System-reliability-based Disaster Resilience Evaluation of Cable-stayed Bridge under Fire Hazard Using Reliability-Redundancy Analysis

Abstract

The concept of disaster resilience recently emerged in efforts to gain holistic understanding of civil infrastructure systems exposed to various natural or human-made hazards. To effectively evaluate the resilience of complex infrastructure systems generally consisting of many interdependent structural components, Lim et al. (2022) proposed a system-reliability-based framework for disaster resilience. In the proposed framework, the disaster resilience of a civil infrastructure system is characterized by three criteria: reliability, redundancy, and recoverability. For comprehensive resilience analyses at the scale of individual structures, the reliability (β) and redundancy (π) indices were newly defined in the context of component- and system-level reliability analysis, respectively. Reliability-redundancy diagram, i.e., the scatter plot of the reliability and redundancy indices computed for each initial disruption scenario, was also proposed to help a decision-maker check whether the corresponding risk is acceptable for the society. In this paper, we demonstrate the framework through its application to a cable-stayed bridge in South Korea, the Seohae Grand Bridge under fire hazards. First, a probabilistic model is developed to describe the hazard of fire scenarios that may occur on the deck of the cable-stayed bridge. Next, finite element simulations are performed to compute the reliability and redundancy indices through component and system reliability analyses for the fire accident scenarios. An adaptive simulation method, AK-MCS (Echard et al. 2011), is employed to overcome the computational cost issue. The example successfully demonstrates that the reliability-redundancy analysis and diagram facilitate a comprehensive assessment of the disaster resilience of a complex civil infrastructure such as a cable-stayed bridge by using sophisticated computational simulations and advanced reliability methods.