Structural Analysis

Abstract

Since several years, seismic input definition is one of the hot topics of earthquake engineering because of its influence on simulations for estimating seismic structural performance. Herein, the efforts of the ReLUIS 2010-2013 project toward the development of practice-ready tools for hazard consistent seismic input definition aimed at seismic structural analysis is shown. Determination of design seismic actions in seismic codes mostly relies on a target spectrum, which is, therefore, also the basis for record selection in seismic input definition when performing nonlinear structural analysis. Since a rational performance target should account for the seismic hazard at the site of interest, the uniform hazard spectrum (UHS), or an approximation of it, is often used as the design spectrum. The UHS is built entering the elastic spectral acceleration, Sa(T), hazard curves for several T values at a specified probability of exceedance of (e.g., 10% in 50 years or, equivalently, 475 years return period, Tr), and plotting the corresponding ordinates versus T. Generally, the signals that can be used for structural simulation are of three types: (1) artificial waveforms; (2) simulated accelerograms; and (3) natural records. Signals of type (1) are often obtained via random vibration theory. Simulation records (2) are obtained via modelling of the seismological source and may account for path and site effects. Finally of type (3) are ground-motion records from real events (Bommer and Acevedo, 2004). As far as it regards real records, given the UHS for the structural limit-state of interest (i.e., the UHS corresponding to a Tr), current or advanced (depending on the context where it is applied) practice today, which may require aid by a seismologist, would select a set of records reflecting the likely magnitudes (M), source-to-site distances (R), and other earthquake parameters thought to drive the probabilistic seismic hazard analysis (PSHA) for the site (McGuire, 2004), and which are believed to matter with respect to structural response (This information comes from a procedure called disaggregation of PSHA). Finally, the records are usually manipulated to match the UHS, individually or in average sense, at the period of the first mode of the structure (T*), Figure 1, or in an interval around it (e.g., Iervolino and Cornell, 2005).