Missouri University of Science and Technology

Scholars' Mine
International Conference on Case Histories in (2004) - Fifth International Conference on Case
Geotechnical Engineering Histories in Geotechnical Engineering

Apr 13th - Apr 17th

Behavior of Piles in Liquefiable Soils During

Earthquakes: Analysis and Design Issues
W. D. L. Finn
Anabuki Komuten and Kagawa University, Japan

N. Fujita
Anabuki Komuten and Kagawa University, Japan

Follow this and additional works at: https://scholarsmine.mst.edu/icchge

Part of the Geotechnical Engineering Commons

Recommended Citation
Finn, W. D. L. and Fujita, N., "Behavior of Piles in Liquefiable Soils During Earthquakes: Analysis and Design Issues" (2004).
International Conference on Case Histories in Geotechnical Engineering. 2.
https://scholarsmine.mst.edu/icchge/5icchge/session00f/2

This Article - Conference proceedings is brought to you for free and open access by Scholars' Mine. It has been accepted for inclusion in International
Conference on Case Histories in Geotechnical Engineering 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@mst.edu.
 Behavior of Piles in Liquefiable Soils during Earthquakes: Analysis and Design Issues
W. D. L. Finn N. Fujita
Anabuki Komuten and Kagawa University, Japan

ABSTRACT

A general picture of the current state of the art and the emerging technology for dealing effectively with the seismic design and
analysis of pile foundations in liquefiable soils is presented. Two distinct design cases are considered and illustrated by case histories.
One is the static response of pile foundations to the pressures and displacements caused by lateral spreading of liquefied ground. The
other is the seismic response of piles to strong shaking accompanied by the development of high pore water pressures or liquefaction.
Design for lateral spreading is examined in the context of developments in design practice and the findings from shake table and
centrifuge tests. Response of piles to earthquake shaking in liquefiable soils is examined in the context of 1.5m cast in place
reinforced concrete piles supporting a 14 storey apartment building.

INTRODUCTION Lateral spreading is particularly damaging if a non-liquefied

layer rides on top of the moving liquefied soil. It is only in the
The seismic design of pile foundations in liquefiable soils last few years that the profession has begun to deal effectively
poses very difficult problems in analysis and design. The pile with these critical design issues. The progress is due to
foundation may undergo substantial shaking, while the soil is developments in analysis and findings from shaking table and
in a fully liquefied state and soil stiffness is at a minimum. centrifuge tests. These developments have allowed more
During this shaking phase, the pile is prone to suffering severe fundamental and comprehensive evaluations of case histories,
cracking or even fracture. Liquefaction may lead also to and a greater appreciation of design problems.
substantial increases in pile cap displacements above those for
the non-liquefied case. After liquefaction, if the residual The objective of the paper is to convey a general picture of the
strength of the soil is less than the static shear stresses caused current state of the art and the emerging technology for
by a sloping site or a free surface such as a river bank, dealing effectively with the design and analysis of pile
significant lateral spreading or down slope displacements may foundations in liquefiable soils taking into account the lessons
occur. The moving soil can exert damaging pressures against from case histories, the effects of earthquake shaking and the
the piles, leading to failure. Such failures were prevalent lateral pressures from post-liquefaction displacements.
during the 1964 Niigata and the 1995 Kobe earthquakes.

Fig.1. Ground displacements in 1964 Niigata earthquake (adapted from Hamada et al.1986).

SOAP 1 Page 1
BEHAVIOR OF PILE FOUNDATIONS DURING during the 1995 Kobe earthquake, is shown in Fig. 3. The
EARTHQUAKES function of these piles was to control settlement. They were
designed primarily for vertical loads and could not carry the
moments and shears caused by strong seismic shaking and
Driven Piles lateral spreading.

During liquefaction, large ground displacements can take However piles can be designed to carry the moments and
place on sloping ground or towards an open face such as a shears generated by earthquake shaking or post-liquefaction
river bank. Displacements from lateral spreading during the ground displacements. Figure 4 shows a bridge on pile
1964 Niigata earthquake are shown in Fig. 1. (Hamada et al foundations. The foundation soils liquefied during the 1983
1986). Nihon-Kai-Chubu earthquake. This led to a failure of the
approach embankments by lateral spreading but the pile
Displacements as large as 10m occurred towards the Shinano foundations survived without damage. A pile supporting a
River. Such displacements were very damaging to pile crane rail on Port Island, just offshore of Kobe City, is shown
foundations and caused the failure of two major bridges. in Fig. 5. The ground moved more than 1.0m in this location
Damage to a pile under a building in Niigata caused by about after liquefaction occurred during the 1995 Hyogo ken Nanbu
1m of ground displacement is shown in Fig 2 (Yasuda et al (Kobe) earthquake. The relative motion between the ground
1990). Complete shearing of a pile supporting a warehouse on and the pile is clearly evident in Fig. 5. However the pile was
Port Island near Kobe, by about 1.5m of ground displacement designed to carry significant shears and moments and survived
without damage.

Performance of CIDH Piles

Matsui and Oda (1996) evaluated the damage to the

foundations of five major elevated expressways in the Kobe
region, Japan, caused by the 1995 Kobe earthquake. They
focused on cast-in-deep-hole (CIDH) reinforced concrete
piles, as these comprised 80% of all foundation types. The
piles were all over 1m in diameter.

Damage was classified into the four categories given in Table

1 and estimates of the residual load resisting capacities of piles
in each damage class are also given. The damage was
assessed by direct observation of pile shafts, examination of
cores taken from the piles and observations made using
borehole television (BHTV) cameras. The BHTV system was
very effective, even hair cracks could be observed in the
images. Non-destructive methods such as velocity logging,
Fig.2. Damage to pile by 2m of lateral ground displacement impact wave and electromagnetic wave methods were also
during 1964 Niigata earthquake (Yoshida et al. 1990). used.

Fig.3. Shearing of a pile by ground displacements in 1995 Fig.4. Bridge on undamaged pile foundations with failed
Kobe earthquake (Finn and Fujita 2002). approaches due to liquefaction (Finn and Fujita 2002).

