WANG Wei-yi, JI Kun, WEN Rui-zhi, REN Ye-fei. SITE-SPECIFIC HAZARD CONSISTENT EXACT CONDITIONAL TARGET SPECTRUM AND SELECTION OF GROUND MOTION RECORDINGS FOR CHINA[J]. Engineering Mechanics, 2024, 41(2): 71-82. DOI: 10.6052/j.issn.1000-4750.2022.03.0250
Citation: WANG Wei-yi, JI Kun, WEN Rui-zhi, REN Ye-fei. SITE-SPECIFIC HAZARD CONSISTENT EXACT CONDITIONAL TARGET SPECTRUM AND SELECTION OF GROUND MOTION RECORDINGS FOR CHINA[J]. Engineering Mechanics, 2024, 41(2): 71-82. DOI: 10.6052/j.issn.1000-4750.2022.03.0250

SITE-SPECIFIC HAZARD CONSISTENT EXACT CONDITIONAL TARGET SPECTRUM AND SELECTION OF GROUND MOTION RECORDINGS FOR CHINA

  • The seismic geological structure and the background seismicity varied between different cities or regions, even the corresponding design seismic intensity are identical. To make the hazard level consistent with the target site, this study proposed to construct the conditional spectrum (CS) based on the output of the fifth-generation seismic zonation map of China. The potential seismic sources and GMPE used in the zonation map were applied to carry out the PSHA and the disaggregation of earthquake scenarios. The 3D (magnitude-longitudes-latitude) disaggregation results were applied to construct the "exact" CS considering the contribution of different strikes of the seismic source area. A city in north China was taken as the example to illustrate the CS computation results. The comparison results indicated that the median value of an exact CS lies within the approximate CS using long axis and short axis, the conditional variance is higher than that of the approximate CS. The exact CS could reflect the spatial distribution of the potential seismic area sources and the difference in seismicity, and more suitable for China GMPE which has long axis and short axis equations. The correlation derived using China ground motion dataset would result in higher CS compared with that using PEER dataset. In this study, the covariance matrix of multivariate Gaussian distribution is adopted to approximately replace the covariance matrix of Gaussian mixture distribution of an exact CS. With greedy optimization algorithm, highly efficient matching of mean, standard deviation, and Sa distribution was realized. The exact CS constructed in this study and corresponding ground motion recordings selection could provide a basis for following performance-based analysis of structures.
  • loading

Catalog

    /

    DownLoad:  Full-Size Img  PowerPoint
    Return
    Return