KINEMATIC SEISMIC RESPONSE O F SINGLE PILES

AND P I L E GROUPS
By Ke Fan, 1 George Gazetas, 2 Amir Kaynia, 3 Eduardo Kausel, 4
and Shahid Ahmad 5
Downloaded from ascelibrary.org by HARVARD UNIVERSITY on 09/26/13. Copyright ASCE. For personal use only; all rights reserved.

ABSTRACT: Results of a numerical study are presented on the kinematic response

of groups of vertical floating piles connected through rigid massless caps and sub-
jected to vertically propagating harmonic 5-waves. Pile-soil and pile-pile interaction
effects are modeled rigorously. Parametric results for the "effective seismic pile-
cap motion," normalized by the "free-field" ground-surface motion, are displayed
in dimensionless form for a number of typical pile-group configurations, in three
idealized soil profiles: a homogeneous half-space, a half-space with modulus pro-
portional to depth, and a two-layered stratum. It is shown that, whereas the influ-
ence of the nature of the soil profile is profound at all frequencies, the effects of
pile-group configuration, number of piles in the group, and relative spacing between
piles are usually insignificant for lateral displacements, but quite important for pile-
cap rotations. Pile-head "fixity" conditions and the pile/soil modulus ratio are found
to affect appreciably and in a similar way the seismic response of a single pile and
of a pile group.

INTRODUCTION

Consider a vertical piled foundation supporting a structure and subjected

to upward propagating ("incident") seismic waves. While an extremely flex-
ible pile might simply follow the seismic motion of the ground, real piles in
general "resist" and, hence, modify soil deformations. As a result, the
incident seismic waves are "scattered" and the seismic excitation to which
the structure base is effectively subjected differs from the free-field motion
and may generally include rotational (in addition to translational) compo-
nents. In turn, piles experience bending, axial, and shearing stresses, which
depend on their overall rigidity relative to the soil. This wave-induced in-
terplay between soil and piles is also affected by the kinematic constraints
imposed at the head of the piles from the cap and the superstructure, and
will be called hereafter kinematic soil-pile-foundation seismic interaction,
or simply kinematic/seismic response.
While, in real-life situations, rigorously modeling all the factors influ-
encing the kinematic response is a formidable task, (especially if nonver-
tically propagating seismic waves are expected to impinge on the piles, or
if substantial soil nonlinearities are likely to develop) practically useful re-
sults and valuable insight to the mechanics of soil-pile interaction during
earthquake shaking can be obtained from proper linear analyses or simple
idealized systems.
Analyses of the kinematic seismic response of single piles and of pile
groups have been reported by Blaney et al. (1976), Takemiya and Yamada
'Grad. Student of Civ. Engrg., State Univ. of New York, Buffalo, NY 14260.
2
Prof. of Civ. Engrg., State Univ. of New York, Buffalo, NY 14260.
3
Assoc. Prof, of Civ. Engrg., Isfahan Univ. of Tech., Isfahan, Iran.
4
Prof. of Civ. Engrg., Massachusetts Inst, of Tech., Cambridge, M A 02139.
5
Assoc. Prof, of Engrg., State Univ. of New York, Buffalo, NY 14260.
Note. Discussion open until May 1, 1992. To extend the closing date one month,
a written request must be filed with the ASCE Manager of Journals. The manuscript
for this paper was submitted for review and possible publication on September 15,
1990. This paper is part of the Journal of Geotechnical Engineering, Vol. 117, No.
12, December, 1991. ©ASCE, ISSN 0733-9410/91/0012-1860/$1.00 + $.15 per page.
Paper No. 26457.

1860

J. Geotech. Engrg. 1991.117:1860-1879.

 SOIL PROFILES

(a) (b) (c)

Soil Young's Modulus

Downloaded from ascelibrary.org by HARVARD UNIVERSITY on 09/26/13. Copyright ASCE. For personal use only; all rights reserved.

FIG. 1. Sketch of System and Soil Profiles Studied

FIG. 2. Idealized General Shape of Kinematic Displacement Factor, /„ = /„(«„)>

Explaining Transition Frequency Factors aal and ao2.

