Вы находитесь на странице: 1из 9
PSEUDOSTATIC APPROACH FOR SEISMIC ANALYSIS OF SINGLE PILES By A. Tabesh' and H. G. Poulos,’ Fellow, ASCE 8 subjected to lateral seisthic excitation. The method involves two main steps: computa free-field soil movements caused by the earthquake and the analysis of the sponse of the pile to the niaximum free-field soil movements (considered as static movements) plus a static loading at the pile head, which depends fon the computed spectral acceleration of the structure being supported. The applicability of this approach has bbeen verified by an independent benchmark analysis developed by the writers. It is demonstrated that the roposed method yields reasonable estimates of the pile maximum moment and shear. The methodology is then used to obtain the response of the Ohba-Ohashi bridge in Japan to one of the earthquakes that occurred in the 1980s. Good agreement is found between the computed and measured pile moments. INTRODUCTION Under the assumption of linear soil behavior, sophisticated ‘numerical approaches based on both the boundary-element ‘method and the finite-element method have been developed to evaluate the response of single piles and pile groups to earth- ‘quake excitation [for example, Kaynia and Kausel (1982), Sen et al. (1985), and Mamoon (1990). These methods are almost ‘entirely in the frequency domain and give the pile response to a single harmonic excitation. Their application to earthquake motion therefore requires application of a Fourier transfor- mation. They are mathematically cumbersome, involve many numerical approximations, and require substantial computa- tional effort, which is justified only for very large projects. Not all of these methods yield identical results, even when the fundamental, assumptions are similar (Tabesh and Poulos 1997). ‘Several simplified approaches for the analysis of single piles hhave also been developed that can be used with little com- putational effort. Methods based on a Winkler hypothesis have siven results that are often in remarkable agreement with the ‘mathematical models. Novak (1974), for example, assured a plane-strain condition at the pile soil interface, and Gazetas, Dobry, Kaynia and others obtained the coefficients of their Winkler models by equating the head displgeement of the pile to the displacement obtained in separate finite-element analy- ses. These simplifying assumptions should be bom in mind ‘when using these methods. For example, if, instead ofthe pile- hhead deflection, the aim is the calculation of the maximum ‘moment ina pile, these Winkler models may give less accurate results. Gazetas and his coworkers {for example, Makris and Gazetas (1992)] have expanded their Winkler model to the case of pile groups and accounted for the effect of pile-to-pile interaction with the help of the Poulos static superposition ap- proach (Poulos and Davis 1980), together with certain simpli- fied attenuation factors for the motion of waves in an elastic medium. In almost all ofthe developed methods, the primary concer has been the evaluation of dynamic response of the superstruc- ture. The evaluation of internal pile response (ie. pile moment and shear) has been a secondary issue. In fact, researchers have generally looked at pile foundations in the same way as they "Se Res, Asst, Dept. of Civ. Engrg. Univ. of Sydney. Sydney, NSW 2006, Australia: currently, No, 169 Dr. Behesti Ave. Tehran, Iran St Prin, Coffey Pariners International, and Prof. of Cv. Engre., Uni, ‘of Sydney, Sydney, NSW 2006, Australia Note. Discussion open until February 1, 2002. To extend the closing date one month, a written request must be fled withthe ASCE Manager ‘of Journals. The manusript for this paper wae submitted for review and possible publication on March 13, 1998; revised April 24, 2001, This paper is part of the Journal of Geotechnical and Gecenvironmental Engineering, Vol. 127, No. 9, September. 