Development of a novel fire following earthquake probabilistic framework applied to a steel braced frame

Abstract

This paper describes a novel probabilistic fire following earthquake (FFE) framework to develop FFE fragility functions for steel braced frames. In particular, an 8-storey steel braced frame, located in a high seismic area, was used to demonstrate the framework. An algorithm was formulated to generate fire scenarios based on seismic damage. Damage caused by the earthquake was expressed in terms of inter-storey drift ratio (IDR) and peak floor acceleration (PFA). Damage was captured in structural and non-structural components, such as glazing and partition walls. In particular, floors with IDR and PFA that exceeded pre-defined thresholds were considered to experience ignition after an earthquake. More than 1100 nonlinear FFE analyses were performed by randomly generating values of window widths, fire load densities, earthquake intensities, and yield strength of steel at ambient and at high temperatures. Historic ground motions and natural fire curves were employed to characterize the hazards. Thermomechanical analyses were completed and failure criteria based on displacement and displacement rate were applied to the girders (primary beams) and columns. The results of simulations were processed to generate FFE fragility curves and surfaces for girder and column failures with respect to the time to failure in the FFE scenario and grouped based on spectral acceleration intervals. The results showed that the girder always failed first given the low structural redundancy due to the end conditions. Moreover, the higher the spectral acceleration, the more uniform the damage across the structure with a lower time to reach failure. It was finally found that the fire load density did not have a significant effect on the probability of failure for the case study under consideration.