A paradigm for seismic resilience assessment of subway system

Seismic resilience of subway systems is a cornerstone for developing resilient megacities. It depends on infrastructures' fragility and aftermath recovery features at both local and global scales. Yet it remains obscure how to propagate the probabilistic seismic performance of single structures to the overall probabilistic seismic performance of subway systems at the global scale. Here, a paradigm is established for the seismic resilience of subway systems by integrating the seismic performance of single structures, interrelations among single structures, and post-seismic recovery. The seismic performance of single structures is assessed by nonlinear timehistory analysis and quantified in terms of analytical fragility curves. The interrelations among single structures are represented by mapping the subway route map into a directed and weighted topological graph, dictated by the graph theory. Subsequently, the overall seismic performance of subway systems is quantified by network indicators and estimated by a Monte Carlo algorithm. The post-seismic recovery is determined by adopting appropriate simplified recovery functions and recovery strategies. The proposed paradigm is implemented to assess the seismic resilience of a typical subway system. It reveals that the seismic resilience of subway system is extremely sensitive to the repair sequence, e.g., decreased by 63% when switching from the order of damage state to the reverse one. Moreover, it is recognized that reliability and uncertainty are coupled, e.g., an increase in the reliability of single structures may depress the uncertainty of the system's seismic performance.