Risk assessment of reinforced concrete buildings against progressive collapse

This paper presents a two-scale computational model for probabilistic analysis of the collapse behavior of reinforced concrete (RC) buildings subjected to local structural damage. In this model, structural members are modeled as elastic blocks connected by a set of nonlinear cohesive elements, which represents various damage zones that could potentially form during the collapse process. The random constitutive behavior of the cohesive element is determined by the fine-scale stochastic finite element simulations of the corresponding potential damage zone under different loading scenarios. The proposed model is validated through the numerical simulations of recent experiments on a RC frame subassemblage and two flat-slab systems. With the proposed model, a stochastic analysis is performed to investigate the probabilistic collapse behavior of a 10-story RC building under various initial damage scenarios, where the random material properties and gravity loading are sampled by using the Latin Hypercube Sampling (LHS) method. Through stochastic simulations, the occurrence probabilities of various collapse scenarios are calculated and the results are compared with those obtained by using the existing deterministic analysis method.