2001. ASCE, ISSN 1030 (0241/01/0009.0757-0765/88,00 + §.50 per page. Paper No. 17875. ook at surface foundations; ie., they have been secking the pile-head impedances and the pile-head motions to be used in the subsequent dynamic analysis of the superstructure, Several extensive parametric studies reported in the literature have been almost entirely confined to the effects of different factors ‘on pile-head impedance and motion and rarely have the effects on pile intemal response been examined, The reason for this appears to be because the internal pile response is heavily dependent upon the dynamic behavior of the superstructure, which is usually a complex multidegree-of-freedom system ‘whose dynamic behavior changes from case to case. ‘Many pile failures during earthquakes have occurred be- cause of the inadequacy of the pile to withstand large induced ‘moments and shears (Mizuno 1987; Mizuno et al. 1996). The pile designer, who often is different from the person who per- forms structural dynamic analysis, therefore needs to know approximately what the maximum seismically induced internal ‘moment and shear of the pile would be. In this paper a simple ppseudostatic approach is evaluated, which appears to provide 2 good approximation to the maximum pile moment and shear ‘and with little computational effort. DEVELOPMENT OF BENCHMARK ANALYSIS As the first step in developing the pseudostatic approach, a benchmark analysis using a time domain method forthe ses- mic analysis of single piles has been developed under the fol lowing assumptions: ‘+ The soil is an ideal elastic, isotropic material. + The stresses developed between the pile and the soil act, normal to the face of the pile, and no account is taken of possible shear stresses developed between the soil and the ‘ides of the pile. + Each pile element is acted upon by a uniform horizontal stress p, which is constant across the width of the pile. It is also assumed that the soil at the back of the pile near the surface adheres to the pile. ‘The pile is modeled as a beam, which bends in the direction, of loading, and in determining its displacements, use is made Of the differential equation for bending of a thin beam. The soil displacements are calculated based on the Mindlin equa- tion (1936). ‘The analysis is preceded by a seismic free-field analysis, which gives the soil movement at the location of the pile el- ‘ement centers at each time step. By assuming that the site ‘consists of horizontally infinite layers, the soil is modeled as 18 multidegree-of-freedom mass-spring-dashpot system. ‘The pile is discretized into n + 1 elements of equal length 8 except those at the top and the tip, which are of length 8/2, and displacement compatibility is enforced between the pile JOURNAL OF GEOTECHNICAL AND GEOENVIRONMENTAL ENGINEERING / SEPTEMBER 2001 /757 at FIG. 1, Pile Elements in Lateral Seismic Analysis of Single Piles; H Is Soil Layer Depth and u, Is Profile of External Soil Movement Caused by Earthquake Hitting Base of Layer ‘and soil at the center of these elements (Fig. 1), which leads to the following incremental equation (Tabesh 1997): 1" ge BAP a, p+ ga] (au) + Ai [ zl Je ey (A 4 cB UBB (si) ~ (Ai) = as + Ey 4a ~ (ada = SE (a) where (D] = matrix of finite-| Ude20 E/e,=1000 veoa Earthquake FIG. 2. Model Pile Embedded in Two-Layer Soil System 1788 | JOURNAL OF GEOTECHNICAL AND GEOENVIRONMENTAL ENGINEERING / SEPTEMBER 2001 Whittier, Varvb=ti6 Noweast kao — serar © Benth i Head rel: dip.) Newcastle 94, Valvb=t/3 kao —— separ Taft, =12m, d=.3m, Es=120MPa, Capa44.et kao oo —— emp 201 0 @ Time (see) FIG. 3. Comparison of SEPAP Results with Kavvadas and Gazetas (1993) Method TABLE 1. Eanhauake Characteristics aration Baribguake record Peak accsleration g o 1994 Newcastle 0.052 2 1969 Tat 0.046 20 1987 Whitiee Narrows 0.18 20 herein. The first comparison is made with a Winkler formu- lation for the frequency domain dysamic linear analysis of pile foundations developed by Kavvadas and Gazétas (1993). "An cend-bearing pile embedded in a two-layer soil system, as shown in Fig. 2, and subjected to earthquake excitation at its tip is considered. The response of the pile for several ratios of the shear-wave velocities of the upper and lower layers, ‘ViJ¥,, is obtained by SEPAP and is compared with the Kav: vadas and Gazetas (K&G in the figures) formulation. AS the ‘Kavadas and Gazetas method is a frequency domain formu lation, the time domain response has been obtained by Fourier transformation. In SEPAP, the analysis is sensitive to element size that should be small enough to ensure unhindered wave ‘Propagation through the discretized model and accurate eval. tation of pile curvature. For the current comparison with