Abstract

In codes’ provisions and design procedures for acceleration-sensitive nonstructural components, seismic demand is commonly defined by means of floor response spectra expressed in terms of pseudo-acceleration. Depending on the considered analysis method, floor response spectra may be derived from floors’ acceleration histories, based on structural response-history analysis, or calculated using a predictive equation from a given input ground motion spectrum. Methods for estimating floor response spectra that are based on the second alternative are commonly called spectrum-to-spectrum methods. The objective of this paper is to briefly review these methods, and to discuss the main assumptions they are based on. Both predictive equations from selected seismic codes and proposals from the literature are included in the review. A new probability-based method, recently developed by the Authors for generating uniform hazard floor response spectra, namely, floor response spectra whose ordinates are characterized by a given target value of the mean annual frequency of being exceeded, is also described. By using this method floor spectra are determined through closed-form equations, given the mean annual frequency of interest, the damping ratio of the spectra, the modal properties of the structure, and three uniform hazard ground spectra. The method is built on a proposal for a probabilistic seismic demand model that relates the ground spectral acceleration with the floor spectral acceleration, and is able to explicitly account for the ground motion variability of the nonstructural response. Results for a case study consisting of a service frame of a visbreaking unit in an oil refinery are presented to show the good predictive accuracy of the method with respect to exact uniform hazard floor response spectra obtained through a standard probabilistic analysis.

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call