(1981), Flores-Berrones and Whitman (1982), Wolf and Von Arx (1978),
Kaynia and Kausel (1982), Gazetas (1984), Barghouthi (1984), and Tazoh
et al. (1988). A recent overview has been presented by Novak (1991).
Continuum as well as Winkler-type models were developed/used in these
studies. However, the amount of published parametric results is very limited,
especially for pile groups, and a number of questions remain unanswered.
By contrast, a wealth of data is presently available for the response of single
piles and pile groups under dynamic force-and-moment loading at the top;
such a loading arises during earthquakes from the inertial forces developing
on the superstructure and the pile cap.
To fill this gap, this paper presents a comprehensive set of dimensionless
1861

J. Geotech. Engrg. 1991.117:1860-1879.

 SEISMIC RESPONSE OE SINGLE PILE
Downloaded from ascelibrary.org by HARVARD UNIVERSITY on 09/26/13. Copyright ASCE. For personal use only; all rights reserved.

° V,

FIG. 3. Effect of Lid and EJES on Kinematic Seismic Response of Single Free-
Head Piles in Homogeneous Soil (p^p,, = 0.7, p = 0.05, and v„ = 0.4)

graphs that could be readily utilized in practical applications. Moreover, a

comparative study of these graphs leads to some interesting conclusions that
may contribute towards an improved appreciation of the nature of seismic
pile-soil-pile interaction.

PROBLEM DEFINITION AND METHODS OF SOLUTION

Three categories of groups of floating vertical piles are studied in this

paper:

• A single free-head or fixed-head pile.

• A rigidly capped pile group consisting of two, three, four, six, or nine
piles in a row.
1862

J. Geotech. Engrg. 1991.117:1860-1879.

 SEISMIC RESPONSE OF SINGLE PILE
Downloaded from ascelibrary.org by HARVARD UNIVERSITY on 09/26/13. Copyright ASCE. For personal use only; all rights reserved.

v.cn
FIG. 4. Effect of Lid and EJES{L) on Kinematic Seismic Response of Single Free-
Head Piles in Nonhomogeneous Soil ( p ^ = 0.7, 3 = 0.05, and vs = 0.4)

• A rigidly capped square group of 2 x 2, 4 x 4, or 6 X 6 piles.

All piles, of diameter d and length L, are considered to be linear elastic

beams with constant Young's modulus, Ep, and mass density, p They are
embedded in three different soil deposits (Fig. 1) in which the Young's
modulus: (1) is a constant (Es); (2) is proportional to depth [Es(z) =
ES(L) • z/L]; or (3) has two distinct values (Esl and Es2) above and below
a depth z = L/2, respectively, with Esl = 0.10£,2 or Esl = 0.30£j2. In all
cases, the soil is assumed to be a linear hysteretic continuum with constant
Poisson's ratio vs, constant material density p^ = 0.7 p P , and constant hys-
teretic damping (3,.
Each pile-foundation-soil system is excited by vertically propagating
harmonic shear (5) waves, which would produce a horizontal oscillation
Uffexp (iutt) at a "free-field" point of the ground surface, i.e., at a location
1863

J. Geotech. Engrg. 1991.117:1860-1879.

 SINGLE PILE L/d = 20

(a)
Downloaded from ascelibrary.org by HARVARD UNIVERSITY on 09/26/13. Copyright ASCE. For personal use only; all rights reserved.

" u.

0 FREE HEAD Ep/£s=1000

" O FREE HEAD Ep/Es=10000

* FIXED HEAD Ep/Es=1000
1 FIXEO HEAD Ep/Es=10000

V,(.L)
FIG. 5. Influence of Pile Fixity on Kinematic Seismic Response of Single Piles in
(a) Homogeneous; and (b) Nonhomogeneous Soils (Lid = 20, p,/pp = 0.7, p = 0.05,
and vs = 0.4)

unaffected by the presence of the piles. Around the piles the "perturbed"
wave field is a complicated combination of incident (upward propagating)
waves, reflected-at-the-surface (downward propagating) waves, and waves
diffracted by the piles and propagating in a primarily horizontal direction.
A rigorous method of solution to this three-dimensional dynamic boundary-
value problem has been developed by Kaynia and Kausel (1982). This method
is in essence a boundary-integral-type formulation in which the Green's
functions, defining the displacement fields due to uniform unit loads acting
on an elemental cylindrical surface and on a circular disk, are computed by
solving the wave equations through Fourier and Hankel transformations.
1864

J. Geotech. Engrg. 1991.117:1860-1879.

 SINGLE PILE Ep/Es1=5000 L/d = 20
Free-Bead Pile \
Downloaded from ascelibrary.org by HARVARD UNIVERSITY on 09/26/13. Copyright ASCE. For personal use only; all rights reserved.