Kav- vvadss and Gazetas method, an element size of 75 em was found to be adequate Figs. 3(2-c) show the profile of moment along the pile at the time step of maximum moment for several soil conditions ‘and the 1987 Whitier and 1994 Newcastle earthquakes (Table 1D. As can be seen, the agreement between the two models is 200d even when the soil medium is strongly nonhomogeneous and is very close indeed for the homogeneous case. The sim ilarity between the two methods holds for different cath: quakes Fig. 3(d) shows the time history of the head response of a 0.3.mdiameter, 12-m-long concrete pile carrying a cap-mass of 48.1 ton and embedded in » homogeneous clay soil with medium stiffness. The agreement is again very good. ‘The methodology presented in this paper has been further ‘compared with the results of a number of centrifuge tests by Finn and Goh! (1987), As an example, the computed and mea~ sured pile-head response time histories in Test No. 12 have been compared in Fig. 4(a) and their frequency content has been depicted in Fig. 4(b). Again the agreement is reasonable. ‘The soil parameters used in the computations are the same as measured by Finn and Gobl (1987). @ “Tine (see) ‘Test 12 -Single pile - Measured and computed pile head acceleration = t » B » i 6 s 0 \ 4 1 0 = oe) sasaatian) —sepap FIG. 4. Comparison of SEPAP Results with Centifige Test Results, MOURNAL OF GEOTECHNICAL AND GEOENVIRONMENTAL ENGINEERING / SEPTEMBER 2001 /759 ‘Several comparisons have been made with the measure- ‘ments carried out by Japan's Shimizu Corp. at the site of the ‘Ohba-Obashi pile foundation in Japan, one of which will be presented later in this paper. ‘The comparisons between SEPAP and both theoretical and ‘experimental results suggest that the SEPAP analyses provide ‘an adequate basis for benchmark solutions to, compare with the approximate analysis. PSEUDOSTATIC APPROACH Abghari and Chai (1995) examined the performance of a ppseudostatic approach for the analysis of soil-pile-superstruc- fue interaction. In their analysis the pile was subjected tothe maximum free-eld soil displacements at each node along its ength. These displacements were obtained from a separate free-field site response analysis. According to the current prac- tice in CALTRANS, in the model, the superstructure was re- placed by @ single-degree-of-freedom system with a natural frequency equal to the fundamental natural frequency of the superstructure. The inertial forces acting on the pile were ob- tained from the product of mass and spectral acceleration. ‘These forces were applied to the pile as static forces. ‘Abghari and Chai (1995) compared the result ofthis static analysis with a dynamic finite-element model and concluded that only when the inertial forces were reduced to 25% for deflection and to 50% for moment and shear did the result of the static analysis approach that of the dynamic analysis. They attributed this to the unlikelihood of the simultaneous occur- rence of the maximum free-feld soil displacement and maxi ‘mum inertial forees. This explanation however isnot sufficient and cannot be generalized. Consider the example of shear: the soil movement always gives a zero shear at the pile head and, ‘when there is a considerable pile-head force, maximum shear ‘occurs at the head and is equal to the applied force. In such a case, the maximum value of the shear is not related to the free-feld motions and its overestimation in the proposed pseu- dostatic methodology cannot be attributed to the unlikelihood of simultaneous occurence of the maximum free-feld move- ‘ment and maximum inertia force. Abghari and Chai used this ‘methodology only in one example and nade no efforts to gen- ceralize the procedure. ‘AS an allernative, the following pseudostatic method is pro- posed here: 1. The superstructure is modeled by a single-degree-of-free- dom system whose natural frequency is equal to the fun- damental frequency of the superstructure. When calcu- lating this fundamental frequency, the effect of pile-head stiffness must be taken into account. This may be ob- tained through published formulas and charts (Gazetas 1991). The simplest way is to reduce the superstructure to a cap-mass. The designer should be careful about such 1 crude approximation; the mere eccentricity of the su- perstructure mass may have a