中文摘要：

冲刷造成桩周土体的剥蚀将会削弱土体对桩基的侧向支撑能力,冲刷效应会对桥梁桩基的地震易损性产生影响,因此有必要对冲刷和地震共同作用下桥梁桩基的易损性进行研究。利用SAP2000软件建立三维桥梁有限元模型,通过非线性时程分析得到桥梁桩基地震响应峰值。采用概率性地震需求分析方法,建立不同冲刷深度下桥梁桩基地震易损性模型,在地震易损性函数假设为对数正态分布函数的基础上,通过回归分析得到概率模型中的参数,进而得到不同冲刷深度下桥梁桩基在不同破坏状态所对应的地震易损性曲线,并分析冲刷深度对桩基破坏概率的影响。研究结果表明：随着冲刷深度的增加,桥梁桩基在地震作用下的破坏概率显著增加。

英文摘要：

The erosion of soil by scour around a pile weakens the lateral support in bridge founda- tions and this phenomenon has an effect on the seismic vulnerability of bridge pile foundations. Therefore, it is important to study the seismic fragility of bridge piles in the presence of scour. We used SAP2000 software to establish a three-dimensional finite element bridge model in the presence of scour. Damage occurs more readily at bridge piers and pile foundations and these com- ponents may become plastic during earthquakes. As such, we simulated these components using an elasto-plastic connection unit. We simulated the mechanical behavior of the soil using nonlinear p-y springs and modeled the loss of soil by removing the p-y springs above the scour surface. We selected 80 acceleration time histories of seismic waves that conformed to site II, and then trans- lated the acceleration time history of each seismic wave into the displacement time history. We then applied the displacement time history to the p-y springs and carried out a nonlinear time history analysis on the bridge model. We obtained the seismic response of the bridge and recordedthe maximum response of the bridge pile under each seismic wave. Using a probabilistic seismic demand analysis approach, we developed an analytical seismic vulnerability model of the bridge pile at different scour depths. We used curvature ductility as a damage index, and according to Hwang＇s suggestion, defined different damage states of the pile foundation. We obtained the boundaries of damage states by applying moment-curvature analysis to the pile cross section, thus obtaining quantitative data of the different damage states of the pile foundation. On this basis, we assumed that lognormal distribution functions were suitable for developing fragility functions, and we used regression analysis to estimate the parameters in the seismic fragility functions. We established the seismic fragility curves of the pile under different scour depths, and then discussed the influence of sc