-
Novak, M. and Aboul-Ella, F. (1978a). "Impedance Pile-structure Systems", Ph.D. Thesis, U~ of
Functions of Piles in Layered Media", J. Eng. California, Berkeley, Grad. Div.
Mech. Div. ASCE, June, vol. 104, no. EM3, Proc.
Paper 13847, 643-661. Pak, R.Y. and Jennings, P.C. (1987). "Elastodyna-
mic Response of Pile Under Transverse Excita-
Novak, M. and Aboul-Ella, F. (1978b). "Stiffness tions", J. Eng. Mech. Div., ASCE, vol. 113,
and Damping of Piles in Layered Media", Proc. 1101-1116.
Earthq. Eng. & Soil Dyn., ASCE Specialty Conf.,
Pasadena, CA, June 19-21, 704-719. Parmelee, R.A., Penzien, J., Scheffey, C.F.,
Seed, H.B. and Thiers, G.R. (1964). "Seismic
Novak, M. and El Hifnawy, L. (1984). "Effect of Effects on Structures Supported on Piles
Foundation Flexibility on Dynamic Behaviour of Extending Through Deep Sensitive Clays", Inst.
Buildings", 8th WCEE, San Francisco, vol. III, Eng. Res., U. of California, Berkeley, Rep. SESM
721-728. 64-2.

Novak, M. and El Sharnouby, B. (1983). "Stiff- Penzien, J. (1970). "Soil-Pile Foundation Inter-
ness Constants of Single Piles", J. Geotech. action", in Earthquake Engineering, Ed. R. L.
Eng. Div. ASCE, vol. 109, no. 7, 961-974. Wiegel, Prentice-Hall, Englewood Cliffs, New
Jersey, 349-381.
Novak, M. and El Sharnouby, B. (1984). "Evalua-
tion of Dynamic Experiments on Pile Groups", J. Penzien, J., Scheffey, C.F. and Parmelee, R.A.
Geotech. Eng., June, vol. 110, no. 6, 738-756. (1964). "Seismic Analysis of Bridges on Long
Piles", J. Eng. Mech. Div. ASCE, no. EM3, 223-
Novak, M. and Grigg, R. F. ( 1976) . "Dynamic 254.
Experiments With Small Pile Foundations",
Canadian Geotech. J., vol. 13, no. 4, 372-385. Poulos, H.G. (1968). "Analysis of Settlement of
Pile Groups", Geotechnique, vol. 18, 449-471.
Novak, M. and Han, Y. (1990). "Impedances of Soil
Layer With Boundary Zone", J. Geotech. Eng. , Poulos, H.G. (1971). "Behaviour of Laterally
June, vol. 116, no. 6, 1008-1014. Loaded Piles. II - Pile Groups", J. Soil Mech.
Foundations Div. ASCE, vol. 97, no. SM5, 733-
Novak, M. and Howell, J.F. (1977). "Torsional 751.
Vibration of Pile Foundations", J. Geotech. Eng.
Div. ASCE, April, vol. 103, no. GT4, 271-285. Poulos, H.G. (1979). "Group Factors for Pile-
Deflection Estimation", J. Geotech. Eng. Div.
Novak, M. and Howell, J.F. (1978). "Dynamic ASCE, no. GT12, 1489-1509.
Response of Pile Foundations in Torsion", J.
Geotech. Eng. Div. ASCE, vol. 104, no. GT5, 535- Poulos, H. (1980). "An Approach for the Analysis
552. of Offshore Pile Groups", Proc. Num. Methods in
Offshore Piling, ICE, London, 119-126.
Novak, M. and Janes, M. (1989). "Dynamic and
static Response of Pile Groups", Proc. 12th Int. Poulos, H.G. and Davis, E.H. (1980). Pile
conf. on Soil Mech. Found. Eng. , Rio de Janeiro, Foundations Analysis and Design, John Wiley &
August, 1175-1178. Sons, 397 p.