profound effect on the re- sponse, If such an approximation is warranted, the nat- tural period of the pile head may be obtained approximately from (2), in which K, is the lateral head stiffness of the pile 2m [PES 2 2. A free-feld site response analysis is performed to obtain both the time history of the motion at the surface and the maximum displacement of the soil mass along the pile. In this analysis the well-known SHAKE (Schnabel et al. 1972) program, or similar computer codes such as the ERLS program (Tabesh 1997) used herein, may be used. The maximum values of the displacements along the pile obtained in Step 2 are treated as a static soil ‘movement profile, although the displacement at each point may have occurred at different times. 3. The surface motion obtained in Step 2 is used in an or. inary spectral analysis of a single-degree-of-freedom system whose period is equal to the period obtained in Step 1. The spectral acceleration is calculated. 4, The lateral force to be applied to the pile is obtained from multiplication of the spectral acceleration obtained in Step 4 and the cap-mass or the mass of the single- degree-of-freedom model of the superstructure calculated in Step 1. 5. A static analysis, in which the pile is subjected to the ‘simultaneous application of a lateral force at its head ‘equal to the force obtained in Step 5, and the soil move- ‘ment profile formed in Step 3, is performed, and the ‘maximum pile moment and shear are obtained, VERIFICATION OF PROPOSED METHOD ‘To examine the performance of the proposed pseudostatic ‘methodology, a hypothetical 16-m-deep soil mass consisting of two layers was considered (Fig. 2, with fy = 4). The mod- ulus of the bottom layer with a depth of 12 m is 150 MPa and the modulus of the 4-m-deep upper layer ranged between 25 and 400 MPa, The diameter of the pile ranged between 0.3 and 1.5m. In each case the pile head was assumed to be free and the pile tip was considered to be fixed. The 1994 New- castle earthquake was used as the excitation source. Eighty TABLE 2. Maximum Moment (MIN-m) Obtained in SEPAP (Dynamic) and Static Analyses Ey (MPa) SEPARd=03m Sut, d= 0.3m _SEPAR d= 09m Stati, = 0.9m _SEPAP. 2s 0.033 0023 0487 0.682 3.079 3.052 100 0.022 0022 0.667 0.660 3.009 2.998 200 ‘oz oo2t 0625 0616 2870 2.805 300 0.020 0.4020 0596 0.586 2738 2.665 400 0019 0019 os 0.562 2631 2552 TABLE 3._Maximum Moment (MN) Obtained in SEPAP (Dynamic) and Stale Analyses En (ipa) SEPAR.d=03m Ststi,d=03m __SHPAR = 09m Sutic,d=09m _SEPAR d= 15 m 25 00m ‘0072 0384 0342 101s 100 02 021 0334 0.330 002 gas 200 0.020 0.020 0313 0.309 ose 0923 300 19 ors 0.299 0.294 ‘0902 0378 40 ors 0018 0287 0282 0.868 0842 "780 / JOURNAL OF GEOTECHNICAL AND GEOENVIRONMENTAL ENGINEERING / SEPTEMBER 2001 Est=25MPa,de3m ene 002, bed ie: Nyaa o ei dd=.6m, Est=25MPa, Test 12 earthqual © (Capemace (ton) Est=25MPa, Cap=260t | o— = 0 6 Moment (aan) i celeiatetetsstetetere © eg) 7 Eet=25MPR, d=.3m (e) Depth im) ‘d=.6m, Est=25MPa, Test 12 i] . is — i | Es a @ Cap-mass (ton), d=,6m, Est=25MPa, Cap=280t o ‘Depth en) FIG. 5. Performance of Peudosttic Approach: (a) Envelope of Positive and Negative Moment along Pile in Dynamic Analysis and Profile of Moment in Stic Analysis: (b) Envelope of Positive and Negative Shear along Pile in Dynamic Analysis and ProGle of Moment in Static Analysis; (¢) Maximmarn ‘Moment Obtained by Dynamic and Static Methods; (d) Maximum Shear Obtained by Dynamic and Static Methods) Envelope of Poitivé and Negative ‘Moment in Dyna of Shear in Static Analysis, different pile-soil configurations were considered for which the envelopes of the positive and negative moment and shear along the pile were obtained by the SEPAP program. The shear ‘and moment distributions along the pile were also calculated from the proposed pseudostatic analysis with the maximum ‘computed free-field soil movements as input. The cap-mass ‘was assumed to be zero in this exercise. Without any excep- tion, excellent agreement was obtained between the bench- ‘mark SEPAP (dynamic) analysis and the pseutiostatic meth- ‘odology. Tables 2 and 3 compare the maximum moment and shear obtained in the two analyses for a selection of the cases analyzed. ‘The closeness of the results is not restricted to the point of ‘maximum response but occurs along the whole length of the