Influence of Mainshock-Aftershock Sequence Selection Techniques in Quantifying Seismic Response

Abstract

Past studies have observed increased vulnerability of the structures when subjected to mainshock-aftershock (MS-AS) ground motion sequences instead of mainshocks (MS) alone. The lack of availability of as-recorded real MS-AS ground motion sequences for the seismic performance assessment of the structure has led to the use of artificially generated sequences. This study aims to quantify the relationship between MS and AS ground motion characteristics in a MS-AS sequence. It also evaluates the need to utilize these relations in developing artificial MS-AS sequences for seismic response evaluation of structures. To this end, a real ground motion database comprising of 192 MS-AS sequences is compiled from different ground motion databases. A univariate and a multivariate linear regression model quantifying the relationship between MS and AS ground motion characteristics in a sequence are developed. Artificial MS-AS sequences are simulated using these regression models. An analytical nonlinear model of four story modern reinforced concrete moment frame building is subjected to sets of real and artificial MS-AS sequences through incremental dynamic analysis (IDA). The IDA results from each set of MS-AS sequences are used to generate seismic collapse fragility curves. The results indicate that the median collapse capacity for the building model calculated using artificial MSAS sequence set based on regression models developed in the study match closely with the real MS-AS sequence set.