SOAP 1 Page 2
 Despite the extensive liquefaction and the severe damage to
the elevated super-structures, damage to the CIDH piles was
negligible. The most extensive damage was along the No. 5
Bay Route of the Hanshin Expressway : 11% B, 37%C and
52% D. On the No. 3 Kobe route, the damage was 16% C and
84% D. There was no instance of A category damage. Matsui
and Oda (1996) explained cracking pattern as follows. The
cracks near the top of the pile are to be expected as this is
usually the location of maximum moment. The cracks lower
down the pile occur at the location of the second largest
moment, at an interface between soft liquefied soils and a
harder formation or where there is an abrupt change in the
density of reinforcement.

The comments on the residual capacity of the damaged piles in

Fig.5. Undamaged pile supporting a crane rail in ground that Table 1 were based on tests conducted on CDIH piles, 1m in
moved about 1m (Finn and Fujita 2002). diameter by the Hanshin Expressway Public Corporation in
1993 (Kimura et al, 1994). The tests involved single piles and
a 3x3 pile group. Data from a typical load test is shown in
Matsui and Oda (1996) found that pile damage correlated with Fig. 6. The piles showed cracking at around 10cm of
sub-soil conditions. Damage was largely confined to areas of displacement. At 40cm displacement, the piles still retained
liquefaction with and without lateral flow of liquefied soil. “sufficient lateral capacity.” Figure 7 shows a photo of the

Fig.7. Damage to 1m diameter CDIH pile at 40cm lateral

displacement (Kimura et al. 1994).
Fig.6. Lateral load test on 1m diameter CDIH pile (Kimura et
al. 1994).

Table 1. Classification of pile damage (adapted from Matsui and Oda, 1966)

Damage Type A B C D
Severe Heavy Light No Damage
Damage Many cracks with Many cracks with Some cracks with Almost
Description concrete separation concrete separation separation near top no cracks
all over pile near pile top
Some cracks in
Buckling of main Many cracks around middle middle and lower
reinforcement and lower end of pile end of pile

Pile shaft separation

Residual Pile Totally inadequate Probably adequate Adequate capacity Adequate capacity
Capacity vertical capacity,
Partial lateral capacity

SOAP 1 Page 3
external conditions at the head of one of these piles probabilistic basis. In engineering practice, the free field
corresponding to a displacement of 40cm. displacements are assumed usually to vary linearly from top to
bottom of the liquefied layer. The deformed shape of a pile
Clearly the CDIH piles behaved very well, more particularly foundation caused by these post-liquefaction displacements is
as they were designed for much less intense ground shaking illustrated in Fig. 8.
than they experienced during the Kobe earthquake.

The review of case histories has clearly demonstrated the Force Analysis
design problems posed by pile foundations in liquefied soils.
To cope with these problems it is essential to have a reliable A force based analysis is recommended in a number of
method of calculating the effects of earthquake shaking and Japanese design codes for analysis of piles foundations in
post liquefaction displacements on pile foundations. An liquefied soils, undergoing lateral flow (JWWA 1997, JRA
overview of the methods used in practice will be given which 1996). The underlying concepts are rational and simple. An
indicates some of the advantages and limitations of the various unliquefied surface layer, which is transported on the moving
methods. The aim of the review is to present a reasonably liquefied soil is assumed to apply passive pressure on the
integrated up to date assessment of the state of the art. foundation. A liquefied layer is assumed to apply a pressure
less than the equivalent hydrostatic pressure on the piles
because of the internal flow resistance of the liquefied sand.
ANALYSIS OF PILE FOUNDATIONS IN LATERALLY The transmitted lateral pressure was found to average about
SPREADING GROUND 30% of the overburden pressure on the basis of back analysis
of case histories. The pressure distribution against the
In the case histories section, it was shown that large post foundation for design is shown in Fig. 9.
liquefaction displacements can occur and that these can be
very damaging to pile foundations. These potential Dobry and Abdoun (2001) and Ramos et al (1999) have
deformations can control design but they are very difficult to studied the behavior of piles in laterally flowing soils by
predict reliably. In engineering practice, the displacements at centrifuge tests. The setup for a typical centrifuge test is
the top of the liquefied layer are often estimated by empirical shown in Fig. 10. Typical test results for moments in the piles
formulas based on field data from past earthquakes. The first are given in Fig. 11. In order to simulate the moments they
predictor equation was developed in Japan by Hamada (1986). adopted the two different pressure distributions: inverted
Very comprehensive predictor equations have been developed triangular and a uniform distribution. The adoption of the
by Youd et al (1999) in the USA which are used in practice in inverted triangular distribution may have been influenced by
North America. An updated version of the Hamada equation the inverted triangular distribution of displacements in the
has been adopted by the Japan Water Works Association liquefied soil. However when there is lateral restraint at the
(JWWA, 1997) based only on ground slope and the thickness pile head, both distributions seem to overestimate the bending
of the liquefied layer. Bardet et al (1998a, b) have developed a moments in the upper part of the pile. Abdoun and Wang
method for predicting post–liquefaction displacements on a (2000) studied the effects of lateral spreading of ground with a
upper lightly cemented layer on piles in centrifuge tests. They
concluded that the moments in the pile were dominated by the
lateral pressures from the cemented layer.

Fig.8. Distortion of pile foundation by lateral soil Fig.9. Design pressures against piles in laterally flowing
displacement. liquefied soils (JWWA 1997).

SOAP 1 Page 4
 Fig.10. Centrifuge test on pile in flowing soil (Ramos et al. 1999).

Fig.12. A Winkler spring model for pile foundation analysis.

Fig.11. Computed and measured pile moments (adapted from

Ramos et al. 1999).

Displacement Analysis

The first step in the analysis is to estimate the post-

liquefaction free field displacements using one of the
empirical formulas. These displacements are usually assumed
to vary linearly from the top to the bottom of the upper
liquefied layer. These linearly distributed displacements are
then applied to the free field ends of the near field springs in
the very general Winkler model shown in Fig.12 and a static
analysis is performed (Finn and Thavaraj, 2001). Degraded
p-y curves have usually been used for this kind of analysis. Fig.13. Calculated pile displacements for specified ground
The effects of lateral spreading on 1.5m diameter CDIH piles flow (Finn 1999).

SOAP 1 Page 5
supporting the structure shown in Fig. 29 were analyzed as
described above. The free field displacements at the surface
were estimated to be between 15 cm and 25cm. The computed
pile displacements, assuming that the pile head is fixed against
rotation, are shown in Fig. 13. The resulting bending
moments are shown in Fig. 14. Note that the maximum
bending moment is near the interface between the liquefied
and non-liquefied layers.

