土层地震反应分析的地面加速度峰值计算研究

针对不同的工程场地采用20组相互独立的基岩加速度时程作为输入地震波,对广东省307个项目的地震安全性评价中的662个地震钻孔剖面进行土层地震反应分析,用相对均方误差来定量评估所得到的地面加速度峰值离散值,发现地面加速度峰值的计算结果具有很大的离散性,最大相对均方误差为19%,平均相对均方误差为7.53%;地震基本烈度高的地区比地震基本烈度低的地区的相对均方误差大,场地类别差的比场地类别好的相对均方误差大;目前较为普遍地用1条或3条基岩加速度时程作为地震动输入是明显不够的,应适当地增加,对于高烈度地区和场地类别较差的场地更是如此,才能得到较为合理的地面加速度峰值。

Study on Calculation of Seismic Peak Ground Acceleration of Soil Layer in Seismic Response Analysis

Seismic peak ground acceleration is an important parameter of earthquake intensity. It is also one of the main results of seismic safety evaluation of engineering sites. To get reasonable seismic peak ground acceleration is of positive significance. While working on seismic safety evaluation of engineering sites in the past decade, the author found that some more serious scatter characters are showed in the seismic peak ground analysis. To evaluate the influence, this paper calculates the seismic peak ground acceleration of 662 real bore sections from 307 project sites in Guangdong province, using 20 groups of mutual independent bedrock acceleration histories as the input seismic waves for different sites. The relative rootmean-square errors are used to evaluate the scatter characters of seismic peak ground acceleration. The maximum relative root-mean-square errors of seismic peak ground acceleration are 19%, while the mean relative root-mean-square errors are 7. 53%. The larger relative root-mean-square errors are in regions with higher magnitude than those with lower and the larger relative root-mean-square errors are in the worse site sort than in the better site sort region. In order to get the reasonable seismic peak ground acceleration using 1 or 3 acceleration histories as the input seismic waves is obviously not enough. The number of the input seismic waves shall be increased especially in regions of magnitude and with worse site sorts.