pile. Figs. 5(a and b) show the envelope of the positive and negative moment and shear along the pile for the case of E,, = 25 MPa. These figures also show the profiles of moment and shear distibution along the pile obtained in the static anal- ysis, which match closely the maximum values obtained in the ‘dynamic analysis. Iti interesting that the static profile in some parts closely matches the dynamic positive envelope and, in ‘ther parts, the dynamic negative envelope. Analysis and Profile of Momeat in Static Analysis; (F) Envelope of Positive and Neg ‘Shear in Dynamic Analysis and Profile ‘What these analyses suggest is that, regardless of the soil ‘nonhomogeneity, the static methodology gives very good re- sults for the maximum values of the moment and shear along the pile. This has far-reaching consequences and means that, when the response of the pile is dictated by the free-field soil ‘movements, the internal response of the pile can easily be estimated by a very simple static analysis. It is emphasized that, whenever the pile cap-mass is small or when the natural period of the pile head is far from the natural period of the soil layer, the response is largely dictated by the free-field movernents. ‘When the effects of cap-mass were taken into account, it was observed that, in many cases, the agreement between the proposed pseudostatic method and dynamic analysis was very ‘close but that, in some others, the pseudostatic approach over- ‘estimated the maximum moment and shear. One of these anal- yses is presented herein. Consider the soil profile of Fig. 2 ‘m, fy = 12 m, Ey, = 25 MPa, and Eq = 150 MPa. ‘The pile diameter is 0.6 m and the cap-mass is changed from 0 to 380 ton, in 20-ton increments. The input is the Test No. 12 earthquake (Finn and Gohl 1987). Figs, 5(c and d) compare ‘the maximum moment and shear obtained in dynamic and the JOURNAL OF GEOTECHNICAL AND GEOENVIRONMENTAL ENGINEERING / SEPTEMBER 2001 /761 proposed pseudostatic analyses for a wide range of values of the cap-mass. These figures show that the static methodology ‘considerably overestimates the maximum values of moment and shear when the cap-mass goes beyond about 160 ton, However the trends in the results of the two methods are sim- ilar, with both methods having a peak at the same cap-mass value. . Figs. 5(e and f) show the envelope of the positive and neg- ative moment and shear along the pile in the dynamic analysis and the profile of moment and shear distributions obtained in the static analysis for cap-mass values that have yielded the maximum difference between the two analyses. The agreement appears to be acceptable for practical purposes. It needs to be ‘emphasized that in the proposed method it is assumed that the ‘maximum free-field ground movement is in phase with the ‘maximum lateral inertial force applied to the pile head by the ‘superstructure. This explains the overestimation of the pseu- dostatic approach in some cases compared to the results of the dynamic SEPAP analysis. CASE STUDY In this section the proposed pseudostatic methodology is used to estimate the maximum moment developed in the ‘Ohba-Ohashi bridge in Japan. This bridge is located in the city of Fujisawa, Kanagawa prefecture, near Tokyo. The seismic ‘observations at this bridge and one of its pile foundations were ‘conducted between 1981 and 1985 by the Shimizu Corp. Dur- ing this period, 14 earthquakes hit the region. Among them, the 12th earthquake induced the largest peak horizontal surface acceleration, which was 0.11g. ‘The bridge is supported by 11 piers and is 4848 m long and 10.75 m wide. The girder is continuous from pier 5 to ier 8. Piers 5, 7, and 8 are equipped with movable bearings, and pier 6 is of the fixed-shoe type. Figs. 6(a and b) show the details of the bridge and the pile foundation of pier 6, where the strain meters were installed. The soil profile shown in Fig. {6(@) was obtained from a borehole near pier 6. It shows that the top soil layers are extremely soft, with N values from the standard penetration test being almost zero. The shear-wave velocity of the top layers was between 40 and 100 m/s, The tips of the piles were embedded in a stiff layer of clay with a shear-wave velocity of 400 mis. The length of the vertical piles ‘was 22 m, 2 m of which were in the stiff clay. ‘The piles were