Probabilistic performance-based evaluation of a tall steel moment resisting frame under post-earthquake fires

This paper aims to develop a framework to assess the reliability of structures subject to a fire following an earthquake (FFE) event. The proposed framework is implemented in one seamless programming environment and is used to analyze an example nine-story steel moment-resisting frame (MRF) under an FFE. The framework includes uncertainties in load and material properties at elevated temperatures and evaluates the MRF performance based on various limit states.

Specifically, this work models the uncertainties in fire load density, yield strength and modulus of elasticity of steel. The location of fire compartment is also varied to investigate the effect of story level (lower vs higher) and bay location (interior vs exterior) of the fire on the post-earthquake performance of the frame. The frame is modeled in OpenSees to perform non-linear dynamic, thermal and reliability analyses of the structure.

Results show that interior bays are more susceptible than exterior bays to connection failure because of the development of larger tension forces during the cooling phase of the fire. Also, upper floors in general are more probable to reach specified damage states than lower floors because of the smaller beam sizes. Overall, results suggest that modern MRFs with a design that is governed by inter-story drifts have enough residual strength after an earthquake so that a subsequent fire typically does not lead to results significantly different compared to those of an event where the fire occurs without previous seismic damage. However, the seismic damage could lead to larger fire spread, increased danger to the building as a whole and larger associated economic losses.

Although the paper focuses on FFE, the proposed framework is general and can be extended to other multi-hazard scenarios.