Estimating the Response of Steel Structures Subjected to Vertical Seismic Excitation: Idealized Model and Inelastic Displacement Ratio
Yang Xiang,Yong-feng Luo,Zhao-chen Zhu,Zu-yan Shen
DOI: https://doi.org/10.1016/j.engstruct.2017.06.043
IF: 5.582
2017-01-01
Engineering Structures
Abstract:Vertical seismic excitations (VSEs) could induce severe damage to building structures and infrastructural facilities, especially to those with large-spans or long cantilevers. Thus, there is a need in engineering practice for estimating the seismic response of structures under VSEs. In this paper, a tentative investigation on VSEs-induced inelastic response of steel structures is proposed. The structure is transformed into an equivalent single-degree-of-freedom (ESDF) system in an extended pushover process, in which the vertical component of response is included. Considering the influence of gravity, a novel ESDF model with asymmetric vibrating characteristic is established. In this ESDF model, the inelastic response only occurs at the "lower" side of vibration, whereas at the other side, the system keeps elastic. Based on the idealized ESDF model and 53 recorded strong VSEs, the vertical inelastic displacement ratio (IDR) is computed. The averaged IDRs are plotted against the elastic vibrating period T, and fitted by a piece wise function to generate the IDR spectra, in which the effects of strength reduction factor R and of post-yield stiffness ratio a are accounted for by fitting coefficients. The overall process for estimating the VSE-induced inelastic responses of steel structures, using the extended pushover procedure and the proposed IDR spectra, is summarized. In the numerical example, seismic responses of a large-span space truss subjected to VSEs ate computed by means of the nonlinear RHA method and the proposed IDR-based method, respectively, the results demonstrate that the IDR-based method could yield an anticipative estimation of structural responses under severe VSEs. For mid-span displacement, error of the IDR-based method is less than 13.6%; for quantity of yielded member, error is less than 14.2%. (C) 2017 Elsevier Ltd. All rights reserved.