steel pipes with a diameter of 0.6 m, the thickness of the vertical piles was 9 mm, and the wall thick- ness of the battered piles was 12 mm. A total of 32 wall strain ‘meters were installed on one vertical pile and one battered pile, at four different elevations (sal ~sa4 and sbl—sb4 in Fig. 6(@)] A total of 11 units of accelerometers of the servo type were installed, with one unit (GS1) at the ground surface, four units (GB1-GB4) at the bearing substratum, three units (Bs1—Bs3) at the footings, two units (BRI and BR3) at the piers, and one ‘unit (BR2) at the girder. The horizontal axes of each acceler- ‘ometer coincided with the bridge axis (HII) and with the di- rection perpendicular to it (H2). ‘The river ran between piers 6 and 7 and the locations of GBI, GB2, and GS1 were 70 m offset from the bridge axis in the H2C+) direction; GB3 was 90 m offset in the H2(—) direction ‘The data obtained from the Ohba-Ohashi observations were first used by Tazoh et al. (1988) in their analysis of seismic response of pile groups. They employed the formulation do- ‘veloped by Tajimi (1969) for the response analysis of single piles embedded in a uniform stratum. As the Tajimi method- ‘ology is only for the case of an end-bearing pile, they first ‘extended this formulation to account for a fixed tip pile, then approximated the top soft layers in the soil profile by a ho- ‘mogeneous layer with an average shear-wave velocity. They assumed that a 2-m embedment in the stiff clay bearing stra- tum is enough to create a fixed tip for the pile. They also developed an approximate methodology to account for the ‘group effects but, because of the complexities of the formu- ation, they used only static interaction factors. Fan (1992) used a number of methods in his analysis of the Ohba-Ohashi data, ‘These methods include an approximate Winkler meth- odology developed by Gazetas and his coworkers (Winkler method) and a rigorous method developed by Kaynia and Kau- sel (1982). Both Tazoh et al. and Fan used primarily the 12th recorded earthquake in their analysis, In this paper the eighth earthquake has been considered and the pseudostatic methodology has been used to estimate the ‘maximum moment developed in the vertical instrumented pile. In this analysis group effects were ignored. In the Ohba- Ohashi bridge measurements, the free-field displacements along the length of the pile were not measured and only the bbase and surface motions were monitored [Fig. 6(c)]. A free field analysis was therefore required, for which the ERLS pro- ‘gram was employed and the motion monitored at GBI was used as the input motion. The soil was modeled as a system of seven horizontal layers, as shown in Fig. 7. As can be seen in this figure, the mass of the superstructure was concentrated. in two points. The maximum value of the free-field response ‘was obtained at 48 points corresponding to the centers of the 48 pile clements. The spectral acceleration corresponding to the pile natural period, and based on the surface motion, was calculated to be 0.092 m/s?. The mass M, was defined as the ‘sum of the weights of the continuous girder from piers P. to P, and half the weight of pier P,, The value of My was defined by adding half the weight of pier P, to the weight of the foot- ing. The values of M, and M, were calculated by Tazoh etal (21988) to be equal to 223.47 and 81.63 ton, The values of m, and mo, which are the portions of mass allocated to each pile, were calculated by assuming that all piles carried equal loads. ‘The pier was very stiff and could be considered to be sol In that case the model in Fig. 7(a) could be reduced to that in Fig. 7(b). The eccentricity was calculated to be 16.3 m, ‘With these assumptions the pseudostatic approach was used to obtain the response of the pile to the combination of the following disturbances: + A soil movement profile formed from the maximum free- field soil movement obtained at each pile element + Ahead force equal 10 F = (iy + Mm )Agec in Which Aya, is the spectral acceleration obtained for'a period equal 0 that ofthe pile and based on the surface motion + Armoment equal to M = Fe, in which eis the eccentricity The profile of the moment along the pile obtained from the pseudostatic method, along with the maximum moments mea- sured at four locations along the pile, are