基于震害调查数据的液化侧向变形预测模型框架

Prediction Model Framework for Assessing Liquefaction induced Lateral Deformation Based on Seismic Damage Investigation Data

摘要:

基于地震统计数据，前人就统计回归分析影响因子和变形值的关系提出若干评估液化侧向变形的方法，能计算得到给定土体条件和地震作用下的侧向变形，但这些方法为确定性描述，无法根据震害历史预测变形，较难为城市抗震合理布局提供切合实际的依据。本文采用地面峰值加速度PGV和震级MW联合分布概率的计算方法，以蒙特卡洛方法模拟已有的地震液化侧向变形回归公式中场地参数的随机性，建立对区域性地震液化侧向变形超过指定阈值的年平均率的预测模型框架，并应用于实例，以期对区域性震害进行液化侧向变形评估及其非确定性描述研究提供思路。

Abstract:

This paper presents a framework for assessing the probability that a specified liquefaction-induced lateral deformation will be exceeded based on seismic damage investigation data.This framework overcomes the defect in traditional empirical formulas and regression models in how the terms of uncertainty are described.It deals not only with the effect of liquefaction but also with the probabilistic characterization of all possible region ground motions at a given site with a known earthquake history （in terms of the joint distribution of PGA and MW）.In the case analysis，we obtain a liquefaction-induced lateral deformation hazard curve by applying the framework to the calculation in order to demonstrate the annual rate of deformation exceedance at a given site for a given exposure time.Thus，this framework has the potential to be a very useful tool in the field of geotechnical earthquake engineering.It can also facilitate further research efforts to describe the uncertainties associated with liquefaction induced lateral deformation.While the proposed framework is simple and effective，more work is necessary，especially on other factors that may be influencing the deformation，before it may be considered to be complete.