Systematic Selection of Ground Motion Intensity Measures Satisfying the Probabilistic Seismic Demand Analysis

In performance-based seismic design theory, the accurate selection of ground motion intensity parameters is crucial for the probabilistic assessment of structural seismic performance. To achieve efficient, sufficient, applicable, and complete prediction of structural seismic demands, this study systematically evaluates the comprehensive performance of 35 commonly used ground motion intensity measures (IMs). The research begins by analyzing 178 real records from the Pacific Earthquake Engineering Research Center (PEER) database, employing the Spearman rank correlation to reveal intrinsic relationships among the parameters. Subsequently, a ground motion database classified according to four site types based on Chinese seismic design codes is established. Combined with seven single-degree-of-freedom (SDOF) structural models of different periods, the performance of the IMs is comprehensively evaluated from four dimensions: correlation, efficiency, applicability, and completeness. Finally, by comparing the evaluation results under both bilinear and Clough stiffness-degrading bilinear hysteretic models, the robustness of the parameter selection is verified. The results demonstrate that: acceleration-related parameters are most suitable for short-period structures, velocity-related parameters for medium-period structures, and displacement-related parameters for long-period structures. Parameters Svavg and Sdavg exhibit consistent performance across all period ranges and under different site conditions, and the evaluation results remain consistent across different hysteretic models. This study provides a systematic basis for the rational selection of intensity parameters in probabilistic seismic demand analysis, significantly enhancing the reliability and precision of seismic performance assessment.