Reliable seismic collapse assessment of short-period structures using new proxies for ground motion record selection

One of the important steps in seismic collapse assessment of structures by using nonlinear dynamic analyses is the appropriate selection of ground motion records. Epsilon (εSa), eta (η), and gamma (γ) for long-period structures, are proxies that have been recently proposed for Ground Motion Record Selection (GMRS). In this study, two parameters, named γs, are proposed, which have considerable correlation with the collapse capacity of short-period structures having fundamental period less than 1 s. One of these parameters is a linear combination of εSa, epsilon of Pseudo Spectral Acceleration (PSA) at 1.5 times of the fundamental period of the structure (εSa(1.5T1)) and εPGV. The other one is a linear combination of εSa, εSa(1.5T1) and epsilon of spectrum intensity, εSI. To obtain and optimize γs, the Particle Swarm Optimization (PSO) algorithm is applied. Since the parameters proposed as γs have significant correlation with the collapse capacity of short-period structures, they can be used as efficient proxies for GMRS in seismic collapse assessment of short-period structures. The results show that GMRS using γs leads to reduction in the dispersion of structural collapse capacity in comparison with GMRS based on εSa or η.