Seismic fragility assessment and maintenance management on regional bridges using bayesian multi-parameter estimation

Abstract

Bridges are the most vulnerable elements in regional transportation networks. Seismic fragility assessment has significant implications regarding the potential damage of regional bridges and their seismic maintenance operations. Unlike individual bridges, assessing bridges at the regional scale involves greater diversity and distinct ground motions. This paper proposes a Bayesian parameter estimation methodology to assess the seismic fragilities of regional beam bridges and schedule their maintenances. The proposed multi-parameter models contain several structural parameters and intensity measures to comprehensively consider the regional diversity. The selected intensity measures represent the diversity of regional ground motions at the peak effect, spectrum intensity, and the duration of strong motion. To demonstrate the applicability of the proposed methods, this paper analyzes the seismic damage results of 13,500 bridge-ground motion pairs generated to train the fragility models. Results of the proposed models are similar to the seismic fragility probability derived from Monte-Carlo simulation results. This study proposes a two-level strategy to effectively manage regional bridge maintenance. These priority maintenance and secondary maintenance levels are determined based on the results of regional seismic fragility analysis. Additionally, significant sequences of structural parameters and intensity measures are identified using the leave-one-attribute-out analysis method. This process helps bridge designers and managers better understand the sensitive parameters pertaining to different components. This study also tests the proposed models on a virtual transportation network to assess the seismic fragility of its associated regional bridges, resulting in a two-level regional maintenance plan for that specific system.