2454
Prakash, S.F. and Chandrasekaran, V. (1977).' Tajimi, H. (1966;1969). "Earthquake Response of
"Free Vibration Characteristics of Piles", Proc. Foundation Structures", Rep. of Fac. Sci. Eng.,
9th Int. Conf. Soil Mech. Found. Eng., Tokyo, Nihon University, 1966.3, 1.1-3. 5 (in Japanese) .
vol. 2, 333-336. See also Taj imi, H. "Dynamic Analysis of a
Structure Embedded in an Elastic Stratum", Proc.
Prakash, s. and Sharma, H.D. (1990). "Pile 4th WCEE, Chile.
Foundations in Engineering Practice", Wiley,
Chap. 7. Tajimi, H. (1977). "Seismic Effects on Piles",
Proc. of Specialty Session 10, State-of-the-Art
Rajapakse, R.K.N.D. and Shah, A.H. (1987a). "On Rep., 9th ICSMFE, Tokyo, July 14, 15-26.
the Longitudinal Harmonic Motion of an Elastic
Bar Embedded in an Elastic Half Space", Int. J. Tajimi, H. and Shimomura, Y. (1976). "Dynamic
Solids Structures, vel. 23, 267-285. Analysis of Soil-Structure Interaction by Thin
Layered Element Method", Trans. Archit. Inst. of
Rajapakse, R.K.N.D. and Shah, A.H. (1987b). "On Japan, May, No. 243 1 41-51 (in Japanese). Also:
the Lateral Harmonic Motion of an Elastic Bar Rep. "Dynamic Analysis of a single Pile Embedded
Embedded in an Elastic Half-Space", Int. J. in Horizontally Layered Soils", College of Sci.
Solids Structures, vel. 23, 287-303. & Tech., Nihon Univ., Tokyo.
Rajapakse, R.K.N.D. and Shah, A.H. (1989). Takemiya, H. and Yamada, Y. (1981). "Layered
"Impedance Curves for an Elastic Pile", Soil Soil-Pile-Structure Dynamic Interaction",
Dyn. and ~arthq. Eng., vel. 8, no. 3, 145-152. Earthq. Eng. and structural Dyn., vel. 9, 437-
458.
Roesset, J .M. (1980). "Stiffness and Damping
Coefficients of Foundations", Proc. of Session Terzaghi, K. and Peck, R.B. (1967). "Soil
on Dyn. Response of Pile Found.: Analytical Mechanics in Engineering Practice", 2nd Ed., New
Aspects, ASCE National Convention, Florida, York: Wiley.
Oct., 1-30.
Trochanis, A.M., Bielak, J. and Christiano, P.
Roesset, J .M. and Angelides, D. (1979). "Dynamic ( 1988). "A Three-Dimensional Nonlinear Study of
Stiffness of Piles: Numerical Methods in Piles Leading to the Development of a Simplified
Offshore Piling", Inst. of Civil Engrs., London, Model", Res. Rep. R-88-176, Dept. of Civil Eng.,
75-82. Carnegie Mellon Univ., Pittsburgh, PA, 175 p.

