Probabilistic seismic demand model for nonstructural components

Past earthquakes have shown the importance for critical facilities to remain functional during seismic events. In the performance assessment of these facilities, it is therefore essential to accurately evaluate the seismic demand of nonstructural components. Evaluation of these components is also important when estimation of nonstructural repair costs is a concern. In this paper, the use of a multivariate demand model for nonstructural components is proposed, in which demand is expressed in terms of both interstory drifts and floor acceleration spectra. A model is built using statistics of the demand vector derived from the results of a limited number of inelastic response history analyses of a structure. The model is then used to simulate any number of additional realizations of the demand vector required for an accurate estimation of the probability of functionality loss. A new proposal for a predictive equation to generate approximate realizations of floor response spectra is presented. A reinforced concrete frame is selected as an illustrative example to show the implementation of the probabilistic seismic demand model and to evaluate the proposed predictive equation for the floor response spectra. The results of the case study are used to demonstrate the importance of accounting for the correlation among the demand parameters when realizations of the seismic demand of nonstructural components are simulated. © 2016 John Wiley & Sons, Ltd.