University of Wollongong

Research Online
Faculty of Engineering - Papers (Archive) Faculty of Engineering and Information Sciences

2005

Pseudostatic Approach for Seismic Analysis of Piles in Liquefying Soil

D. S. Liyanapathirana
University of Wollongong, saml@uow.edu.au

H. G. Poulos
University of Sydney

Publication Details
This article was originally published as: Liyanapathirana, DS & Poulos, HG, Pseudostatic Approach for Seismic Analysis of Piles in Liquefying Soil, Journal of Geotechnical and Geoenvironmental Engineering, 2005, 131(12), 1480-1487. Copyright American Society of Civil Engineers. Original journal available here.

Research Online is the open access institutional repository for the University of Wollongong. For further information contact the UOW Library: research-pubs@uow.edu.au

Pseudostatic Approach for Seismic Analysis of Piles in Liquefying Soil

D. S. Liyanapathirana, M.ASCE,1 and H. G. Poulos, F.ASCE2
Abstract: The performance of pile foundations during an earthquake signicantly inuences the integrity of structures supported by them. Therefore, in the overall seismic design process of the structures, modeling of the soilpile-superstructure interaction is an essential part. Although nite element based coupled analysis of the soilpile-superstructure interaction models have the potential to provide accurate results, they are computationally expensive and often complex to utilize. In practice, many geotechnical engineers tend to use simple methods for obtaining the internal response of piles subjected to earthquake loading. Therefore this paper presents a simple pseudostatic approach where a single pile is considered, including the contribution of the superstructure to the pile and the interaction between the pile and the soil. The method involves two main steps. First a nonlinear free-eld site response analysis is carried out to obtain the maximum ground displacements along the pile and the degraded soil modulus over the depth of the soil deposit. Next a static load analysis is carried out for the pile, subjected to the maximum free-eld ground displacements and the static loading at the pile head based on the maximum ground surface acceleration. The method has been veried using an independent dynamic pile analysis program developed by the writers for the seismic analysis of piles in liquefying soil. It is demonstrated that the new method gives good estimates of pile bending moment, shear force, and displacement, despite its relative simplicity. The method is then used to compute the response of pile foundations during the Kobe 1995 earthquake and some centrifuge tests found in the literature where extensive soil liquefaction has been observed. Very good agreement is observed between computed and recorded pile bending moments. DOI: 10.1061/ASCE1090-02412005131:121480 CE Database subject headings: Seismic analysis; Piles; Liquefaction; Earthquakes. and Kausel 1982; Kavvadas and Gazetas 1993. Winkler models that take into account the nonlinear soil behavior have been developed by Penzien 1970; Kagawa 1980; Kagawa and Kraft 1981; Norris 1994; El Naggar and Novak 1996; Nogami and Konagai 1988; and Tabesh and Poulos 2000, 2001a but they lack the ability to predict pile behavior when the soil around the pile starts to liquefy. For the seismic analysis of piles in liquefying soil, Winkler type models have been developed by Kagawa 1992; Yao and Nogami 1994; Fujii et al. 1998; and Liyanapathirana and Poulos 2005. Although these models are one-dimensional, a dynamic nite element analysis has to be carried out to obtain the pile response in liquefying soil. Recently, pseudostatic approaches for the seismic analysis of pile foundations have emerged. In pseudostatic approaches, a static analysis is carried out to obtain the maximum bending moment and shear force developed in the pile due to earthquake loading. These methods are attractive for design engineers when compared to difcult but more complex dynamic analyses. For piles in nonliquefying soil Abghari and Chai 1995 and Tabesh and Poulos 2001b have developed pseudostatic approaches. When liquefaction is of concern, the stiffness of the soil is dramatically reduced and the effect of the reduced stiffness should be incorporated in the analysis. Therefore in this paper a pseudostatic approach, which requires relatively little computational effort, is presented for the analysis of piles in liquefying soil. Results obtained from the pseudostatic approach are compared with the results given by a dynamic analysis and centrifuge data and despite its simplicity, the pseudostatic approach results, which are in good agreement.