FIG. 6. Kinematic Seismic Response of Single Free-Head Pile in Two-Layered

Soil with Different EJE,2 (EpIEsX = 5,000, Lid = 20, pjpp = 0.7, 0 = 0.05 and vs
= 0.4)

These functions yield the dynamic soil flexibility matrix that is combined
with the analytically derived pile flexibility matrix, while enforcing com-
patibility of deformations at the pile-soil interface. For an arbitrary pile
group, the method gives the horizontal translation and rotation of the mass-
less cap, both of which are complex-valued functions of frequency, for a
given soil profile. The Kaynia and Kausel (1982) formulation was used to
obtain most of the results presented herein. In addition, finite-element
(Blaney et al. 1976) and boundary-element (Ahmad 1986) formulations were
used in a few cases for obtaining the response of single piles, and for es-
tablishing the very small sensitivity of the results to using different numerical
solution procedures. A simple model developed for the problem at hand
by Makris and Gazetas (1993) is also used for qualitatively explaining some
of the trends observed in the numerical results.

RESULTS AND OBSERVED TRENDS

The effects of soil-piled-foundation kinematic interaction are portrayed

in the form of two kinematic response factors
1865

J. Geotech. Engrg. 1991.117:1860-1879.

 Normalized Displacements Normalized Displacements
Soil Profile

20 -20 20 0 E„
Downloaded from ascelibrary.org by HARVARD UNIVERSITY on 09/26/13. Copyright ASCE. For personal use only; all rights reserved.

pile

free field

a„-0.05

FIG. 7. Distribution of Free-Field Soil and Pile Displacements along Depth of Two-
Layered Soil (EJEs2 = 0.1, Lid = 20, p,/p,, = 0.7, (3 = 0.05, and vs = 0.4) aa = 0.05

Normalized Displacements Normalized Displacements

Soil Profile

10 0 Ec

pile

a 0 = 0.15

REAL PART IMAG. PART

FIG. 8. Distributions of Free-Field Soil and Pile Displacements along Depth of

Two-Layered Soil (EJEs2 = 0.1, Lid = 20, pjpp = 0.7, (3 = 0.05, and vs = 0.4) a0
= 0.15

\UP \%\d
/,. = and L, (1)
u„
Jff U„
Uff
plotted as functions of the frequency factor

a = (2)
° v*
1866

J. Geotech. Engrg. 1991.117:1860-1879.

 Normalized Displacements Normalized Displacements
Soil Profile

5 0 E,

pile . free field ^ -. free field

Downloaded from ascelibrary.org by HARVARD UNIVERSITY on 09/26/13. Copyright ASCE. For personal use only; all rights reserved.

! pile ";

a0=-0.45

FIG. 9. Distributions of Free-Field Soil and Pile Displacements along Depth of

Two-Layered Soil (EJEs2 = 0.1, Lid = 20, ps/pp = 0.7, 0 = 0.05, and vs = 0.4) aa
= 0.45

where V* = a characteristic value of the soil S-wave velocity profile (in

this paper V* is taken equal to Vs for the homogeneous profile, to VS(L)
for the linearly inhomogeneous profile, and to Vsl for the two-layered pro-
file. The horizontal displacement, Up, and angle rotation, <£>p, at pile-cap
level constitute the "effective pile-cap input motion"; they are complex
numbers because of the generation of both radiation damping (due to dif-
fracted waves spreading away from each pile) and material damping (due
to hysteretic action in the soil). Only the amplitudes (absolute values) of
Up and <£>p are used in the definitions of kinematic response factors [(1)].
The general shape of /„ versus frequency is plotted in Fig. 2.
The dimensionless parameters whose influence on /„ and /,,, is investigated
for a given pile group and soil profile include:

• The ratio of the effective pile modulus to a characteristic soil Young's

modulus.