Roesset, J. M. , Stokoe, K. H. , Baka, J. E. and Kwok, Waas, G. and Hartmann, H.G. (1981). "Pile
S.T. (1986). "Dynamic Response of Vertically Foundations Subjected to Dynamic Horizontal
Loaded Small-Scale Piles in Sand", Proc. 8th Loads", European simulation Meeting "Modelling
European Conf. Earthq. Eng., Lisbon, vol. 2, and Simulation of Large Scale Structural
5.6/65-72. systems", Capri, Italy, Sept., (also SMIRT,
Paris), 17 p.
Sanchez-Salinero, I. (1982). "Static and Dynamic
Stiffnesses of Single Piles", Geotech. Eng. Rep. Waas, G. and Hartmann, H.G. (1984). "Seismic
GR82-31, Civil Eng. Dept., U. of Texas at Analysis of Pile Foundations Including Pile-
Austin, August, 225 p. Soil-Pile Interaction", Proc. 8WCEE, San
Francisco, vol. v, 55-62.
Sen, R., Davies, T.G. and Banerjee, P.K. (1985).
"Dynamic Analysis of Piles and Pile Groups Wolf, J.P. (1985). "Dynamic soil-structure
Embedded in Homogeneous Soils", Earthq. Eng. Interaction", Prentice-Hall Inc.
Struct. Dyn., vel. 13, 53-65.
Wolf, J.P. and von Arx, G.A. (1978). "Impedance
Sen, R., Kausel, E. and Banerjee, P.K. (1985). Functions of a Group of Vertical Piles", Proc.
"Dynamic Analysis of Piles and Pile Groups in ASCE Specialty Conf. on Earthq. Eng. & Soil
Non-Homogeneous Soils", Int. J. Num. and Anal. Dyn., Pasadena, CA, vol. II, 1024-1041.
Meth. in Geomech., vel. 9, 507-524.
Wolf, J.P. and von Arx, G:A· (1982). "Horizon-
Sheta, M. and Novak, M. (1982). "Vertical tally Travelling Waves ~n a Group of Piles
Vibration of Pile Groups", J. Geotech. Eng. Di v. Taking Pile-Soil-Pile Interaction Into Account",
ASCE, April, vol. 108, no. GT4, 570-590. Earthq. Eng. & Struct. Dyn., vel. 10, no. 2,
225-237.
Shimizu, N., Yamamoto, s. and Koori, Y. (1977).
"Three-Dimensional Dynamic Analysis of Soil- Wolf, J.P. and Weber, B. (1986). "Approximate
Structure System by Thin Layer Element Method", Dynamic stiffness of Embedded Foundation Based
Trans. Archit. Inst. of Japan, March, no. 253, on Independent Thin layers With Separation of
31-41. Soil", Proc. 8th European Conf. on Earthq. Eng.,
Lisbon, vel. 2, 5.6/33-5.6/40.
Sooysmith, c. (1896). "Concerning Foundations
for Heavy Buildings in New York City", Trans. Wolf, J.P., von Arx, G.A., de Barros, F.C.P. and
ASCE, Vol. 35, 459-469. Kakubo, M. (1981). "Seismic Analysis of the
Pile Foundation of the Reactor Building on the
Swane, I.e. and Poulos, H. G. (1982). "A NPP Angra 2", Nuclear Eng. & Design, val. 65,
Theoretical study of the Cyclic Shakedown of no. 3, 329-341.
Laterally Loaded Piles", Res. Rep. No. R415,
Univ. of Sydney, School of Civil and Mining
Eng., July.

2455
Woods, R.D. (1984). "Lateral Interaction Between
Soil and Pile", Proc. Int. Symp. on Dyn. Soil-
Structure Interaction, Minneapolis, MN, Sept.,
47-54.
Yan, L. (1990). "Hydraulic Gradient Similitude
Method for Geotechnical Modelling Tests With
Emphasis on Laterally Loaded Piles", Ph.D.
Thesis, Fac. of Grad. Studies, Univ. of British
Columbia, Canada.

Papers Presented to This conference (1991)

Ahmad, s. "Seismic response of floating piles to
obliquely incident waves".
Arya, A. and Arya, A.S. "Pile group stiffness
for seismic soil-structure interaction".
Baba, K. "Dynamic analysis of soil-piles
interaction systems under earthquake type
loading".
Kagawa, T. "Seismic response of axially loaded
pile group".
Kobori, T., Nakazawa, M., Hijikata, K.,
Kobayashi, Y., Miura, K., Miyamoto, Y. and
Moroi, T. "Study on dynamic characteristics of
a pile group foundation".
Nogami, T., Jones, H.W. and Mosher, R.L.
"Seismic response analysis of pile supported
structure: Assessment of commonly used
approximations".
Nomura, S., Tokimatsu, K. and Shamoto, Y. "Soil-
pile-structure interaction during liquefaction".
Otani, J., Nogami, T. and Konagai, K. "Non-
linear time domain numerical model for pile
group under transient dynamic forces".
Purkayastha, R.D. and Dey, s. "Behaviour of
cyclically loaded model piles in soft clay".
Wu, S., Chen, Y., Cai, Y. and Chen, L. "Analysis
of pile-soil dynamic interaction by combination
of BEM and FEM''.

2456