Introduction
Liquefaction of saturated soil subjected to earthquake loading is one of the major factors affecting the behavior of pile foundations and subsequent building failure in seismically active areas. This has been clearly demonstrated during past earthquakes that have occurred in the USA e.g., 1989 Loma-Prieta, Japan e.g., 1995 Kobe, and Mexico e.g., 1995 Manzanillo. At many instances, pile failure in liqueed ground occurred due to the inadequacy of the pile to sustain large shear forces and bending moments developed during an earthquake event. Hence there is a great demand for numerical procedures which can be used to predict pile behavior in liquefying ground during an earthquake event. Although one-dimensional Winkler models have become popular for the seismic analysis of pile foundations, most of them can be used only for the linear analysis of pilesoil interaction in nonliquefying soil e.g., Novak 1974; Dobry et al. 1982; Kaynia
1 Senior Lecturer, Dept. of Civil, Mining and Environmental Engineering, Univ. of Wollongong, Northelds Ave., Wollongong, NSW 2522, Australia. 2 Senior Principal, Coffey Geosciences Pty. Ltd. and Emeritus Professor of Civil Engineering, Dept. of Civil Engineering J05, Univ. of Sydney, Sydney, NSW 2006, Australia. Note. Discussion open until May 1, 2006. Separate discussions must be submitted for individual papers. To extend the closing date by one month, a written request must be led with the ASCE Managing Editor. The manuscript for this paper was submitted for review and possible publication on December 23, 2002; approved on August 17, 2004. This paper is part of the Journal of Geotechnical and Geoenvironmental Engineering, Vol. 131, No. 12, December 1, 2005. ASCE, ISSN 10900241/2005/12-14801487/$25.00.

1480 / JOURNAL OF GEOTECHNICAL AND GEOENVIRONMENTAL ENGINEERING ASCE / DECEMBER 2005

Pseudostatic Approach
The pseudostatic methods were presented by Abghari and Chai 1995 and Tabesh and Poulos 2001b for the nonliquefying soil, and the inertial force acting at the pile head is represented by the product of the cap-mass and the spectral acceleration Dowrick 1977. By comparing the results given by the pseudostatic method with the results given by a dynamic nite element analysis, Abghari and Chai 1995 concluded that the inertial force should be reduced to 25% for the pile deection and to 50% for the bending moment and shear force to obtain the results in agreement with the dynamic nite element analysis. They made this conclusion by analyzing only one example and the method has not been generalized. However, they have taken into account the nonlinear behavior of the soil in their analysis. Tabesh and Poulos 2001b compared results given by pseudostatic and dynamic analyses for different pile and soil properties. They applied the full inertial force at the pile head and observed an excellent agreement with the dynamic analysis for the cases without cap-mass, but when the cap-mass increased they showed that the pseudostatic approach overestimates the maximum bending moment and shear force developed in the pile. However, they carried out an elastic free-eld site response analysis and in the pile analysis the nonlinear behavior of the soil is also not taken into account. In the pseudostatic approach proposed by Ishihara and Cubrinovski 1998 for the pile foundations in the soil deposits subjected to lateral spreading, inertial force at the pile head has not been considered to obtain the maximum bending moment and shear force developed in the pile. Here the pseudostatic approach has been extended for a liquefying soil, where the degradation of shear modulus of the soil occurs with the generation of pore water pressure in the soil. Although spectral acceleration has been used by Abghari and Chai 1995 and Tabesh and Poulos 2001b, it has been found that the inertial force at the pile head calculated using the spectral acceleration, based on the effective stress analysis, is overestimated when the surrounding soil starts to liquefy. The numerical studies carried out using the dynamic nite analysis show that when the surrounding soil starts to liquefy, maximum pile head acceleration closely agrees with the maximum ground surface acceleration. Hence pseudostatic analysis has been carried out by applying inertial force at the pile head calculated based on the maximum ground surface acceleration, instead of the spectral acceleration. In the pseudostatic approach presented here, maximum pile bending moment, shear force, and displacement are obtained by performing a static load analysis for the pile, involving two main stages as follows. 1. A free-eld site response analysis is carried out to obtain the maximum ground displacement and the minimum effective vertical stress at each depth of the soil deposit and the maximum ground surface acceleration during the earthquake loading. 2. Next a static load analysis is carried out for the pile, subjected to the maximum free-eld ground displacements and the static loading at the pile head, which is given by the maximum ground surface acceleration multiplied by the capmass. Here the maximum ground surface acceleration, minimum effective stress, and the maximum ground displacement at each depth have been obtained from the free-eld site response analysis developed by Liyanapathirana and Poulos 2002b,c. The

static load analysis of the pile is carried out by modeling the pile as a nonlinear beam. Soilpile interaction is modeled using the method of a beam on a nonlinear Winkler foundation. Cap-mass at the pile head represents the mass of the superstructure. Although the superstructure supported by pile foundations is a multi-degree of freedom system, in the design of pile foundations, it is reduced to a single mass at the pile head, to simplify the analysis. The partial differential equation of a beam on a Winkler foundation is given by E PI P

