| description abstract | Characterizing structural resiliency after severe damage to a few load-carrying members is challenging. Engineers use various computational approaches to assess the vulnerability of structures and to evaluate the parameters that affect their response. Few of these approaches are capable of predicting the actual peak load-carrying capacity a damaged structure can withstand before experiencing total collapse, and practically none of them have been verified against actual test results. In this paper, experimental data from large-scale tests on steel–concrete composite floor systems under different column loss scenarios were used to develop and to validate a high-fidelity numerical modeling approach capable of predicting the response of the tests up to total collapse. This approach incorporates geometric and material nonlinearity, explicit modeling of steel and concrete failure, and contact modeling using LS-DYNA version R10.2.0. Nearly all specimen components were modeled using brick elements, including the concrete slab, steel members, bolts, and other connecting elements. The corrugated metal decking was represented with shell elements, and beam elements represent the reinforcing steel and shear studs. The predicted response and ultimate load–carrying capacity up to total collapse show good agreement with the results of the experimental tests. Validating the numerical models revealed the sensitivity of various modeling parameters and demonstrated the potential for inaccurate predictions of response when certain parameters were not correctly specified. The most important of these parameters are described in this manuscript. Lessons learned from the current study are helpful for understanding the mechanisms that have the greatest impact on collapse of composite floor systems, and these lessons can be used to gain insight on the collapse potential of other structures with different geometries or configurations. | |