shown in Fig. 8. In this figure the envelope of the positive and negative moment along the pile obtained from an independent benchmark SEPAP analysis is also shown, ‘AS can be seen in this figure, the proposed static method cology has given very close estimates of the measured values. ‘This is despite the fact that the Ohba-Ohashi bridge is very complicated and has been highly simplified. Similar results ‘were obtained for several other measured earthquakes (Tabesh 1997), CONCLUSIONS ‘The results of the analyses outlined in previous sections ‘suggest that the pseudostatic methodology explained in this ppaper has promise. It can be used to obtain a reasonable es- imate of the maximum moment and shear that is likely to be 1762 JOURNAL OF GEOTECHNICAL AND GEOENVIRONMENTAL ENGINEERING / SEPTEMBER 2001 Organic En Battered Pile Vertical Ple Vertical pile with strain gauges Battered pile with strain gauges Earthquake No. 8 - H1 20 ost 15 —sst 10 6 ‘Ace. (gal) © Time (sec.) FIG. 6. (a) Obba-Ohashi Bridge Pile Foundation; (b) Arrangement of Piles in 8 X 8 Pile Group of Ohba-Ohashi Pier 6 Foundation; (c) Base and ‘Surface Acceleration-Time Histories tor Earthquake 8 in HI JOURNAL OF GEOTECHNICAL AND GEOENVIRONMENTAL ENGINEERING / SEPTEMBER 2001 /763 « sexmoem Acc mee ° . mH ON 43 | 45 o 100 er mh sas | 6s 40 05 8 aan ons om ons FIG. 8. Comparison of Pseudos developed in a pile when itis subjected to earthquake exci- tation This static analysis is likely to give very good results in ‘many practical cases, but it may overestimate the maximum ‘moment and shear in certain other cases. ‘At the event of an earthquake, for a pile with a cap-mass, both the free-ield soil movement and the inertia force due to the cap-mass determine the response of the pile. If the cap- mass is not large enough to bring the natural period of the Depth tt fe Method with Measured Moment along Vertical Pile at Ohba-Ohashi Pile Foundstion pile-cap-soil system within the range of dominant periods of the surface motion, it will often have a negligible effect on the response and the response will be determined by the free-feld ‘movement. In such a case the pseudostatic approach is likely to give an accurate result if, instead of maximum spectral ac: celeration in Step 4, the maximum surface acceleration is used If the cap-mass is such that the period of the pile-cap-soil system is within the range of dominant periods of the surface motion, the cap-mass is likely to have a dominant effect on ‘764 / JOURNAL OF GEOTECHNICAL AND GEOENVIRONMENTAL ENGINEERING / SEPTEMBER 2001 the response of the pile. In such a case, the effect of the eap- mass will be grossly underestimated if, instead of maximum spectral acceleration, the maximum surface acceleration is used in Step 4, The use of the maximum spectral acceleration in this step will give a conservative result. ‘The static analysis provides viable estimates for'the moment and shear that would be obtained in an elastic dynamic anal- ysis. Soil yielding, which may have significant effects on the Pile moment and shear, is not taken into account. Such effects are discussed by Tabesh (1997). ‘The maximum moment and shear obtained in the static anal ysis is only for a lateral motion in one direction. The stressing, (of the pile by motions in other directions, and the axial stresses ‘caused in piles in a pile group due to the rocking of the group, hhave not been taken into account, but could be incorporated in an extension of the present analysis. ACKNOWLEDGMENTS ‘This work has been partly supported by a numberof grants from the ‘Australian Research Council and scholarship from Iran's Minty of Clore and Higher Education. Acknowledgment is made to Japan's Shi ‘misu Corp, which supplied the mescured data atthe Ohbe-Ohachi pile foundation, and Prof. W.D. L. Fin, who made the results of a number of centrifuge tests available, REFERENCES ‘Abghari, A., and Chai, J. (1995). “Modeling of soi pile superstructure Tnweraction for bridge foundations.” Performance of deep foundations tinder seismic loading, J.P. Turner, ed, ASCE, New York, 45-59. Berger, E, Lysmer, J, and Seed, H. B. (1975). "ALUSH, computer [rogram for seismic response of aisyrnmetic sllstracte systems." Fep. No. 75-31, Envir. Engrg. Res. Council, University of California, Berkeley, Calif Fan, K. (1992). “Seismic response