4U P = KxU f f U P + M amax z4

where E P = Youngs modulus of the pile material, I P = inertia of the pile, U P = pile displacement, U f f = free-eld lateral soil displacement, Kx = spring coefcients of the Winkler model, M = cap-mass, and amax = maximum ground surface acceleration. Eq. 1 is solved using the nite element method. The spring coefcients of the Winkler model have been obtained by integrating Mindlins equation over a rectangular area as explained in the companion paper by Liyanapathirana and Poulos 2005. At each depth, the spring coefcients are calculated based on the minimum effective stress obtained from the freeeld site response analysis. Usually it is assumed that the liqueed soil does not have any stiffness. However, due to the stiffness contrast between the pile and the liqueed soil, computationally it is difcult to carry out an analysis with a near-zero shear modulus. Therefore in the numerical studies, a lower limit has been set for the initial effective vertical stress, below which effective vertical stress is not allowed to decrease and pore pressures are not allowed to build up. By analyzing eld data recorded at the Port Island site during the Kobe 1995 earthquake, Davis and Berril 1998 reported that the shear wave velocity of the liqueed region is about 25 m / s. Ishihara and Towhata 1982 also suggested that since shear stress application during earthquakes is multidirectional, even when shear stresses are reduced to zero in one direction, there will always be some shear stress left in the soil. This was demonstrated in the rotational simple shear tests performed by Ishihara and Yamasaki 1980. During the one-dimensional free-eld site response analyses carried out by Ishihara and Towhata 1982, this lower limit of effective stress is set at 3% of the initial effective overburden pressure. The degradation of soil modulus gives rise to nonhomogeneity of the soil prole, even if it was initially homogeneous. The use of the Mindlin equation is of course approximate for soils, which are not homogeneous and isotropic, but can give results of adequate accuracy for many cases of nonuniform soil proles Poulos 1982. The nonlinear behavior of soil at the pilesoil interface has been modeled using a plastic slider in series with spring coefcients of the Winkler model, which represents the lateral pressure at the pilesoil interface, as shown in Fig. 1. This lateral pressure is monitored and an iterative procedure is used to keep it below the ultimate lateral pressure of the soil. According to Broms 1964, for noncohesive soils, the ultimate lateral pressure is given by Pu = 3 v

1 + sin 1 sin

In this model, displacement of the soil adjacent to the pile wall is represented by the displacement of the plastic slider, which is different from the displacement of the soil away from the pile

JOURNAL OF GEOTECHNICAL AND GEOENVIRONMENTAL ENGINEERING ASCE / DECEMBER 2005 / 1481

5.

6.

in Step 1. The maximum soil displacement prole calculated in Step 1 is applied to the pile through these springs as shown in Fig. 1. A plastic slider is used in series with each spring to limit the pressure at the pilesoil interface to the ultimate lateral pressure given by Eq. 2. A nonlinear static load analysis is carried out to obtain the prole of maximum pile displacement, bending moment, and shear force along the pile by applying the lateral force at the pile head calculated in Step 3 and the maximum soil displacement prole along the pile calculated in Step 1 simultaneously to the pile as shown in Fig. 1.

Verication of the Proposed Method

Fig. 1. Beam on Winkler foundation model for pseudostatic analysis

represented by the maximum displacement at each depth obtained from the free-eld site response analysis. The calculation steps involved in this new approach can be summarized as below. 1. First, a free-eld site response analysis is performed by taking into account the pore pressure generation and dissipation in the soil deposit due to the earthquake loading. From this analysis, the maximum ground surface acceleration, maximum ground displacement along the length of the pile, and the minimum effective stress level attained during the seismic activity can be obtained. 2. The superstructure is modeled as a concentrated mass at the pile head. Generally the superstructures supported by pile foundations are multi-degree of freedom systems, but in the design of pile foundations, the superstructure is reduced to a single mass at the pile head to simplify the analysis. 3. The lateral force to be applied at the pile head is the capmass multiplied by the maximum ground surface acceleration obtained from the free-eld site response analysis, as shown in Fig. 1. 4. The pilesoil interaction is modeled using springs as shown in Fig. 1. Spring coefcients are calculated by integrating the Mindlins equation as described by Liyanapathirana and Poulos 2002a, 2005, using the minimum shear modulus corresponding to the minimum effective vertical stress calculated

