Various Near-Fault Ground Motions for Seismic Analysis of High-Speed Railway Track Bridges

Abstract

In order to investigate the applicability of the ground motion generated by different methods in the seismic study of high-speed railways, this paper presents a multiple comparative analysis of the ground motion produced by the ground motion prediction equation method (GMPE) and the physically-based seismic simulation method (PBS), taking the Chi-Chi earthquake as an example. First, the ground motions of typical stations were obtained by GMPE and PBS methods and compared with the original seismic records. It is found that the PGAs of the simulated ground motions are closer to the original seismic records, and the median value of the PGAs of the ground motions selected by the GMPE method is slightly larger. There are differences between the two ground motion acceleration response spectra and the original seismic record, but there are also partially valid intervals. After that, the PBS ground motion (PBS-GM), GMPE ground motion (GMPE-GM), and original seismic record (Oril-SR) were inputted into the High-Speed Railway Track Bridge System (HSRTBS) for the dynamic history analysis, respectively. It was found that the PBS-GM overestimated the dynamic response of the HSRTBS at locations closer to the epicenter and underestimated it at locations farther away, but the error was small. In contrast, the GMPE-GM accurately estimates the dynamic response of the HSRTBS at locations moderately close to the epicenter and overestimates it at locations closer to the epicenter but with flat topography. In summary, the PBS-GM has a reliability that is not inferior to that of the GMPE-GM.