(3)
E; E,(LY E,X

for the homogeneous, inhomogeneous, or layered profiles of Fig. 1,

respectively.
• the ratio of the spacing between the closest piles to the diameter
of the piles, sld (hereafter called simply spacing ratio).
• the pile slenderness ratio, Lid.

A few results are also displayed in the form of distributions along the
depth of kinematic pile displacements and free-field soil displacements.
1867

J. Geotech. Engrg. 1991.117:1860-1879.

 2 PILES Ep/Es=1000 L / d = 20
Downloaded from ascelibrary.org by HARVARD UNIVERSITY on 09/26/13. Copyright ASCE. For personal use only; all rights reserved.

/.. = \u.
:

0.06-:
\±\d

- B D • D B-B—-j]

"TQ oo o tj> o s ft^ 8 8....$

0.2 0.3
cod
°° = ~
FIG. 10. Effect of Pile Spacing on Kinematic Interaction of Relatively Compress-
ible Fixed-Head Piles in Homogeneous Soil (EPIES = 1,000, Lid = 20, ps/pp = 0.7,
P = 0.05, and v, = 0.4)

The dimensionless graphs are arranged according to pile configuration as

follows:

• Figs. 3-9 display results from a single pile.

• Figs. 10-18 are for groups of piles in a row.
• Figs. 19-22 are for square pile groups.

The presented figures fin conjunction with the results of Kaynia and
Kausel (1982), of Gazetas (1984), and of additional analyses by the present
authors that are not included herein for lack of space], reveal the following
significant trends.

1. The general shape of the kinematic displacement factor, /„ = Iu(a0), is

idealized in the sketch of Fig. 2. It consists (for single piles and pile groups) of
three fairly distinct regions in the frequency range of greatest interest for earth-
quake loading (a0 < 0.5):
1868

J. Geotech. Engrg. 1991.117:1860-1879.

 2 PILES Ep/Es= 10000 L/d = 20
Downloaded from ascelibrary.org by HARVARD UNIVERSITY on 09/26/13. Copyright ASCE. For personal use only; all rights reserved.

FIG. 11. Effects of Pile Spacing on Kinematic Interaction of Relatively Rigid Fixed-
Head Piles in Homogeneous Soil (EPIE, = 10,000, Lid = 20, pjpp = 0.7, p = 0.05,
and vs = 0.4)

A low-frequency region (0 < a0 < aal) in which I„ = 1—the end result

of the pile(s) following closely the relatively large-wavelength defor-
mations of the ground, as illustrated in Fig. 7 for a pile in the two-
layered soil with Esl/Es2 = 0.10.
An intermediate-frequency region (aoX < a0 < ao2) characterized by /„
declining rapidly with frequency—a direct consequence of the pro-
gressively increasing incompatibility between the "wavy" pattern of soil
movements and the flexurally deforming pile(s), as illustrated in Fig. 9
for a pile in the two-layered soil with EJEs2 = 0.10.
A relatively high-frequency region a0 > ao2 in which Iu(a0) fluctuates
around an essentially constant value of about 0.20-0.40—since at such
frequencies the increasing "waveiness" of the soil deformations is largely
counterbalanced by the generally decreasing amplitude of the ground-
surface motion. Indeed, a„ > ao2 corresponds to seismic-excitation fre-
quencies, /, being of the order of five to ten times greater than the
natural frequency, /„, of the soil deposit in S-waves; and it is well known
from the one-dimensional soil amplification theory (Roesset 1977) that

1869

J. Geotech. Engrg. 1991.117:1860-1879.

 2 PILES Ep/Es(L) = 500 L/d = 20
Downloaded from ascelibrary.org by HARVARD UNIVERSITY on 09/26/13. Copyright ASCE. For personal use only; all rights reserved.

V.W

FIG. 12. Effect of Pile Spacing on Kinematic Interaction of Relatively Compress-

ible Fixed-Head Piles in Nonhomogeneous Soil (EP/E,(L) = 500, Lid = 20, ps/pp =
0.7, 0 = 0.05, and i\ = 0.4)

a linearly hysteretic soil would experience at such frequencies Uff values

not very different from the values of the base-rock displacement).