The proposed pseudostatic approach has been veried for soil deposits with uniform relative density and for two-layer soil deposits using the dynamic benchmark analysis described in the companion paper by Liyanapathirana and Poulos 2005. Soil Deposits with Uniform Relative Density In this section the proposed pseudostatic approach has been veried for soil deposits with uniform relative density. Results have been obtained for a soil deposit with 50% relative density by changing the length of the pile and the diameter of the pile. It is assumed that the pile extends down to the bottom of the soil deposit. The shear modulus of the soil is assumed to vary with the effective stress level of the soil as shown below

Gs = G0

1 + 2K0 v 3 100

0.5

MPa

where v = effective stress level of the soil, K0 = coefficient of earth pressure at rest, and the initial shear modulus, G0, is assumed to be 30. Tables 1 and 2 show the maximum pile bending moment obtained for different pilesoil congurations. The depth of the soil deposit ranges between 15 and 30 m and the pile diameter ranges from 0.3 to 1.2 m as given in Table 1. The concrete piles used for the analysis have a Youngs modulus of 3 104 MPa and a den-

Table 1. Maximum Bending Moment MN m Obtained from Dynamic and Pseudostatic Analyses Dr = 50% d = 0.3 m Length m 15 20 25 30 Dynamic 0.13 0.07 0.10 0.14 Pseudostatic 0.12 0.06 0.95 0.14 Dynamic 0.86 0.46 0.57 0.53 d = 0.6 m Pseudostatic 0.93 0.46 0.57 0.69 Dynamic 4.11 2.16 2.30 2.50 d = 0.9 m Pseudostatic 4.39 2.16 2.55 2.70 Dynamic 12.1 6.08 7.0 7.41 d = 1.2 m Pseudostatic 13.4 6.08 7.27 7.41

Table 2. Maximum Bending Moment MN m Obtained from Dynamic and Pseudostatic Analyses Dr = 60% d = 0.3 m Length m 15 20 25 30 Dynamic 0.12 0.17 0.13 0.16 Pseudostatic 0.12 0.16 0.10 0.15 Dynamic 0.86 1.20 0.63 0.80 d = 0.6 m Pseudostatic 1.05 1.24 0.50 0.70 Dynamic 4.03 5.36 1.98 2.70 d = 0.9 m Pseudostatic 4.03 5.36 1.98 2.99 Dynamic 11.0 12.9 5.93 8.31 d = 1.2 m Pseudostatic 11.3 12.9 5.93 7.31

1482 / JOURNAL OF GEOTECHNICAL AND GEOENVIRONMENTAL ENGINEERING ASCE / DECEMBER 2005

Fig. 2. Amount of pore pressure generation and free-eld displacement along the depth of the soil deposit and the maximum ground surface acceleration

Fig. 3. Variation of pile moment, shear, and displacement along depth for 15 m pile with diameters 0.3 and 1.2 m Dr = 50%

sity of 2,400 kg/ m3. The soil deposits used for the analysis have a density of 1,900 kg/ m3, permeability of 5.5e5 m / s, and a friction angle of 30. It is assumed that the water table is 2.0 m below the ground surface. First, a free-eld site response analysis is carried out to obtain the maximum soil displacements along the depth of the soil deposit, maximum ground surface acceleration, and the minimum effective vertical stresses along the depth of the soil deposit. The 1995 Kobe earthquake record given in the companion paper by Liyanapathirana and Poulos 2005 scaled to 0.25g has been used as the excitation source. As discussed in a previous section, soil deposits retain some shear strength even after liquefaction. Therefore, during the free-eld site response analysis, effective stress of the soil is reduced only up to 2% of the initial effective overburden pressure at each depth. Fig. 2 shows the amount of pore pressure generation, positive and negative ground displacement envelopes, and the maximum ground surface acceleration for the four soil deposits considered for the analysis. The liqueed depth ranges between 6 and 8 m for these soil deposits. During the pseudostatic analysis of the pile, the pile head is assumed to be restrained against rotational movement and the pile tip is assumed to be restrained against lateral movement. The cap-mass carried by each pile conguration is calculated based on the ultimate load carrying capacity of piles in sand with a factor of safety of 2.5.

