地震时饱和砂土的失效概率

用Excel 2 0 0 0中的规划求解软件对大量的液化和未液化数据做了可靠度计算。由可靠性分析所得的液化概率 ,本文将其定义为失效概率Pf。以这些分析结果为基础讨论了加速度变量的分布和它的变异系数对失效概率的影响。对于给定的概率分布 ,失效概率在很大程度上与它的变异系数有关 ;但是 ,似乎与其分布关系不大 (随机变量的变异系数保持不变 )。同样 ,也以这些分析结果为基础 ,讨论了失效概率函数Pf=f(Fs)。Fs 为安全系数。本文着重指出 ,基于饱和砂土极限状态函数的可靠度分析 ,如果对随机变量的不确定性做了合理的考虑 ,则由可靠度分析所得的失效概率函数Pf=f(Fs)和Juang等人用Bayesian定理所得的液化概率函数PL =f(Fs)几乎是一样的。液化概率函数可为场地液化势的合理评估和工程风险设计决策提供依据。

Probability of Failure of Saturated Sandy Soil During Earthquake

In the paper, the reliability analysis is conducted for a large database about liquefied and unliquefied cases using optimization tool of a spreadsheet (Excel 2000). The probability of liquefaction P f from the reliability analyses is defined as the probability of failure. Based on the analytical results, the influence of coefficient of variation of ground acceleration and its probability distribution on probability of failure is discussed. It appears that the coefficient of variation of ground acceleration is a great influence on the probability for a given probability distribution, but the probability of failure obtained through the reliability analyses for various probability distribution of ground acceleration is almost not different, if their coefficients of variation are kept unchanged. Similarly, based on the analytical results the probability function (P f =f(F s))is also discussed. F s is used as the safety measure (factor of safety). The authors put emphatically out that the reliability analyses based on the limit state function of saturated sands during an earthquake is conducted if the uncertainty of random variables is reasonably considered, then yielded the mapping function (P f =f(F s)) could compares very well with that obtained through the Bayesian approach which was developed by Juang et al. The mapping function can provide a basis for evaluating reasonably liquefaction potential at sites and for making risk-based design decisions.