The reader should recall that kinematic displacements induced on rigid shallow
foundations by seismic waves impinging at an angle (with or without random
phase "incoherence") follow a qualitatively similar trend versus frequency with
that discussed above for the kinematic displacements atop piles and pile groups.
Excellent presentations on this subject for shallow foundations can be found in
Scanlan (1976), Luco and Wong (1986), Pais and Kausel (1985), and Veletsos
and Prasad (1988).
2. Whereas this general shape of /,,(«„) is approximately valid in all studied
cases, four factors seem to affect significantly the transition frequencies aal and
ao2: (1) The type of soil profile; (2) the relative rigidity of the pile; (3) the pile-
head fixity conditions; and (4) the pile slenderness.
Specifically, the significant factor controlling the magnitude of aol and ao2,
and thereby the kinematic response of single piles and pile groups, is the nature
1870

J. Geotech. Engrg. 1991.117:1860-1879.

 2 PILES Ep/Es(L) = 5000 L/d=20
Fixed-Head Piles
1 ^ s/d = 3 :_
v
\\ s/d = 5 :
0.8- i\ .o s/d=10 :-
Downloaded from ascelibrary.org by HARVARD UNIVERSITY on 09/26/13. Copyright ASCE. For personal use only; all rights reserved.

\\ o s/d=20 :
0.6-
TO o s/d=40 j

0.4- :-

0.2-

0,0- " ' 1 -

l*ld

0.00+

• V.CD
FIG. 13. Effect of Pile Spacing on Kinematic Interaction of Relatively Rigid Fixed-
Head Piles in Nonhomogeneous Soil (EPIES{L) = 5,000, Lid = 20, pjpp = 0.7, p =
0.05, and vs = 0.4)

of the soil profile as expressed by the variation of soil modulus, Es, versus depth.
In strongly nonhomogeneous deposits, as the one having modulus proportional
to depth, a0l is very small—of the order of merely 0.05 (e.g., Figs. 4, 12, 13,
16, 17, 21, and 22), depending of course on the value of the other three factors.
By contrast, in a homogeneous stratum or in a stratum with a thick homogeneous
top layer, aDl may be as high as 0.20-0.30 (e.g., Figs. 3, 5, 6,10, and 19). (Thus,
in terms of actual frequencies, co, in deposits with the same average wave velocity
the decaying branch of /„ will start earlier—by a factor of about two—in non-
homogeneous profiles.) Similarly, ao2 is about 0.10-0.20 in the linear-modulus
profile "b" compared to ao2 usually exceeding 0.40 in the two other profiles "a"
and "c." The practical implications of these differences are worthy of note: In
nonhomogeneous profiles, piles and pile groups will depress a much wider
spectrum of the harmonic components of the incident seismic excitation (and
thereby their heads will experience smaller "effective" horizontal motions) than
pile(s) in a homogeneous soil.
The relative rigidity of the pile(s), expressed through the aforementioned

1871

J. Geotech. Engrg. 1991.117:1860-1879.

 2 PILES Ep/Es1=5000 ES1/ES2 = 0.3 L/d = 20
1.2+r^ I I I.
Downloaded from ascelibrary.org by HARVARD UNIVERSITY on 09/26/13. Copyright ASCE. For personal use only; all rights reserved.

I = ! !— 0.04
• U

FIG. 14. Effect of Pile Spacing on Kinematic Interaction of Relatively Rigid Fixed-
Head Piles in Two-Layered Soil {EpIEsl = 5,000, Lid = 20, pjpp = 0.7, p = 0.05,
and vs = 0.4) EJEs2 = 0.3

moduli ratios [(3)] also affects aal and ao2- As expected, the stiffer pile(s) is
more effective in depressing a seismic soil movement and, and hence, its ki-
nematic response is characterized by smaller values of aoX and ao2, compared
with those of the softer piles (e.g., see Figs. 3, 4, and 5, and compare Fig. 10
with Fig. 11 and Fig. 19 with Fig. 20).
Increasing the degree of fixity at the pile-cap level (from hinged- or free-head
to fixed-head piles) has an effect similar to the effect of increasing EpIEs: aol
and ao2 tend to decrease and, hence, the "effective" pile-cap input motion in
an earthquake excitation will tend to be less severe (see Fig. 5). An additional
influence of pile-head fixity conditions has been observed with pile(s) embedded
in homogeneous deposits and in deposits with a homogeneous top layer (like
profile "c"): free-head single piles and hinged-at-the-cap piles and pile groups
1872

J. Geotech. Engrg. 1991.117:1860-1879.

 2 PILES Ep/Es1=5000 Es1/Es2 = 0.1 L/d = 20
1.2 L I I I I.
Downloaded from ascelibrary.org by HARVARD UNIVERSITY on 09/26/13. Copyright ASCE. For personal use only; all rights reserved.