Bending Moment (kNm)

-5000 0 5000 10000
0
Free Field
2 Displacement
at the Pile Top (cm)
4
15
6
25
8
Depth (m)

10 Fig.15. Japanese computational model for pile groups (JRA

12 1996).

14 evaluated by plate loading tests or correlations with the SPT-

16 N measurements and therefore includes some nonlinear
effects. In some design offices the spring constant K is taken
18 as zero in liquefied soil. The typical Japanese computational
20 model for pile groups is shown in Fig.15.

. The JRA (1996) code for highway bridges recommends

Fig.14. Pile moments induced by field displacements (Finn reductions in the spring stiffness for use in liquefiable soils
1999). that depend on the factor of safety, FL, against liquefaction.
The reduction factors are given in Table 2. The resistance to
liquefaction, RL, is determined by cyclic triaxial tests on
Soil Properties for Displacement Analysis undisturbed samples obtained by in-situ freezing techniques.
This strength is modified depending on whether Type 1 or
The selection of soil properties for post–liquefaction flow Type 2 motions are used in design, by a factor cw. Then R= cw
deformation analysis will be discussed in the context of RL is the dynamic shear strength ratio in Table 2. The factor
engineering practice. cw has a value of 1 for Type1 motions and a value in the range
1.0-2.0 for Type 2 motions. The code should be consulted
Japanese practice. In Japanese practice the springs in the for details of the 2 types of motions. Generally Type 1
Winkler model are linearly elastic-plastic. The elastic soil motions are the design motions before the Kobe earthquake.
stiffness is determined by semi-empirical code formulas Type 2 motions were introduced to provide protection against
related to the elastic modulus of the soil. This modulus is another earthquake like Kobe.

Table 2. Reduction coefficients for soil constants due to liquefaction (JRA 1996)

Depth from the Present Dynamic Shear Strength Ratio R

Range of FL
Ground Surface x (m) R ≦ 0.3 0.3 < Ra
0 ≦ x ≦ 10 0 1/3
FL ≦ 1/3
10 < x ≦ 20 1/3 1/3

0 ≦ x ≦ 10 1/3 2/3
1/3 < FL ≦ 2/3
10 < x ≦ 20 2/3 2/3
0 ≦ x ≦ 10 1/3 1
2/3 < FL ≦ 1
10 < x ≦ 20 1 1

SOAP 1 Page 6
North American Practice. There is no general consensus in centrifuge at UC Davis. Pipe piles were used. The single
North American practice on the appropriate modeling of the piles had prototype diameters of 0.36m, 0.73m, and 1.45m: the
Winkler springs for post-liquefaction analysis. The basis of piles in the pile group were 0.73m in diameter. The
most analyses is a degraded form of the API (1995) p-y curves foundation soil profile sloped gently towards a channel at one
or curves due to Reese (1974). The practice is to multiply the end of the shear box as shown in Fig. 17. It consisted of a
p-y curves, by a uniform degradation factor p, called the p- non-liquefiable layer of clay, with a thin sand cover, underlain
multiplier, which ranges in value from 0.3 to 0.1. This follows by a liquefiable layer of sand with a relative density of 35%
from the original work of Dobry et al. (1995). They found and a base layer of dense sand at a relative density of 85%.
that bending moments could be predicted adequately using a
Winkler analysis, if the commonly used p-y curves were
uniformly degraded by multiplying by a degradation factor p
that appeared to diminish with increasing pore water pressure
to a value of 0.1 at 100% excess pore water pressure. Wilson
et al (1999) confirmed these results but showed that the p-
multiplier for fully liquefied soil depended also on relative
density, ranging in value from 0.1-0.2 for sand at about 35%
relative density and 0.25-0.35 for a relative density of about
55%.

They also found that the resistance of the loose sand did not
pick up even at substantial strains but the denser sand, after an
initial strain range in which it showed little strength, picked up
strength with increasing strain. This finding suggests that the
good performance of the degraded p-y curves which did not Fig.17. Centrifuge-model-test (Brandenburg et al. 2001).
include an initial range of low or zero strength, must be test
specific and the p-multiplier may be expected to vary from one The responses of the piles to lateral spreading were analyzed
design situation to another. using a Winkler model based program, LPILE (Reese et al
2000). Matlock’s 1970) static p-y relation for soft clay and
The very low initial strength range in the laboratory p-y curves Reese’s (1974) static p-y relation for sand were used to
followed by a range of increasing strength is related to the represent the non-linear springs. A p-multiplier p=0.1 was
dilatancy characteristics of sand at low effective stresses. used for fully liquefied sand.
Similar behavior is observed in tests in which undrained
monotonic loading is conducted on sand specimens after The responses of the piles to lateral spreading were analyzed
cyclic loading to liquefaction. Typical examples of this using a Winkler model based program, LPILE (Reese et al
phenomenon are shown in Fig. 16 (Yasuda et al 1999). Vaid 2000). Matlock’s (1970) static p-y relation for soft clay and
and Thomas (1995) found similar results and also showed that Reese’s (1974) static p-y relation for sand were used to
the strain range of very low undrained resistance after represent the non-linear springs. A p-multiplier p=0.1 was
liquefaction depends on the number of cycles of stress reversal used for fully liquefied sand.
the sand experiences after liquefaction, before the undrained
monotonic loading is applied. Three cases were considered: (1) original p-y curves for loose
sand with p=0.1 and only the properties in the loose liquefied
sand were degraded for pore pressure effects ; (2) original p-y
curves for loose sand with p=0.1and reductions in p-y stiffness
and capacity of the dense sand due to pore water pressures in
that layer; (3) the same as case (2) except that the standard p-y
adjustment factors to the static p-y curves for cyclic loading
were made also.. As Brandenburg et al (2001) point out these
latter adjustments were developed for the large number of
water wave generated stress cycles associated with a major
offshore storm and are probably not applicable to the far fewer
significant stress cycles associated with earthquake shaking.
Comparison of measured and computed responses led to a
Fig.16. Post-liquefaction undrained stress-strain behavior of number of important conclusions. The three most important
sand (Yasuda et al 1999). ones are quoted verbatim below.
• the recorded responses of the three single piles and
Brandenburg et al (2001) conducted a very comprehensive the one group of two piles could be modeled within
series of tests to determine the effects of various parameters on the range of parameter variations that were studied,
pile performance in laterally spreading ground. Centrifuge but all the responses could not be accurately modeled
tests on single piles and 2-pile groups were conducted on the with the same set of input parameters.