Despite its simplicity, Table 1 shows that the pseudostatic analysis gives results in close agreement with the benchmark dynamic analysis. The agreement between results is not only conned to the point of maximum bending moment but occurs along the whole length of the pile. Fig. 3 shows the maximum positive and negative bending moment, shear force, and displacement envelopes along the pile obtained from the dynamic analysis, for the 15 m pile with 0.3 and 1.2 m diameters, given in Table 1 during the earthquake loading. These gures also show the maximum bending moment, shear force, and displacement obtained from the pseudostatic analysis, and they demonstrate the close agreement between the dynamic and pseudostatic analyses along the length of the pile. It is interesting to see that in some parts, the static prole matches with the maximum positive envelope and in other parts, it matches with the maximum negative envelope. In Fig. 4, the maximum bending moment and shear force proles along the 15 m long pile with Dr = 50% are given for a free head pile where the pile extends 1 m above the ground surface. The maximum ground surface acceleration, pore pressure ratio, and ground displacements at each depth are given in Fig. 2a. The concrete pile used for the analysis and the soil have the same properties as in the previous analysis. Maximum bending moment proles are given for pile diameters of 0.3, 0.6, 0.9, and 1.2 m. When carrying out the pseudostatic analysis, the pile head is assumed to be at the ground surface. Hence in addition to the iner-

JOURNAL OF GEOTECHNICAL AND GEOENVIRONMENTAL ENGINEERING ASCE / DECEMBER 2005 / 1483

Fig. 4. Variation of pile moment along depth for a free head pile with diameters 0.3, 0.6, 0.9, and 1.2 m Dr = 50%

Fig. 5. Variation of pile moment and shear along depth for a 15 m pile with diameter 0.3 m with and without cap-mass Dr = 50%

Two-Layer Soil Deposits tial force, the moment due to eccentricity of the inertial force acting at the pile head is applied to the pile in the pseudostatic analysis. Here the inertial force at the pile head is given by the product of capmass and the ground surface acceleration. The capmass is calculated based on the ultimate load carrying capacity of piles in sand with a factor of safety of 2.5. The agreement between dynamic and pseudostatic analyses conrms that, irrespective of the boundary conditions at the pile head and the pile tip, the pseudostatic approach can be used to estimate the internal response of the pile. Fig. 5 shows the internal response for the 0.3 m diameter and 15 m long pile without cap-mass and with a 50 t cap-mass, founded in the soil deposit shown in Fig. 4a. In Fig. 3a the internal response for the same pile is given when the cap-mass is 20 t. When the pile does not carry a cap-mass, excellent agreement between the pseudostatic and dynamic analyses can be observed in Fig. 5. With the increase in cap-mass, the agreement between the pseudostatic and dynamic analyses reduces and the pseudostatic method overestimates the maximum pile bending moment. In the proposed pseudostatic method, the inertial force based on the maximum ground acceleration and the maximum ground displacement prole is applied to obtain the internal pile response. In the dynamic analysis, the maximum free-eld ground displacment will not occur in phase with the maximum ground surface acceleration. As a result, the pseudostatic method overestimates the maximum pile bending moment in some cases compared to the pile response predicted by the dynamic benchmark analysis. Although results are given here only for some selected cases, the method has given reasonable agreement with the dynamic analysis for different soil conditions and pile congurations. It has been found that the maximum values are given at the same depth and the difference in magnitude is less than 25%, which is generally acceptable for practical pile design purposes. In this section the psedostatic method proposed in a previous section has been veried for two-layer soil deposits with a nonliquefying soil layer overlain by a liquefying soil. The relative density of the liquefying soil is 50% and that of the nonliquefying layer is 90%. For the Dr = 50% case, the friction angle is 30 and G0 Eq. 3 is 30 MPa, while for the Dr = 90% case, friction angle is 35 and G0 is 35 MPa. It is assumed that both layers have a density of 1,900 kg/ m3 and permeability of 5.5e5 m / s. The water table is considered to be at the ground surface.

