Mainshock-aftershock fragility and risk analysis for RC frame structures based on Bayesian networks

Mainshock-aftershock (MS-AS) sequences significantly exacerbate the seismic fragility and risk of structures, necessitating comprehensive probabilistic frameworks for assessment. This study develops a comprehensive Bayesian network (BN)-based methodology for MS-AS fragility and risk analysis of reinforced concrete (RC) frame structures. The proposed framework elaborates on the mathematical basis for modeling MS-AS probabilistic risk analysis using BNs, integrating both intensity-dependent and state-dependent fragility modeling. By leveraging the probabilistic reasoning capabilities of BNs, the model seamlessly combines probabilistic seismic hazard analysis (PSHA), structural response analysis, and loss estimation within a single graphical framework. A case study of a RC frame in Xi’an, China, demonstrates the model’s efficacy in deriving mainshock and aftershock physical vulnerability distributions, fragility curves, and probabilistic loss assessment. Results illustrate the significant influence of mainshock damage on aftershock fragility and the corresponding escalation in economic loss. The BN approach not only enhances computational transparency and flexibility but also supports both forward prediction and backward inference, offering a powerful tool for performance-based earthquake engineering.