Method for Probabilistic Evaluation of Seismic Structural Damage

Abstract

This paper presents a systematic approach for estimating fragility curves and damage probability matrices for different structural systems. Both fragility curves and damage probability matrices express the probabilities that a structure will sustain different degrees of damage at given ground motion levels. In contrast to previous approaches, this paper presents a method that is based on nonlinear dynamic analysis of the structure rather than on heuristics or on empirical data. The ground motion level for fragility curves is characterized by spectral acceleration. For damage probability matrices, modified Mercalli intensity is used as the ground motion parameter. The probabilities associated with the different damage states at a specified ground motion level are evaluated using the Monte Carlo-simulation technique. The nonstationary autoregressive moving average (ARMA) model is used for the generation of earthquake time histories. The approach presented in this paper is used to obtain fragility curves and damage probability matrices for reinforced concrete frames. Three different classes of reinforced concrete frames, based on the number of stories, are considered. Park and Ang's damage index is used to identify the different degrees of damage.