Fig. 6. Maximum ground displacement envelopes along the depth and maximum ground surface acceleration for layered soil deposits

1484 / JOURNAL OF GEOTECHNICAL AND GEOENVIRONMENTAL ENGINEERING ASCE / DECEMBER 2005

Fig. 7. Variation of pile bending moment along depth for a 25 m long pile with diameter 0.6 m in a two-layer soil deposit

Fig. 8. Variation of pile bending moment along depth for a 25 m long pile with diameter 0.9 m in a two-layer soil deposit

The total thickness of the two-layer soil deposit is 25 m and the analysis is carried out by varying the thickness of the top liquefying soil layer. The four cases considered have top liquefying layers of 4, 8, 12, and 16 m. The 1995 Kobe earthquake record scaled to 0.25g shown in Fig. 3 has been used as the excitation source. The free head piles used for the analysis extends up to the bottom of the soil deposit and the spring coefcients used to represent pilesoil interaction are calculated based on the minimum effective stress at each depth obtained from the free-eld site response analysis. Two pile diameters are considered, 0.6 and 0.9 m, and each pile carries a cap-mass of 8.4 104 and 1.66 105 kg, respectively. The concrete piles used for this analysis have a Youngs modulus of 3 104 MPa and a density of 2,400 kg/ m3. Fig. 6 shows the positive and negative ground displacement envelopes and the maximum ground surface acceleration obtained from the free-eld site response analysis. Figs. 7 and 8 show the bending moment along the pile obtained from the pseudostatic approach and the maximum positive and negative bending moment envelopes obtained from the benchmark analysis for piles with diameters 0.6 and 0.9 m, respectively. It can be seen that for all cases, the maximum bending moment given by the pseudostatic approach reasonably agrees with those given by the dynamic bench mark analysis.

Hyogoken-Nambu 1995 earthquake reported by Ishihara and Cubrinovski 1998. The centrifuge test by Abdoun et al. 1997 was carried out to study the pile response during lateral spreading. Details of this test are given in the companion paper by Liyanapathirana and Poulos 2005. Fig. 9 shows the maximum bending moment envelope obtained from the pseudostatic analysis and the measured maximum bending moments at several depths during the centrifuge test. In this case, pile does not carry a cap-mass and pile head is free. Therefore the pseudostatic analysis is carried out by applying only the maximum free-eld ground displacements at each depth along the pile. It can be seen that the calculated values agree well with the values recorded during the centrifuge test. The eld measurements made in the piles at Pier 211 in Uozakihama Island after the Hyogoken-Nambu earthquake oc-

Comparison with Field and Centrifuge Data

In this section, the proposed pseudostatic method has been used to estimate the maximum bending moments developed in a pile used for a centrifuge test carried out by Abdoun et al. 1997 and the bored piles at Bridge Pier 211 in Uozakihama Island after the

Fig. 9. Comparison of maximum bending moment along pile with centrifuge data from Abdoun et al. 1997

JOURNAL OF GEOTECHNICAL AND GEOENVIRONMENTAL ENGINEERING ASCE / DECEMBER 2005 / 1485

Fig. 12. Bending moment of the pile at Bridge Pier 211 in Uozakihama Island calculated from the pseudostatic approach and Ishihara and Cubrinovski 1998 results with = 1 102 Fig. 10. Cracks observed in the piles at bridge pier 211 in Uozakihama Island Ishihara and Cubrinovski with permission, 1998

curred on January 17, 1995, and reported by Ishihara and Cubrinovski 1998, have been simulated using the pseudostatic approach presented in this paper. Fig. 10 shows the crack distributions observed in piles after the earthquake. The concrete piles at bridge Pier 211 are 46 m long and the diameter is 1.5 m. The water table is 2.0 m below the ground surface and the upper 20 m of this site consists of Masado sand with an initial shear modulus of 57.8 MN/ m2 and density of 2,000 kg/ m3 Tokimatsu et al. 1998. Soil liquefaction was observed in the Masado sand layer below the water table only. Therefore only the top 20 m layer was analyzed using the effective stress method and incorporating pore pressure generation and dissipation. For this analysis the cyclic shear strength curve for the Masado sand given by Ishihara 1997 was used. It was assumed that the base rock had a density of 2,200 kg/ m3 and shear modulus of 75 GN/ m2. The lower end of the RC pile was assumed to be xed while the pile head was assumed to be xed to the footing but free to move in the horizontal direction.