SOAP 1 Page 7
 • the parameter studies also showed that the standard
adjustments to p-y relations for cyclic loading would
have resulted in substantial under-prediction of lateral
loads from the clay layer
• the calculated bending moments were more sensitive
to the strength and p-y parameters for the upper clay
and sand cover layers, and less sensitive to the p-
multiplier assigned to the liquefied layer.
These findings pose clear warnings for anyone contemplating
analyses of piles in laterally spreading soils using the standard
North American p-y curves. The crucial factors seem to be;
the dominating role of the non-liquefiable layer, the
inappropriateness of using the standard cyclic loading
reduction factors for earthquake shaking and the large
uncertainty associated with the results of any analysis.

Some of the problems of arriving at a generally acceptable set

of Winkler non-linear p-y curves for analysis arise from the
assumed form of the curves. If the form is incompatible with
the actual stress-strain behavior of the soil, problems in
simulating the responses of different pile foundations with one
set of p-y curves is not surprising. The North American p-y
curves are concave downwards and this is not compatible with
the post-liquefaction undrained behavior of liquefied sand
under monotonic loading which is concave upwards as shown
in Fig. 16 above.

Weaver et al (2002) conducted full scale cyclic loading tests in

the field on a 0.6m diameter cast-in steel-shell (CISS) pile in Fig.18. Comparison of standard p-y curves with curves back-
liquefied soil to assess the accuracy of the p-y type of analysis. figured from test data at depths of (a) 0.2m and (b) 2.3m from
The test site is on Treasure Island in San Francisco Bay which a full scale pile test (Weaver et al 2002).
is the location of the National Geotechnical Experimentation
Site. Therefore soil conditions at the site are very well known.
Liquefaction was caused by blasting and the cyclic loading
was conducted using a high speed hydraulic actuator. The
back figured p-y curves for the liquefied sand differed
significantly in shape from the standard p-y curves modified
by the p-multiplier. The slope of the standard p-y curve is
greatest at small displacements and eventually decreases to
zero at large displacements. The back calculated p-y curves
show no resistance for a range of displacements between
20mm and 50mm. The soil resistance increased thereafter and
was still increasing after 150mm. The shape of the back-
calculated p-y curves are shown in Fig. 18. The standard p-y
curves including the p-multiplier effect are also shown for
comparison. The two sets of p-y curves have distinctly
different shapes and give different estimate of soil resistance.
The shapes of the Weaver et al (2001) curves are consistent
with the post-liquefaction undrained monotonic loading test
data from Yasuda et al (1999).
Fig. 19. Comparison of standard p-y curves with curves back-
The p-y curves from the full scale field test share some figured from centrifuge data (Wilson et al 2000).
characteristics with the p-y curves obtained by Wilson et al
(2000) from centrifuge tests. The Wilson data for one cycle of Dobry (1995) found similar results and concluded that
loading at different depths are shown in Fig. 19. The liquefied loose sands provide very little resistance. In this
hysteresis loops are very attenuated showing almost no case it may be reasonable to ignore the effects of liquefied
pressure being exerted against the pile. The standard p-y loose sand as far as pressure on the piles is concerned even in
curves, with the p-multiplier effect included, are shown for the force analysis. This is consistent with the judgment of
comparison. Again the shapes are radically different. Liu and some Japanese designers who assign zero stiffness to the

SOAP 1 Page 8
elastic springs in their form of Winkler displacement analysis with any structural mass included with the pile, are excited by
as discussed earlier. the seismic base motions and free field motions applied to the
end of each Winkler spring. The free field motions at the
desired elevations in the soil layer are computed by 1-D
DYNAMIC ANALYSIS OF PILE FOUNDATIONS IN dynamic analyses using a computer analysis program such as
LIQUEFIABLE SITES SHAKE (Schnabel et al. 1972).

In the previous section, the more or less passive response of An alternative to the Winkler type computational model is to
piles to pressures from laterally spreading ground due to use a finite element continuum analysis based on the actual
liquefaction was investigated. The dynamic response of piles soil properties. Dynamic nonlinear finite element analysis in
in liquefied soil in response to earthquake shaking will now be the time domain using the full 3-dimensional wave equations
considered. The issues will be explained in the context of the is not feasible for engineering practice at present because of
behavior of CDIH piles. A major research project on the the time needed for the computations. However, by relaxing
seismic behavior of these piles is underway at Kagawa some of the boundary conditions associated with a full 3-D
University, supported by Anabuki Komuten, a major analysis, it is possible to get reliable solutions for nonlinear
construction firm with headquarters in Takamatsu. The response of pile foundations with greatly reduced
company uses CDIH piles almost exclusively for supporting computational effort.
their buildings on reclaimed land. Such land is highly
susceptible to liquefaction during earthquake shaking. Since seismic response analysis is usually conducted assuming
Potential methods of analysis will be reviewed and some that the input motions are horizontally polarized shear waves
examples from building studies will be presented. propagating vertically, the PILE-3D model retains only those
parameters that have been shown to be important in such
analysis. These parameters are the shear stresses on vertical
Overview of Analysis and horizontal planes and the normal stresses in the direction
of shaking. The soil is modeled by 3-D finite elements as
The pile foundation-structure system vibrates during shown in Fig. 20. The pile is modeled using beam elements or
earthquake shaking as a coupled system. Logically it should
be analyzed as a fully coupled system. However this type of
analysis is not feasible in engineering practice. Many of the
popular structural analysis programs cannot include the pile
foundation directly into a computational structural model.
Therefore various approximate methods of analysis are used.

