基金项目:国家重点研发计划专题(2018YFC1503606G06)资助.
备注
以某Ⅷ度设防区基岩场地地震危险性计算为基础,拟合不同随机相位的人造地震动时程作为输入,采用一维等效线性化方法,计算了Ⅱ类、Ⅲ类和Ⅳ类典型场地的土层地表地震反应。结果表明:①在不同地震动强度、不同随机相位基岩时程输入条件下,对不同类型场地土层地震反应计算得到的地表加速度峰值和反应谱值相对极差差别较大,地震动相位特征对土层地震反应的影响不可忽略; 在反应谱特征周期2.0 s内,地表峰值和反应谱值变异系数随输入地震动强度的增大有增大趋势; ②采用统计学方法计算给出了不同场地类别的基岩输入随机相位样本时程的必要数量,不同场地类别不同地震强度输入条件下所需要的最少样本量不同。在输入地震动强度不大(PGA<0.20 g)且满足反应谱变异系数在均值加1倍标准差范围内时,不同类别场地至少需要15组不同随机相位的基岩时程,基本能满足均值统计要求; 在输入地震动强度较大(PGA≥0.20 g),满足反应谱变异系数在均值加1倍标准差范围内时,至少需要30组不同随机相位的基岩时程,才能满足均值统计要求。
On the basis of the seismic hazard calculation of some bedrock site in Intensity 8 earthquake fortification zone,and using one-dimensional equivalent linearization method,we input the fitted time history of artificial ground motion with different random phases,and calculated the ground response spectrum of Site Class Ⅱ,Ⅲ and Ⅳ.The following conclusions are reached:(1)the earthquake responses of different types of sites are carried out by using bedrock time histories with different seismic intensities and random phases.The difference of relative extrema of the peak ground acceleration values and the response spectrum values calculated is significant.The influence of phase characteristics of the time history of ground motion on the soil earthquake response is innegligible.Within 2.0 s of the characteristic period of the response spectrum,the variation of the ground motion values and the response spectrum values increases with the increase of intensity of input ground motion.(2)the necessary quantity of the time histories with random phases on different sites is calculated by the statistical method.The minimum sample size depends on the following conditions:when the input ground motion is small(PGA<0.20 g),and the variation coefficient of response spectrum is within the range of mean value plus one-fold standard deviation,at least 15 pairs of time histories are needed to meet the average statistics.When the input ground motion is large(PGA≥0.20 g),and the variation coefficient of response spectrum is within the range of mean value plus one-fold standard deviation,30 pairs of time histories are needed to meet the average statistics.