Integration of Element Inspection Data in Model Updating and Performance Evaluation of In-Service Bridge Superstructures

With the volume of aging bridges approaching the end of their service lives nationwide, maintenance of these deteriorated structures is a growing challenge for transportation agencies. To manage this infrastructure, inspectors are tasked with providing an average rating based on element-level condition measurements. According to the National Bridge Inspection Standards (NBIS), many of the proposed inspection methods are visual and as a result, the evaluated condition states are generally qualitative and in most cases based on engineering judgments. To combat this issue, agencies have begun to utilize a variety of technologies to monitor the behavior and detect different sources of damage in the bridges and generate higher fidelity condition metrics. These condition metrics are used in the load rating process, which mainly focuses on the behavior of degraded elements to assess the remaining capacity of the bridge system. However in these practices, less attention is attributed to the overall system-level response under the reported condition states. An appropriate understanding of in-service performance requires knowledge of characterizing the influence of damage and deterioration on the overall system behavior. Today’s computational capabilities provide a tool to assess these characteristics using numerical and theoretical models representing the bridge structural system. Moreover, recent advances in monitoring and inspection techniques provide a mechanism to efficiently capture the damage data necessary for model updating. Using non-linear FE analysis, this paper presents a case study on evaluating the behavior of representative in-service bridges in the Commonwealth of Virginia, under the effect of deteriorating conditions. The geometrical characteristics of the selected structures represent common features of steel-concrete composite bridges serving within the state. The range of the corresponding damage characteristics were also selected through a questionnaire which collected inspection data from the VDOT engineers in nine different districts across the state.