The most common approach to the analysis of pile foundations

is to use Winkler springs to simulate soil-pile interaction. The
springs may be elastic or nonlinear. Some organizations such
as the American Petroleum Institute give specific guidance for
the development of nonlinear load-deflection (p-y) curves as a
function of soil properties that can be used to represent
nonlinear springs [API 1995]. The API (p-y) curves, which
are the most widely used in engineering practice, are based on
data from static and slow cyclic loading tests in the field. The
reliability of these (p-y) curves for the analysis of pile
foundations even under static and slow cyclic loading has been
questioned (Murchison and O’Neill 1984). The effectiveness
of p-y curves for seismic loading conditions is poorly
established. Researchers trying to simulate the seismic
response of piles in centrifuge tests usually resort to back- Fig. 20. Computational model in Pile-3D.
figured p-y curves and, even then, find that no one set of p-y
curves can be used for general analysis (Brandenburg et al. volume elements. The pile is assumed to remain elastic. This
2001). Finn and Thavaraj (2001) have shown by analysis of assumption is in keeping with the design philosophy that the
the response of single piles in dry sand in centrifuge tests that structural elements of the foundation should not yield. In the
a Winkler computational model with API p-y curves gave analysis of concrete piles, the cracked section moduli are used,
poor results for strong shaking but very good results for low when deformations exceed the cracking limit. A full
level shaking. description of this method, including validation studies, has
been presented by Wu and Finn (1997a, b). The method is
A general Winkler dynamic model is shown in Fig. 12 above.. incorporated in the computer program PILE-3D. The results
The near field interaction between pile and soil is modeled by are quite accurate for excitation due to horizontally polarized
springs and dashpots. The near field pile-soil system, together shear waves propagating vertically.

SOAP 1 Page 9
An effective stress version of this program, PILE-3D-EFF, has The soil profile consists of two level layers of Nevada sand,
been developed by Finn and Thavaraj (1999) and validated by each approximately 10m thick at prototype scale. Nevada
Finn et al (1999) and Finn and Thavaraj (2001) in cooperation sand is a uniformly graded fine sand with a coefficient of
the geotechnical group at the University of California at Davis. uniformity of 1.5 and mean grain size of 0.15 mm. Sand was
In support of the subsequent analyses of CDIH piles in air pluviated to relative densities of 75%-80% in the lower
liquefied sands, some excerpts from the validation study with layer and 55% in the upper layer. Prior to saturation, any
UC Davis for a single pile are given here. entrapped air was carefully removed. The container was then
filled with a hydroxy-propyl methyl-cellulose and water
mixture under vacuum. The viscosity of this pore fluid is
Analysis of Centrifuge Tests at UC Davis about ten times greater than pure water to ensure proper
scaling. Saturation was confirmed by measuring the
Dynamic centrifuge tests of pile supported structures in compressive wave velocity from the top to the bottom of the
liquefiable sand were performed on the large centrifuge at soil profile.
University of California at Davis, California. The models
consisted of two structures supported by single piles, one The shear strain dependencies of the shear modulus and
structure supported by a 2×2 pile group and one structure damping ratio of the soil were defined by the curves suggested
supported by a 3×3 pile group. The typical arrangement of by Seed and Idriss (1970) for sand. The friction angles of the
structures and instrumentation is shown in Fig. 21. Full upper and the lower sand layers were taken as 35° and 40°,
details of the centrifuge tests can be found in Wilson et al. respectively. Increments in seismic pore water pressures at any
(1997). The model dimensions and the arrangement of time were generated in each individual element depending on
bending strain gauges for the single pile are shown in Fig. 22. the accumulated volumetric strain prevailing in that element at
Model tests were performed at a centrifugal acceleration of that time and the current increment in volumetric strain, using
30g. the pore water pressure model proposed by Martin et al
(1975). The moduli and shear strengths of the foundation soils
GP1
were modified continuously to account for the effects of the
changing seismic pore water pressures.
Porepressure SP1 Strain Gauge
Displacement Accelerometer

Results of Single Pile Analysis

Acceleration Response. Figure 23 shows the measured and

computed acceleration response of the superstructure. There is
generally good agreement between them, especially around the
time period of peak response.

0.5
Measured
Acceleration (g)

Computed
Fig. 21. Layout of models for centrifuge tests.
0.0
Mass=1.82 kg
12.70
15.24
17.78

-0.5
20.32
25.40

0 5 10 15 20
30.48

SG1
Time (sec)
40.64

SG2
43.70

SG3

SG4 Fig. 23. Comparison of measured and computed

superstructure acceleration time histories.
68.58

SG:Strain Gauge
Pore Water Pressure Response. Figure 24 shows comparisons
between measured and computed pore water pressures at three
different depths; 1.14 m, 4.56 m, and 6.78 m in the free field.
There is generally good agreement between the measured and
computed pressures.

Bending Moment Response. Figure 25 shows the measured

and computed bending moment time histories at two different
All dimensions are in cm model scale
depths; 0.76 m and 1.52 m. There is a very good agreement
Fig. 22. Instrumented pile for single pile test. between the measured and computed time histories. Figure 26

SOAP 1 Page 10
shows the profiles of measured and computed maximum Bending Moment (MNm)
bending moments with depth. The comparison between -2.0 -1.0 0.0 1.0
measured and computed moments is adequate for engineering 0.0
purposes, although the maximum moment is overestimated by
10%-15% between 1 m and 4 m depths.
Porepressure Ratio (%)

100 4.0

Depth (m)
50

0
at Depth, D= 1.14 m 8.0

0 5 10 15 20 Measured
Time (sec) Computed
Porepressure Ratio (%)

12.0
100
Fig. 27. Comparison of measured and computed maximum
bending moments profiles along the pile.
50

at Depth, D= 4.56 m
0 ANALYSES OF CDIH PILES

0 5 10 15 20 Seismic response analyses were conducted on a 1.5 m

Time (sec) diameter cast-in-place reinforced concrete pile supporting a
Porepressure Ratio (%)

column of the 14 storey apartment building using PILE 3-D-

100 Measured
EFF. The soil conditions and pile are shown in Fig.28.
Computed Slightly idealized site conditions shown in Fig. 29. The upper
50 10m are expected to liquefy during the design earthquake.
The mass mounted on the pile in Fig. 29 represents the portion
at Depth, D= 6.78 m of the total mass supported by the pile. The purpose of
0
placing the mass on the pile is to model approximately the
0 5 10 15 20 inertial interaction between the super-structure and the pile
Time (sec) foundation. It is mounted on the pile head by a flexible
support that gives the mass a period of vibration of 1.4s that is
Fig. 25. Comparison of measured and computed porewater the estimated fundamental period of the prototype structure.
pressure time histories at three depths.
1.5
N
Measured
Moment (MNm)

Computed

0.0

at Depth, D= 0.76 m
-1.5
Depth [m]

0 5 10 15 20
Time (sec)

1.5
Measured
Moment (MNm)

Computed

0.0

at Depth, D= 1.52 m
-1.5
0 5 10 15 20
Time (sec)
Fig. 26. Comparison of measured and computed bending
moment time histories at two depths. Fig. 28 Site in reclaimed land.

