Reliability assessment of the robustness of reinforced concrete frame under column loss scenario

Current code rules do not provide details in identifying the safety level associated to progressive collapse scenario regarding reinforced concrete (RC) buildings. This study deals with a probabilistic approach to assess the robustness of 2D RC moment resisting (MR) frames designed assuming a high ductility class. The assessment is based on the removal of the central supporting column as critical failure scenario for three distinct frames: one designed according to current code rules and other two with longitudinal reinforcement arrangement modified to enhance structural robustness. Employing a full probabilistic approach, 100 samples are generated for each frame, including uncertainties on both material properties and loads. Pushdown non-linear finite element (NLFE) analyses are carried out for the 300 sampled models for increasing displacements imposed at the top of the removed column to determine energy-based dynamic amplification factors. Then, 300 probabilistic static-equivalent NLFE analyses are performed by simulating column removal and properly amplifying gravity loads. The strains in concrete and reinforcement in different cross-sections of the sampled frames are monitored and probabilistically modelled. Finally, failure probabilities with respect to the ultimate limit state are computed between aleatory strains and corresponding ultimate thresholds. The results highlight higher reliability for the RC MR frames with specific robustness improvements.