Fig. 11 shows the maximum ground displacement at each depth obtained from the site response analysis and the free-eld ground displacements used by Ishihara and Cubrinovski 1998. During the earthquake, only the ground surface displacements were recorded. Based on these data, the lateral ground surface displacement at the vicinity of bridge Pier 211 was about 1.0 m, which agrees well with the maximum ground surface displacement obtained from the numerical model. Ishihara and Cubrinovski 1998 used a cosine function through the liqueed layer down to a depth of 20 m to distribute the ground surface displacement, as shown in Fig. 11. There is a slight discrepancy between this assumed displacement distribution and the displacement distribution obtained from the numerical model. Fig. 12 shows the maximum bending moment prole along the pile obtained from the pseudostatic approach and those calculated by Ishihara and Cubrinovski 1998. The predictions made by the pseudostatic approach agree well with Ishiharas results. The yield moment for these piles was about 5 MN m. The computed maximum bending moment prole exceeds the yield moment near the pile head and around the boundary between the liqueed and nonliqueed layers. This is consistent with the location of cracks observed after the earthquake shown in Fig. 10.

Conclusions
This paper has described a pseudostatic approach that can be used to compute the maximum bending moment and shear force developed in a pile founded in liquefying soil. An effective-stressbased free-eld site response analysis is rst carried out and the resulting ground displacements, degraded soil stiffness, and inertial force at the pile head, based on the cap-mass and the maximum ground surface acceleration, are applied to the pile statically to obtain the internal pile response. The spring coefcients of the Winkler model used in the pseudostatic analysis are derived from Mindlins equations. The results presented in the paper suggest that the new method has promise in practical applications. For a few cases the new method overestimated the pile bending moment and shear force but the values were within 25% of those obtained from the dynamic analysis. Both dynamic and pseudostatic analyses give peak values at the same locations. Also the pseudostatic method has been veried for two layer soil deposits with liquefying and

Fig. 11. Maximum ground displacement at each depth obtained from the site response analysis and estimated by Ishihara and Cubrinovski 1998 based on eld measurements

1486 / JOURNAL OF GEOTECHNICAL AND GEOENVIRONMENTAL ENGINEERING ASCE / DECEMBER 2005

nonliquefying soil layers. The pile performance observed during a centrifuge test and a real earthquake has been simulated using the pseudostatic approach. It is found that the pile response calculated from the pseudostatic approach is consistent with the observed pile behavior.

Acknowledgments
This work is part of a project on Design of Pile Foundations for Seismically Active Areas funded by the Australian Research Council and this support is gratefully acknowledged. Also the writers would like to thank the reviewers for their thorough review and useful comments.

References
Abdoun, T., Dobry, R., and ORouke, T. D. 1997. Centrifuge and numerical modeling of soil-pile interaction during earthquake induced soil liquefaction and lateral spreading. Observation and Modeling in Numerical Analysis and Model Tests in Dynamic Soil-Structure Interaction ProblemsProc., Sessions held in conjunction with Geo-Logan 97, Logan, Utah, 7690. Abghari, A., and Chai, J. 1995. Modeling of soil-pile-superstructure interaction for bridge foundations. Proc., Performance of Deep Foundations under Seismic Loading, J. P. Turner, ed., ASCE, New York, 4559. Broms, B. B. 1964. Lateral resistance of piles in cohesive soils. J. Soil Mech. Found. Div., 902, 2763. Davis, R. O., and Berrill, J. B. 1998. Energy dissipation and liquefaction at Port Island, Kobe. Bull. New Zealand Natl. Soc. Earthquake Eng., 31, 3150. Dobry, R., Vicente, E., ORourke, M. J., and Rosset, J. M. 1982. Horizontal stiffness and damping of single piles. J. Geotech. Eng. Div., Am. Soc. Civ. Eng., 1083, 439459. Dowrick, D. J. 1977. Earthquake resistant design: A manual for engineers and architects, Wiley, New York. El Naggar, M. H., and Novak, M. 1996. Non-linear analysis for dynamic lateral pile response. Soil Dyn. Earthquake Eng., 154, 233244. Fuji, S., Cubrinovski, M., Tokimatsu, K., and Hayashi, T. 1998. Analyses of damaged and undamaged pile foundations in liqueed soils during the 1995 Kobe Earthquake. Proc., 1998 Conf. on Geotechnical Earthquake Engineering and Soil Dynamics III, Seattle, Wash., 2, 11871198. Ishihara, K. 1997. Geotechnical aspects of the 1995 Kobe Earthquake. Terzaghi Oration, Proc., 14th Int. Conf. on Soil Mechanics and Foundation Engineering, Hamburg, Germany. Ishihara, K., and Cubrinovski, M. 1998. Performance of large-diameter piles subjected to lateral spreading of liqueed deposits. Thirteenth Southeast Asian Geotechnical Conf., Taipei, Taiwan. Ishihara, K., and Towhata, I. 1982. Dynamic response analysis of level ground based on the effective stress method. Soil Mechanics Transient and Cyclic Loads, G. N. Pande and O. C. Zienkiewicz, eds., Wiley, New York, 133171.