FIG. 15. Effect of Pile Spacing on Kinematic Interaction of Rigid Fixed-Head Piles
in Two-Layered Soil (Ep/Esl = 5,000, Lid = 20, p^/p,, = 0.7, p = 0.05, and v„ = 0.4)
EJE<, = 0.1

embedded in such soils experience in the low to intermediate frequency range

/„ values exceeding unity. This appears to be the only case where a small de-
viation from the aforescribed general shape of 4(a 0 ) has been observed. It
implies an "effective" pile top motion, \UP\, greater than the free-field Uff\ Fig.
8 elucidates this possibility by comparing the pile and soil displacement profiles
(both real, i.e., in phase, and imaginary, i.e., 90° out of phase, components of
displacement) for the two-layered profile with Esl/Es2 = 0.10. With fixed-head
piles, this tendency for larger pile-top motion is completely suppressed, as evi-
denced in Fig. 5 for a single pile and in numerous additional results for pile
groups (not shown herein).
3. Pile-group configuration ("row" versus "square"), number of piles in the
group (1, 2, 3, 36), and pile-spacing ratio (s/d = 3, 5 and 10) make little
1873

J. Geotech. Engrg. 1991.117:1860-1879.

 1*4 PILES Ep/Es(L) = 500 L/d = 20

Fixed-Head Piles
Downloaded from ascelibrary.org by HARVARD UNIVERSITY on 09/26/13. Copyright ASCE. For personal use only; all rights reserved.

FIG. 16. Kinematic Seismic Response of 1 x 4 Relatively Compressible Fixed-

Head Pile Groups in Nonhomogeneous Soil (EP/ES(L) = 500, Lid = 20, ps/pp = 0.7,
p = 0.05, and vs = 0.4)

difference on /„, in the low (a0 < aol) and intermediate (aol <a0< ao2) frequency
ranges. This conclusion is valid (within engineering accuracy) for most studied
soil profiles and relative pile rigidities. It implies that with seismic excitation
there is little pile-to-pile interaction in this frequency range, even for close pile
spacing. By contrast, with inertial excitation at the top, pile-to-pile interaction
has been shown to play a dominant role in the response of pile groups. (The
only small deviation from this general trend is depicted in Fig. 11 for two piles
in a homogeneous half-space.)
This lack of interaction between the piles of a group for very low frequency
factors occurs when seismic wavelengths are then so large that piles follow almost
exactly the free-field ground movements, and there is hardly any scattering of
the vertically propagating seismic waves.
However, the explanation is not so simple for low and intermediate frequency
factors, at which pile and (free-field) soil motions are different, and the incident
seismic waves are therefore diffracted by the piles. A qualitative explanation is
suggested herein with the help of Fig. 9, which compares the seismic displace-
ment profiles of the two-layered soil deposit with EJEs2 = 0.10 and of a pile
1874

J. Geotech. Engrg. 1991.117:1860-1879.

 Ep/Es(L)=5000 L/d=20 s/d=5
Downloaded from ascelibrary.org by HARVARD UNIVERSITY on 09/26/13. Copyright ASCE. For personal use only; all rights reserved.

« V,

0.00+

V.U)
FIG. 17. Kinematic Seismic Response of 1 x N Relatively Rigid Fixed-Head Pile
Groups in Nonhomogeneous Soil {EPIES(L) = 5,000, Lid = 20, ps/pp = 0.7, p =
0.05, and vs = 0.4; N = 1, 2, 3, 4, 6, and 9)