SOAP 1 Page 11
 Displacement [cm]
SDOF System
Period, T=1.4 sec
Mass, M=650 Mg 0.0 1.0 2.0 3.0 4.0 5.0 6.0
0

Sand Concrete Pile 4

10m (N1)60= 5 Diameter, D=1.5m
Flexural Rigidity, EI=4850 MNm2 6
φ =30o Unit Weight, =25 kN/m3
8

Depth [m]
10
3m (N1)60
12

Sand 14
7.5m (N1)60= 50
16
φ =40o fixed
18
not fixed
20
4.5m

Fig .30. Pile deflections at maximum pile head displacement.

Amax=0.25g
Bending Moment [kNm]
x 1.0E+03
Fig. 29. Model of soil-pile-structure system.
-6.0 -4.0 -2.0 0.0 2.0 4.0 6.0
Two kinds of analyses were conducted; total stress dynamic 0
analysis in which seismic pore water pressures and 2
liquefaction are ignored and effective stress analysis that
4
automatically takes the seismic pore water pressures into
account. In general, soil properties are adjusted continuously 6
to maintain compatibility with current pore water pressures 8
Depth [m]

and shear strains. The peak acceleration of the input

10
acceleration record is 0.25g and is amplified to 0.4g at the
surface. The surface accelerations become negligible after 12
liquefaction has occurred. Dynamic effective stress analyses 14
of this system were conducted for two conditions: including
16
both inertial and kinematic interaction, and with kinematic fixed
interaction only. The latter analyses did not include the mass 18
not fixed
of the superstructure. Data from these analyses are compared 20
to evaluate the significance of kinematic interaction.

Fig. 31. Pile moments at maximum pile head displacement.

RESULTS OF ANALYSES

Analyses with Inertial interaction The displacements are more than twice as large when the pile
head is free to rotate. The maximum moment occurs at the
Pile displacements and moments for the 14 storey building, at pile head, when the pile head is fixed against rotation, but
the instant of maximum pile head displacement, are shown in significant moment also occurs at the boundary between the
Fig. 30 and Fig. 31 respectively. Approximately the top 10 m softer and stiffer soils. When the pile head is not fixed against
liquefy or develop very high pore water pressures during rotation, the maximum moment occurs at the boundary
earthquake shaking. Results are shown for two conditions; between the stiffer and softer soils. This moment is
the pile head is fixed against rotation and the pile head is free approximately equal to the pile head moment, when the pile
to rotate. There is generally a greater degree of fixity in head is fixed against rotation. The results show that when
Japanese buildings because much deeper grade beams are used designing piles or evaluating pile foundations in potentially
to tie adjacent pile caps together than in North America, as liquefiable soils for earthquake loading, it is important to
shown in Fig. 32. The large grade beams provide considerable make a realistic assessment of pile head restraint against
restraint against rotation and so they mobilize much higher rotation and to be aware of the potential for large moments at
inherent structural stiffness in the pile. the interfaces between soft and hard layers.

SOAP 1 Page 12
 Bending Moment [kNm]
x 1.0E+03
-6.0 -4.0 -2.0 0.0 2.0 4.0 6.0
0

6
8

Depth [m]
10

12
14

16
fixed
18
not fixed
20

Fig. 34. Pile moments at maximum pile head displacement.

has more than doubled compared to the previous case of no

stiff upper layer. This is due to the restraint of the upper layer
Fig. 32. Large grade beam for 14 storey building. and the movement of that layer as a rigid body after
liquefaction develops. The stiff upper layer greatly increases
At some sites a thick surface layer of non-liquefiable soil may the bending moment demands on the pile during earthquake
lie over the liquefaction zone. A stiff upper layer is excitation.
incorporated into the original site of the 14 storey building.
Deflections and moments for this case, at the instant of The moments at the pile head and at the interface between the
maximum pile head displacement, are shown in Fig. 33 and soft and stiff soils have increased by 30%, compared to the
Fig. 34 respectively. As before, the results are shown for two case without the upper layer. When the pile is fixed against
pile head conditions, no pile head rotation and the pile head is rotation the moments at the pile head and the interfaces
free to rotate. between layers are about the same. The behavior of the upper
layer is clarified further in the next section which presents
Deflections and moments for this case, at the instant of results from kinematic analyses.
maximum pile displacement, are about the same whether the
pile head is fixed against rotation or not. Also the deflection
of the pile head, when the pile head is fixed against rotation, KINEMATIC ANALYSIS

Displacement [cm] Kinematic analyses were conducted on the 1.5 m diameter pile
to assess the importance of kinematic interaction. Analyses
0.0 1.0 2.0 3.0 4.0 5.0 6.0
were conducted with and without the stiff surface layer and, in
0
each case, the pile head was considered either fixed against
rotation or not. The kinematic analyses were conducted after
2 removing the superstructural mass in Fig. 29.
4
The pile and free field displacements at the instant of
6
maximum pile head displacement are shown in Fig. 35 for the
8 case when there is a stiff surface layer. It is evident that the
Depth [m]

10 stiff surface layer is moving as a rigid body at the time of

maximum pile head displacement which occurs after the
12
incidence of liquefaction. At this time it also appears to be
14 driving the pile, so that the pile and surface layer undergo
16
about the same displacements. Consequently when the stiff
fixed surface layer is present, the kinematic pile head moments
18 shown in Fig. 36 are about the same as the moments, when
not fixed
20 both inertial and kinematic interactions are included (Fig. 34 ).
This indicates that, in this case, the kinematic moments
Fig. 33. Pile deflections at maximum pile head displacement. dominate the moment response of foundation.

SOAP 1 Page 13
 Displacement [cm] accompanied by the development of high pore water pressures
or liquefaction.
0.0 1.0 2.0 3.0 4.0 5.0 6.0 Design for lateral spreading is examined in the context of
0 developments in design practice and the findings from shake
table and centrifuge tests. Response of piles to earthquake
2 shaking in liquefiable soils is examined in the context of 1.5m
4 cast in place reinforced concrete piles supporting a 14 storey
6
apartment building.

