A new indicator of elastic spectral shape for the reliable selection of ground motion records

Abstract

How to select a limited number of ground motion records (GMRs) is an important challenge for the nonlinear analysis of structures. Since epsilon (ε Sa ) is an indicator of spectral shape, which has a significant correlation with the non-linear response of a structure, the selection of GMRs based on the hazard-related target ε Sa is a reasonable approach. In this paper, an alternative indicator of spectral shape is proposed, which results in a more reliable prediction of the non-linear response for the structures with the natural period of 0.25 to 3.0 s. This new parameter, named eta ( ), is a linear combination of ε Sa and the peak ground velocity epsilon (ε PGV ). It is shown that , as a non-linear response predictor, is remarkably more efficient than the well-known and convenient parameter ε Sa . The influence of -filtration in the collapse analysis of an eight-story reinforced concrete structure with special moment-resisting frames was studied. Statistical analysis of the results confirmed that the difference between ε-filtration and -filtration can be very significant at some hazard levels. In the case of this structure, the resulting annual frequency of collapse was found to be lower in the case of -based record selection, in comparison with the ε-based record-selection approach.

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

“…In combination 11, which contains " Sa and " PGV , the value obtained for the coe cient of " PGV is {0.836. This value is very similar to that of {0.826 obtained by Mousavi et al [12] as the coe cient of " PGV in parameter . Thus, considering the di erences between the SDOF systems and ground motion records used in both studies, it can be inferred that the optimization algorithm used does not have signi cant e ect on the results.…”

“…Mousavi et al [12] used the genetic algorithm [30] to propose parameter . As described previously, in this study, the PSO algorithm was used to optimize coe cients k i in each combination of " IM parameters.…”

“…To propose an e cient proxy for GMRS, Mousavi et al [12] investigated the linear combinations of " Sa with epsilons of di erent peak ground IMs (i.e., PGA, PGV, and PGD) to be used instead of " Sa . They proposed a new proxy termed \eta" (), which is a linear combination of " Sa and the peak ground velocity epsilon, " PGV (see Eq.…”

The dependence of conditional spectra on the choice of target periods

“…(2), Epsilon is the key parameter in the development of CMS and it contains information from the elastic spectral shape of the considered record. On the other hand, a new parameter, named Eta, has been proposed in literature in order to improve the conventional Epsilon e ciency [9,10]. The Peak Ground Velocity (PGV) has been utilized to CMS amplitude is usually less than UHS amplitude.…”

“…To clarify, error terms were de ned as written in Eqs. (9) and (10) in the cases of CMS and ECMS, respectively. A schematic description of the error terms is also shown in 10) where n is the number of CMS cases, T is the target period, is the critical damping ratio, and T 0 is the upper bound of the considered period for the error integration.…”