embedded in it, at a frequency a0 ~ ao2- We notice that the difference (Up -

Uff), which is responsible for the generation of waves at the pile-soil interface
("diffraction"), is practically zero within the bottom stiff layer. In the top soft
layer, however, this difference is positive from the ground surface down to z ~
Ad and negative for greater depths. Therefore, the diffracted waves start with
similar amplitudes but opposite phases from the two parts of the pile—a situation
reminiscent of rocking, which is conducive to considerable "destructive wave
interference" and hence to rapidly diminishing amplitudes away from the source
pile, leading eventually to only a small influence of one pile on another.
On the other hand, at high frequencies (a0 > aa2), pile spacing, total number
of piles, and pile-group configuration appear to affect (although perhaps not
significantly from a practical viewpoint) the fluctuations of the /„(#„) curves.
4. Although the foregoing discussion focused only on displacements, it is
important to recognize that both hinged- and fixed-head pile groups experience
cap rotations during seismic-wave excitation. Thus, a rocking component of
"effective" cap motion develops, the significance of which in rocking of slender
1875

J. Geotech. Engrg. 1991.117:1860-1879.

 Ep/Es=10000
Downloaded from ascelibrary.org by HARVARD UNIVERSITY on 09/26/13. Copyright ASCE. For personal use only; all rights reserved.

/..-
" u,

U,,

0.02-
+ +
++ ++ ++ ++ ° ° 'v j
^11* o o o o o o

0.2 0.3 0.4

tod
a.
FIG. 18. Kinematic Seismic Response of 1 x JV Relatively Rigid Fixed-Head Pile
Groups In Homogeneous Soil (E,JES = 10,000, Lid = 20, pjpp = 0.7, (3 = 0.05, and
v5 = 0.4; N = 1, 2, 3, 4, 6, and 9)

structures must be properly evaluated. Understandably, the amplitude of this

kinematic rotation is quite sensitive to: pile spacing ratio s/d, pile-group con-
figuration, total number of piles in the group, and pile-soil modulus ratio EJ
Es. For instance, /,,, decreases as both sld and the number of piles in the group
increase—apparently as a result of increasing overall rotational stiffness.

CONCLUSION

Dimensionless graphs have been presented for dynamic horizontal dis-

placements and rotations developing at the cap level of single piles and pile
groups, embedded in several idealized soil profiles and subjected to verti-
cally propagating harmonic waves. These graphs should be of practical value
in determining the "effective" seismic input motion at the base of structures,
if the free-field motion is known (e.g., in the form of a "design" spectrum
of the seismic code). A discussion of the presented results has focused on
elucidating the role of the key parameters and has aimed at developing
engineering insight into kinematic soil-pile and pile-pile interaction during
1876

J. Geotech. Engrg. 1991.117:1860-1879.

 2 »2 PILES Ep/Es=1000 L/d=20 |

Fixed-Head Piles }
1 i

0.8- ^*%^ \-
Downloaded from ascelibrary.org by HARVARD UNIVERSITY on 09/26/13. Copyright ASCE. For personal use only; all rights reserved.

0.6^
\ J \ . ;-
0.4- A s/(J=3
D s/d=5
0.2-
0 s/d=10 N:
0 0H

FIG. 19. Kinematic Seismic Response of 2 x 2 Relatively Compressible Fixed-

Head Pile Groups in Homogeneous Soil (EPIES = 1,000, Lid = 20, ps/pp = 0.7, p =
0.05, and vs = 0.4)
2.2 PILES Ep/Es= 10000 L/d = 20

FIG. 20. Kinematic Seismic Response of 2 x 2 Relatively Rigid Fixed-Head Pile

Groups in Homogeneous Soil (EPIES = 10,000, Lid = 20, p,./pp = 0.7, (3 = 0.05, and
vs = 0.4)

1877

J. Geotech. Engrg. 1991.117:1860-1879.

 2*2 PILES Ep/Es(L) = 500 L/d^20

Fixed-Head Piles
A s/d=3
P s/d=5
O s/d=10
Downloaded from ascelibrary.org by HARVARD UNIVERSITY on 09/26/13. Copyright ASCE. For personal use only; all rights reserved.

