Probabilistic Demand Models and Fragility Curves for Reinforced Concrete Frames

Abstract

Fragility curves are constructed to assess the seismic vulnerability of a hypothetical two-story reinforced concrete frame building designed only for gravity loads. Fragility curves are also developed for the same building modestly retrofitted by means of column strengthening. A Bayesian methodology is used to construct probabilistic demand models to predict the maximum inter-story drifts, given the spectral acceleration at the fundamental period of the building. The data for the models are obtained using two-dimensional inelastic time history analyses of the building for a suite of synthetic ground motions, developed for the Memphis region. The models are developed using both equality data and lower bound data, and are developed to properly account for both aleatory and epistemic uncertainties. In the absence of probabilistic capacity models for gravity load designed structures, capacity limit states are considered based on FEMA 356 guidelines and deterministic nonlinear pushover analyses. The results quantify the vulnerability of low-rise reinforced concrete frame buildings and show the effectiveness of seismic retrofitting in reducing the probability of failure.