8 Two methods for design against lateral spreading, a force

Depth [m]

10 based method which is specified in Japanese codes and a

displacement based method which is sometimes used in North
12
America are presented. The Japanese method is based on
14 studies of case histories from past earthquakes, especially the
16 pile foundation failures caused by lateral spreading during the
free field Kobe earthquake and is very simple to apply. The pressures
18
pile on the pile foundation are specified as follows; liquefied soil
20 exerts a pressure equal to 30% of overburden pressure and an
unliquefied surface layer exerts passive pressure. Data from
Fig. 35. Displacements of pile and free field at maximum pile head simulated earthquake loading of model piles in liquefiable
displacement. sands in centrifuge tests indicate that the force method is an
adequate design method.
Bending Moment [kNm]
x 1.0E+03 The displacement method requires the prediction of surface
-6.0 -4.0 -2.0 0.0 2.0 4.0 6.0
displacements which are then distributed linearly over the
liquefied layer and the analysis of pile response to these
0
displacements by a static analysis using a Winkler model or a
2 finite element method. Two factors make this method appear
4 quite unreliable. The surface displacements are predicted by
empirical formulas which can err by a factor of 2 and there is
6
no agreement yet on a standardized set of p-y curves or stress-
8 strain curves for representing the post-liquefaction stress-strain
Depth [m]

10 behavior of the soil. Recent centrifuge and shake table tests

are contributing significantly to a framework of understanding
12 about how piles and soils interact after liquefaction during
14 lateral spreading.
16
fixed The behavior of piles in liquefied ground was studied in the
18 context of large diameter CDIH reinforced concrete piles.
not fixed
20 These piles are often used to support buildings in reclaimed
land in Japan and as combined foundation–piers for bridges
Fig. 36. Kinematic moments at maximum pile at maximum pile head worldwide. Analyses show that large bending moments
displacement. develop in critical areas such as at the pile head, when it is
fixed against rotation, and the boundary between liquefied and
Clearly analyses that neglect kinematic effects may in some non-liquefied layers. The analyses also demonstrate that if a
situations underestimate significantly design moments and stiff surface layer overlies the liquefied zone, then the moment
shearing forces in foundation piles. and deflection demands on the pile may be substantially
increased over the case when the stiff upper layer is not
present.
CLOSING REMARKS
Restraint against pile head rotation has a significant effect on
A general picture of the current state of the art and the the response of piles in liquefied soils, when the surface layer
emerging technology for dealing effectively with the seismic liquefies. The pile cap displacements may be up to three times
design and analysis of pile foundations in liquefiable soils is larger, if the restraint is low compared to full fixity.
presented in this paper. Two distinct design cases were
considered and are illustrated by case histories. One is the If an unliquefiable surface layer covers the liquefied stratum,
static response of pile foundations to the pressures and large kinematic moments may develop in the pile, especially if
displacements caused by lateral spreading of liquefied ground. the surface layer is stiff and relatively thick. After
The other is the seismic response of piles to strong shaking liquefaction the surface layer tends to move as a rigid body

SOAP 1 Page 14
and drives the pile to greater displacements. The increased Dobry, R. and Abdoun, T. [2001]. “Recent studies on seismic
displacements and the greater fixity against rotation of the pile centrifuge modeling of liquefaction and its effect on deep
head are responsible for the increase in moments. foundations,” Proc. 4th Int. Conf. on Recent Advances in
Geotechnical Earthquake Engineering and Soil Dynamics, S.
The keys to good design are reliable estimates of Prakash, Editor, San Diego, CA, March 26-31.
environmental loads, realistic assessments of pile head fixity
and the use of methods of analysis that can take into account Finn, W. D. Liam [1999]. “Lessons from Recent Earthquakes
adequately all the factors that control significantly the on Foundation Performance, Design and Analysis”, Proc.
response of the pile-soil-structure system to strong shaking Inaugural Anabuki Chair Symposium, Kagawa Faculty of
and /or lateral spreading in a specific design situation. Not all Engineering, FEKU TN009, May.
factors are important all the time but an informed background Finn, W. D. Liam and Fujita, N. [2002]. “Piles in liquefiable
is essential in making decisions about what can be ignored. soils: seismic analysis and design issues”, Int. J. of Soil
Dynamics and Earthquake Engineering, Vol. 22, 731-742.

ACKNOWLEDGEMENTS Finn, W. D. Liam and Thavaraj, T. [1999]. “Pile-3D-EFF, A

program for nonlinear dynamic effective stress analysis of pile
The research project on the seismic design and analysis of pile foundations”, Anabuki Chair of Foundation Geodynamics,
foundations is funded by Anabuki Komuten, Takamatsu, Kagawa University, Japan.
Japan. The support of the company is gratefully
acknowledged. Finn, W. D. Liam and Thavaraj, T. [2001]. “Deep foundations
in liquefiable soils: Case histories, centrifuge tests and
methods of analysis”, CD-ROM Proc. 4th Int. Conf. on Recent
REFERENCES Advances in Geotechnical Earthquake Engineering and Soil
Dynamics, S. Prakash, Editor, San Diego, CA, March 26-31.
Abdoun, T. and Wang, Y. [2002]. “Performance of retrofitted
pile foundations subjected to seismically induced lateral Finn, W. D. Liam, Thavaraj, T., Wilson, D. W., Boulanger, R.
spreading,” CDROM Proc. 8th US-Japan Workshop on W. and Kutter, B. [1999]. “Seismic analysis of piles and pile
Earthquake Resistant Design of Lifeline Facilities and groups in liquefiable sand,” Proc. 7th International Symposium
Countermeasures against Liquefaction, Tokyo, Japan. on Numerical Models in Geomechanics, NUMOG VI, Graz,
Austria, September, 287-292.
API [1995]. “Recommended practice for planning, designing,
and constructing fixed offshore platforms”, API RP 2A, Hamada, M., Yasuda, S., Isoyama, R.,and Emoto, K. [1986].
American Petroleum Institute. “Study on liquefaction induced permanent ground
displacements”, Report for the Association for Development
Bardet, J. P., Mace, N., Tobita, T. and Hu, J. [1999a]. “Large of Earthquake Prevention, Tokyo, Japan.
scale modeling of liquefaction-induced ground deformation,
Part I: A four –parameter MLR model”, Proc. 7th U.S.-Japan JRA [1996]. “Seismic design specifications of highway
Workshop on Earthquake Resistant Design of Lifeline bridges”, Japan Road Association, (English summary), in
Facilities and Countermeasures against Liquefaction, Earthquake Resistant Design Codes In Japan, Japan Society of
MCEER, November, 155-174. Civil Engineers, Tokyo, 2000, Ch.2, 1-21.