Ishihara, K., and Yamasaki, F. 1980. Cyclic simple shear tests on saturated sand in multi-directional loading. Soils Found. 201, 4559. Kagawa, T. 1980. Soil-pile-structure interaction of offshore structures during an earthquake. Proc., Annual Offshore Technology Conf., Houston, Tex., 235245. Kagawa, T. 1992. Lateral pile response in liquefying sand. Proc., 10th World Conf. on Earthquake Engineering, Madrid, Spain, Paper No. 1761. Kagawa, T., and Kraft, L. M. 1981. Lateral pile response during earthquakes. J. Geotech. Eng. Div., Am. Soc. Civ. Eng., 10712, 17131731. Kavvadas, M., and Gazetas, G. 1993. Kinematic seismic response and bending of free-head piles in layered soils. Geotechnique, 432, 207222. Kaynia, A. M., and Kausel, E. 1982. Dynamic behavior of pile groups. Proc., Second Int. Conf. on Numerical Methods in Offshore Piling, 509532. Liyanapathirana, D. S., and Poulos, H. G. 2002a. Numerical model for seismic analysis of piles in liquefying soil. Deep Foundation Congress, Orlando, Fl., 274289. Liyanapathirana, D. S., and Poulos, H. G. 2002b. Numerical simulation of soil liquefaction due to earthquake loading. Soil Dyn. Earthquake Eng., 22, 511523. Liyanapathirana, D. S., and Poulos, H. G. 2002c. A numerical model for dynamic soil liquefaction analysis. Soil Dyn. Earthquake Eng., 22, 10071015. Liyanapathirana, D. S., and Poulos, H. G. 2005. Seismic lateral response of piles in liquefying soil. J. Geotech. Geoenviron. Eng., 13112, 14661479. Nogami, T., and Konagai, K. 1988. Time domain exural response of dynamically loaded single piles. J. Eng. Mech., 1149, 15121525. Norris, G. M. 1994. Seismic bridge pile foundation behavior. Proc., Int. Conf. on Design and Construction of Deep Foundations, 1, 27136. Novak, M. 1974. Dynamic stiffness and damping of piles. Can. Geotech. J., 11, 574598. Penzien, J. 1970. Soil-pile foundation interaction in earthquake engineering, R. L. Wiegel, ed., Prentice-Hall, Englewood Cliffs, N.J. Poulos, H. G. 1982. Developments in the analysis of static and cyclic lateral response of piles. Proc., 4th Int. Conf. on Numerical Methods in Geomechanics, Canada, 11171135. Tabesh, A., and Poulos, H. G. 2000. A simple method for the seismic analysis of piles and its comparison with the results of centrifuge tests. Proc., 12th World Conf. on Earthquake Engineering, Auckland, New Zealand, Paper No. 1203. Tabesh, A., and Poulos, H. G. 2001a. The effect of soil yielding on seismic response of single piles. Soils Found., 413, 116. Tabesh, A., and Poulos, H. G. 2001b. Pseudostatic approach for seismic analysis of single piles. J. Geotech. Geoenviron. Eng., 1279, 757765. Toimatsu, K., Oh-Oka, H., Satake, K., Shamoto, Y., and Asaka, Y. 1998. Effects of lateral ground movements on failure patterns of piles in the 1995 Hyogoken-Nambu Earthquake. Proc., Geotechnical Earthquake Engineering and Soil Dynamics III, Reston, Va., 11751186. Yao, S., and Nogami, T. 1994. Lateral cyclic response of piles in viscoelastic Winkler subgrade. J. Eng. Mech., 1204, 758775.

JOURNAL OF GEOTECHNICAL AND GEOENVIRONMENTAL ENGINEERING ASCE / DECEMBER 2005 / 1487