Fractile-Based Mean Spectral Matching with Dispersion Control

Performing nonlinear response history analyses for structural design and site ground motion response analyses requires selecting a suite of representative input ground motions as input to the analyses. Each selected ground motion can be amplitude-scaled such that the mean response spectrum (or resultant spectrum) of the entire suite matches approximately to a target spectrum. Spectral matching, such as tight spectral matching and mean spectral matching, can also be used to further adjust the selected suite of ground motions to produce a closer match of the suite mean to the target spectrum. This study proposes a new approach to mean spectral matching practice that defines the target variation (dispersion) within the ground motion suite to the probabilistic seismic hazard analysis framework. By using a fractile analysis, the dispersion of ground motions can be quantified to produce a more realistic variability that is consistent with the variability from the probabilistic seismic hazard analysis. A case study and scaling algorithm is presented. Two spectral matching tools, namely RSPMatch and REQPY, are used to modify the seed ground motions, and the results show the proposed method can be a potential expansion to the mean spectral matching method.