Probabilistic progressive collapse assessment for RC framed-wall structure

Abstract

Since the progressive collapse of structures could cause disastrous consequences for humans and society, therefore assessing and improving the vulnerability of our buildings against progressive collapse is urgently needed. The current research presents a high-fidelity simulation procedure for an existing RC framed-wall structure and aims to investigate the structure’s resistance to progressive collapse under column/shear wall removal scenarios at different levels through deterministic analysis. Based on the considered removal scenarios, the dynamic effect of the sudden column/shear wall removal is studied, and a simple procedure to predict the vertical reaction of the neighboring columns/shear walls to the removed one is proposed with a dynamic amplification factor (DAF) 1.80 and 2.40 for the column and shear wall removal, respectively. The results revealed that the sudden column/shear wall removal at the higher floor is more critical than the lower one; since more floors participate in absorbing the released energy in the latter. Moreover, in some cases, the horizontal displacement resulting from the sudden shear wall removal is found to be more critical than the vertical displacement. Since there are several sources of uncertainties that significantly affect the building resistance to progressive collapse and the deterministic analysis neglects these uncertainties; therefore, it cannot quantify the failure probability of the structures. Consequently, based on the deterministic analysis results, the critical removal scenario is selected to perform the stochastic analysis via the probability density evolution method (PDEM), where both the stochastic response and the instantaneous probability density function (PDF) can be obtained. Finally, the failure probability of the studied structure is assessed using the PDEM combined with the absorbing boundary condition. The stochastic analysis revealed that the maximum vertical displacement increased to roughly 1.6 times its value compared with the deterministic analysis. Simultaneously, a failure probability of approximately 40% is recorded through the stochastic analysis, while no global or even local collapse is observed in the deterministic analysis.