of pile foundations evaluate through ‘case histories.” PAD thesis, State University of New York, Buffalo, Fina, W. D. L. and Gohl, W. B. (1987). "Centifuge model studies of piles under simulated earthquake Intetal loading.” Dynamic response 2f pile foundations: Experiment observation and analysis, T. Nogami. fed, ASCE, New York, 21-38. Gazsis, G. (1991). “Foundation vibeations." Foundation engineering hhandbook, 2nd 1a., Y. Fang, ed, Van Nostrand Reinbold, New York, 553-593, ‘Gazets, G., and Dobry, R. (1984), “Horizontal response of ples in ered soil." J. Georech. Engrg, ASCE, L10(1), 20-0. Kawadas, M., and Gazetas, G. (1993). “Kinemati seismic response and bending of fee-head piles in layered sll.” Géorechnlque, London, 430), 207-272. Kaynia, A. M. (1988), “Dynamic interaction of single piles under lateral ‘and seismic loads." Eueghlal J. Engr. 6, 5-26 (in Persian) Kaynia, A.M. (1993). “Chapter 7." Boundary element techniques in ‘Zeomechanics, G. D. Manolis and T. G. Davies, ds.,copublished by ‘Computational Mechanics Publications, Southhampton, U.K. and El sevier Science, New York Kaynia, A.-M, and Kausel, E, (1982). “Dynamic behaviour of pile ‘groups. Proc, 2nd Int. Conf on Numer. Methods in Offihore Piling, 509-332. Mamoon, S. M. (1990). “Dynamic and seismic behavior of deep fous: dations.” PhD dissertation, State University of New York at Bufal, Buffalo. Markis, N, and Gazetas, G, (1992), “Dynamic pile sol pile interaction. Par I: Lateral and seismic response." Earthquake Engrg. and Struct Dyn. 21, 145-162. ‘Mindlin,R. D. (1936), “Force a point n the interior ofa semi infinite solid.” Phys. (5), 195-202. Mizuno, H. (1987). “Pile damage during earthquake in Japan (1923 1983)." Dynamic response of pile foundations: Experiment, observa tion and analysis, T. Noga ed, ASCE, New York, 33-78 Minino, H.liba, M., and Hirade, T. (1996). "Pie damage during 1995 Hyougoken-Nenbu earthquake in Japan." Proc, {Ith World Conf. on Earthquate Engrg, Paper No. 977, International Asociation of Earth ‘quake Engineering Novak, M. (1974). "Dynamic stiffvess and damping of piles.” Can. Geo- tech’ J, Otawa, 1. 574-598, Novak, M, Sheta, M.,El-Hifaawy, L., BLMarsafaw, H., and Ramadan, (©. (199%), “Dynad user manual” Rep. No. GEOP 93.01, University fof Wester Ontario, London, Canada. Poulos, H. G. (1992). Program ERCAP, user manual, Coffey Pariers Theratonal Py Lid. Sydney. Austalia Poulos. H. G., and Davis, E. H. (1980). Pile foundation analysis and design, Wiley, New York Sehnabel,B. Lysmer, J, and Seed, B, H. (1972). “SHAKE, «computer program for earthquake response analysis of horizontally layered ste.” ‘Rep. No. EERC 72-12, Earthquake Engrg. Res. CX, National Infor ‘mation System for Earthquake Engineering ‘Sen, R, Davies, T.G., and Banerjee, P.K. (1985), “Dynamic analysis of piles and pile groupe embedded in homogeneous soll.” Earthquake Engrg. and Siuct. Dyn, 13, 53-65. “Tabeshs A. (1997), “Lateral seismic analysis of piles." PAD thesis, Dept. Civ. Engrg. University of Sydney. Sydney, Australia, ‘Tabesh, A. and Poulos, H. G. (1997). “A review ofthe dynamic analysis ‘of plie foundations. Res. Rep. No. R757, Ct for Geotech, Res. U versity of Sydney. Sydney, Australia ‘Tabesh, A. and Poulos, HG. (1999). “Kinematic versus static interaction ‘Of pile and soil." Proc, Sik Australia New Zealand Conf. on Geome: ‘hans, Vo. Is Institute of Engineers, Australia, 45450. “Talim, H. (1969), “Dynamic analysis of a structure embedded in an elastic suatum." Proc, 4th World Conf. Earthquake Engineering, VO 3, International Assocation of Earthquake Enginoering, 53-69. ‘Tazoh, T, Shimizu, K, and Wakahars, T. (1988). “Selamie observations ‘and analysis of grouped piles.” Shimizu Tech, Res. Bull, 1, 17-32. NOTATIONS The following symbols are used in this paper: € = damping matrix; = damping coefficient of Winkler model; matrix of finite-difference coefficients; pile diameter; pile rigidity: soil modulus matrix of soil displacement-influence factors; pile-head stiffness: Spring coefficient of Winkler model; pile length; ‘mass matrix; ‘mass per unt length of pile; shear wave; pile displacement; vector of pile displacement; vector of external soil movement; free field soil movement, and shear velocity, JOURNAL OF GEOTECHNICAL AND GEOENVIRONMENTAL ENGINEERING / SEPTEMBER 2001 /765,

Вам также может понравиться