本研究之土質參數折減係數模式，係以一維波動方程模擬地盤液化狀態下之樁基礎動力反應，依據現場土層鑽探資料與鄰近測站之地震紀錄，由土壤液化潛能評估法配合日本相關規範求得液化土層之土質參數折減係數，以此係數對土壤強度模數進行折減，並使用集中質塊法求取自由場之地盤液化反應，再以此為前置解作用於樁基礎之波動方程求解，樁身剛度以簡化之Bouc-Wen模式模擬其非線性行為，以瞭解液化地盤內基樁之變形行為與破壞機制。本研究另與ABAQUS有限元素法軟體及日本地震時之基樁破壞案例驗證本模式之合理性。 研究結果顯示：（1）液化土層中，土質參數折減係數值變化處易發生基樁變形之轉折點，而衍生較大之內力致使基樁發生破壞；（2）觀察發現，基樁之最大位移易發生於樁頂，其最大彎矩常發生於液化層下方交界處而造成破壞，且損害形式可能以不同型態方式發生；（3）由參數研究可知，SPT-N值液化潛能評估方法中，以日本道路協會簡易經驗法（1996）評估結果最為保守，NCEER修正之Seed簡易經驗法（1997）次之，Tokimatsu and Yoshimi簡易經驗法（1983）評估之地盤液化潛能最低。分析時，若各液化潛能評估法所得之液化層厚度相近，則土質參數折減係數對樁頂最大位移之影響較小，對內力影響較大；（4）影響基樁行為之因子中，以地盤種類影響最大，最大地表加速度次之，地下水位及細粒料含量影響最小；另外，基樁之樁身土壤彈簧及阻尼對樁基礎變形影響較小，對內力變化影響較大；（5）由實際案例研究發現，NCEER修正之Seed液化潛能評估法較能符合實際案例情形，故建議可以NCEER法作為主要之液化潛能評估方法；（6）本研究建立之土質參數折減係數模式，可涵蓋土層液化時之弱化現象，清楚掌握地盤液化反應與樁基礎之動力行為，有助於提供樁基礎耐震分析使用。 This study uses the soil parameter reduction coefficients to model the seismic pile responses affected by liquefied soils. One dimensional wave equation analyses were conducted for the solutions. Based on the bore-hole data and nearby seismic record, the soil parameter reduction coefficients of liquefied layers can be obtained through a pre-analysis for liquefaction potential of the site. The reduction coefficients were used to reduce the soil modulus for liquefied soils, and the lumped mass analysis is performed to obtain the free-field response of the site. The ground deformations are superimposed onto the pile elements for discrete wave equation analysis, and the non-linear pile responses can be simulated through the modified Bouc-Wen model. Thus the deformations and failure mechanism of the pile are able to investigate. This analysis is also validated by linear ABAQUS analysis and the case studies of pile failure due to earthquake in Japan. The results of this study are concluded as follows: (1) In the liquefied layers, the sharp changes of soil parameter reduction coefficients may cause greater internal forces to occur and the failure of pile. (2) Largest pile displacements would occur at the pile head. A very large bending moment of the pile that exceeds the ultimate moment of the pile would occur at the interface between the liquefied layer and the underneath non-liquefiable layer. Damage pattern might occur in many ways. (3) For liquefaction potential analyses based on SPT-N values, it is found that the NJRA (1996) method is the most conservative one, where the T&Y (1983) method is the optimistic one. The NCEER (1997) method ends up in the median; In the analyses, if the thickness of liquefied layers are about the same, the reduction coefficient will affect significantly the internal forces of the pile, and relatively insignificant to the pile displacements. (4) For the influence factors on the pile behavior, the ground stiffness has the most influence. PGA is the 2nd significant one. Ground water table and fine content of soil have smaller effects. Furthermore, soil spring and damping have minor influences on pile’s deformations, but they are sensitive to the internal pile forces. (5) The NCEER method is found to provide comparative results to the field observations of the case studies, therefore it is suggested to use NCEER in the corresponding analysis. (6) The reduction coefficient model is rather simple to simulate degradation of the soil modulus, it is conjunct with the liquefaction potential analysis of the site and coincident with common geotechnical engineering practice. This model would be very helpful to the seismic performance analysis of the pile foundations.