Bardet, J. P., Mace, N., Tobita, T. and Hu, J. [1999b]. “Large JWWA. [1997]. “Seismic design and construction guidelines
scale modeling of liquefaction-induced ground deformation, for water supply facilities.” (in English) Japan Water Works
Part II: MLR model, applications and probabilistic model”, Association, Tokyo, Japan. 150pp.
Proc. 7th U.S.-Japan Workshop on Earthquake Resistant
Design of Lifeline Facilities and Countermeasures against Kimura, M., Kosa, K. and Morita, Y. [1994]. “Full –scale
Liquefaction, MCEER, November, 175-190. failure tests on laterally loaded group piles”, Proc. 3rd Int.
Conf. Deep Foundation Practice incorporating PILETALK
Brandenburg, Scott J., Singh, Priyanshu, Boulanger, Ross W. International ’94, Singapore, 147-154.
and Kutter, Bruce L. [2001]. “Behavior of piles in laterally
spreading ground during earthquakes”, CDROM Proc. 6th Liu, L. and Dobry, R. [1995]. “Effect of liquefaction on lateral
CALTRANS Seismic Research Workshop, Sacramento, CA. response of piles by centrifuge model tests”, Nat. Ctr. for
Earthquake Engrg. Res. (NCEER) bull., Vol. 9(1), 7-11.
Dobry, R., Taboada, V. and Liu, L. [1995]. “Centrifuge
modeling of liquefaction effects during earthquakes,” Proc. 1st Martin, G.R., Finn, W.D. Liam & Seed, H.B. [1975].
Int. Conf. on Earthquake Engineering, K. Ishihara, Editor, “Fundamentals of liquefaction under cyclic loading.” J. of
Tokyo, Japan, Vol.3, 1291-1324. Geotech. Engrg. Div. ASCE, Vol. 101, GT5, 423-438.

SOAP 1 Page 15
Matlock, H. [1970]. “Correlations for design of laterally Wilson, D. W., Boulanger, R. W. and Kutter, B. L. [2000].
loaded piles in soft clay”, Proc. Offshore Technology “Observed lateral resistance of liquefying sand,” J. of
Conference, Houston, Texas, Paper 1204. Geotech. And Geoenvir, Engrg., ASCE, 126 (10), 898-906.

Matsui, T. and Oda, K. [1996]. “Foundation damage of Wu, G. and Finn, W. D. Liam [1997a]. “Dynamic elastic
structures, Soils and Foundations”, Special Issue, Japan analysis of pile foundations using the finite element method in
Geotechnical Society, Jan., 189-200. the frequency domain”, Canadian Geotechnical Journal, 34,
34-43.
Murchison, J. M. and O’Neill, M. W. [1984]. “An evaluation
of p-y relationships in cohesionless soils,” Proc. ASCE Sym. Wu, G. and Finn, W. D. Liam [1997b]. “Dynamic nonlinear
Analysis and Design of Pile Foundations, ASCE National analysis of pile foundations using the finite element method in
Convention, San Francisco, California, J. R. Meyer, Editor, the time domain”, Canadian Geotechnical Journal, 34, 144-
174-191. 152.

Ramos, R., Abdoun, T. and Dobry, R. [1999]. “Centrifuge Yasuda,, Yoshida, N., Kiku, H.and Adachi, K. [1999]. “A
modeling of effects of superstructure stiffness on pile bending simplified practical method for evaluating liquefaction-
moments due to lateral spreading”, Proc. 7th U.S.-Japan induced flow”, Proc. 7th U.S.-Japan Workshop on Earthquake
Workshop on Earthquake Resistant Design of Lifeline Resistant Design of Lifeline Facilities and Countermeasures
Facilities and Countermeasures against Liquefaction, against Liquefaction, November, 311-320.
November, 599-608.
Youd, T. L., Hansen, C.M., and Bartlett, S. F. [1999].
Reese, L. C., Wang, S. T. and Koop, F. D. [1974]. “Analysis “Revised MLR equations for predicting lateral spread
of laterally loaded piles in sand”, Proc. Offshore Technology displacement”, Proc. 7th U.S.-Japan Workshop on Earthquake
Conference, Houston, Texas, Vol. II, Paper 2080, 473-484. Resistant Design of Lifeline Facilities and Countermeasures
against Liquefaction, MCEER, November, 99-114.
Reese, L. C., Wang, S. T., Isenhower, W. M., Arrellaga, J. A.
and Hendrix, L. [2000]. “LPILE Plus, Version 4.0m,” Ensoft
Inc., Austin, Texas.

Schnabel, P.B., Lysmer, J. and Seed, H.B. [1972]. “SHAKE:

A Computer Program for Earthquake Response Analysis of
Horizontally Layered Sites,” Report EERC 71-12, University
of California at Berkeley, CA.

Seed, H.B. and Idriss, I. M. [1970]. “Soil moduli and

damping factors for dynamic response analysis,” Earthquake
Engineering Research Center, University of California,
Berkeley, CA., Report No. EERC70-10.

Thavaraj, T. and Finn, W. D. Liam [2000]. “A program for

dynamic analysis of a single pile using a Winkler model and
p-y curves”, Anabuki Chair of Foundation Geodynamics,
Kagawa University, Japan.

Vaid, Y. P.and Thomas, J. [1995]. “Liquefactionandpost-

liquefactionbehavior of sand,” J. of Geotech. And Geoenvir,
Engrg., ASCE, 126 (10), 898-906.

Weaver, T. J., Ashford, S. A. and Rollins, K. M. [2001].

“Development of p-y curves for a 0.6 m diameter CISS pile in
liquefied sand,” CDROM Proc. 6th CALTRANS Seismic
Research Workshop, Sacramento, CA.

Wilson, D.W., Boulanger, R.W. & Kutter, B.L. [1997]. Soil-

pile-superstructure interaction at soft or liquefiable sites.
Centrifuge data report for CSP1-5. Report No. UCD/CGMDR-
97/01-05, Center for Geotechnical Modeling, Dept. of Civil
and Environmental Eng., University of California at Davis,
CA.

SOAP 1 Page 16