\_±]_d
L-

a =
FIG. 21. Kinematic Seismic Response of 2 x 2 Relatively Compressible Fixed-
Head Pile Groups in Nonhomogeneous Soil (EPIES(L) = 500, Lid = 20, pjpp = 0.7,
(3 = 0.05, and vs = 0.4)
Ep/Es(L) = 5000 L/d = 20 s/d = 5

fixed-Head piles

uod
a
°~V2n
FIG. 22. Kinematic Seismic Response of N x N Relatively Rigid Fixed-Head Pile
Groups in Nonhomogeneous Soil (EJES{L) = 5,000, Ud - 20, ps/p„ = 0.7, p =
0.05, and vs = 0.4; N = 2, 4, and 6)

1878

J. Geotech. Engrg. 1991.117:1860-1879.

 earthquakes. However, the conclusions drawn in this paper should not be
unduly generalized. One possible exception to the general trends observed
in Figs. 2 - 2 2 is with piles in soil deposits comprising a very shallow (thickness
<5d) and very soft (Esl < Es2l25) top soil layer.
Downloaded from ascelibrary.org by HARVARD UNIVERSITY on 09/26/13. Copyright ASCE. For personal use only; all rights reserved.

ACKNOWLEDGMENT

This work was partially supported by the National Center for Earthquake
Engineering Research, State University of New York at Buffalo (NCEER
89-3306).

APPENDIX. REFERENCES

Ahmad, S. (1986). "Linear and nonlinear dynamic analysis by boundary element

method," thesis presented to State University of New York at Buffalo, Buffalo,
New York, in partial fulfillment of the requirements for the degree of Doctor of
Philosophy.
Barghouthi, (1984). "Pile response to seismic waves," thesis presented Univ. of
Wisconsin, at Madison, Wisconsin, in partial fulfillment of the requirements for
the degree of Doctor of Philosophy.
Blaney, G. W., Kausel, E., and Roesset, J. M. (1976). "Dynamic stiffness of piles."
Proc. Second Int. Conference Numerical Meth. in Geomech., ASCE, 1001-1012.
Flores-Berrones, R., and Whitman, R. V. (1982). "Seismic responses of end-bearing
piles." J. Geotech. Engrg., ASCE, 108, 555-569.
Gazetas, G. (1984). "Seismic response of end-bearing piles." Int. J. Soil Dynamics
Earthquakes Engrg., 3(2), 82-93.
Kausel, E. (1981). "An explicit solution for the Green functions for dynamic loads
in layered media." Research Report, Dept. of Civil Engrg., MIT, Cambridge, Mass.
Kaynia, A. M., and Kausel, E. (1982). "Dynamic stiffnesses and seismic response
of pile groups." Research Report, Dept. of Civ. Engrg., MIT, Cambridge, Mass.
Luco, J. E., and Wong, H. L. (1986). "Response of a rigid foundation to a spatially
random ground motion." Earthquake Engrg. and Struct. Dynamics, 14, 891-908.
Makris, N., and Gazetas, G. (1993). "Dynamic pile-soil-pile interaction II: Lateral
and seismic response." Earthquake Engrg. and Struct. Dynamics, 22(2) (in press).
Novak, M. (1991). "Piles under dynamic loads." Proc. 2nd Int. Conjerence on Recent
Advances in Geotech. Earthquake Engrg. and Soil Dynamics, III, Univ. of Mis-
souri-Rolla.
Pais, A., and Kausel, E. (1985). "Stochastic response of foundations." Research
Report R85-6, MIT, Cambridge, Mass.
Roesset, J. M. (1977). "Soil amplification of earthquakes", Numerical Methods in
Geotechnical Engineering, C. S. Desai, and J. T. Christian, eds., McGraw-Hill.
Scanlan, R. H. (1976). "Seismic wave effects on soil-structure interaction." Earth-
quake Engrg. Struct. Dynamics, 4, 379-388.
Tazoh, T., Wakahara, T., and Shimizu, K. "Effective motion of group pile foun-
dation." Proc. of 9th World Conference on Earthquake Engrg., III.
Takemiya, H., and Yamada, Y. (1981). "Layered soil-pile-structure interaction."
Earthquake Engrg. Struct. Dyn., 9, 437-457.
Veletsos, A. S., and Prasad A. M. (1988). "Seismic interaction of structures and
soils: Stochastic approach." Technical Report NCEER-88-0021, State Univ. of New
York, Buffalo, N.Y.
Wolf, J. P., and Von Arx, G. A. (1978). "Impedance function of a group of vertical
piles." Soil dynamics and earthquake engineering, ASCE, 2, 1024-1041.

1879

J. Geotech. Engrg. 1991.117:1860-1879.