Bridge Rating Based on In-Situ Weigh-In-Motion and Health Monitoring Data

Abstract

Parallel to the bridge design methodology changes from the Allowable Stress Design to the Load Factor Design, and then to the reliability based Load and Resistance Factor Design (LRFD), bridge load rating method has also been evolving. Applying the reliability theory to the bridge load rating is more complex than applying to the LRFD since any conservatism can have a significant effect on the assessment of bridge capacity, particularly in load posting and bridge replacement. Although the current Load and Resistant Factor Rating (LRFR) method applying the concept of reliability analyses, it uses very limited site-specific data due to practical constraints and the limited availability of site-specific data. The objective of this study is to develop a reliability based rating approach, grounded in in-situ responses from long-term structural health monitoring systems and actual unbiased traffic data from weigh-in-motion stations. Rating bridges that use actual bridge in-service measurements and site-specific traffic can remove conservatism and uncertainties in association with load distribution factors, dynamic impact, and secondary and non-structural element effects. The end goal is to achieve a continuous bridge evaluation model for real-time vehicle loads, which in turn can be used for speedy truck permitting, bridge management, and identifying sudden condition changes to ensure public safety. The bridge site-specific truck data and bridge peak strains under ambient traffic for the instrumented bridge have been continuously collected for over a year. The time dependent values of the maximum live load effects are obtained from the statistical analysis of the in-service responses and traffic data. The site-specific live load distribution factors are developed and live load factors are re-calibrated based on reliability analysis. Statistical distribution and projection methods have been compared and validated. This study suggests that the Gumbel distribution and the Parent Tail projection method will be the most suitable methods for the live load distribution and maximum live load effect projection. The reliability-based in-service traffic rating result is compared to three other rating methods: the simplified distribution method, the finite element method, and the live load testing method. The load rating results based on the updated load and load distribution have improved tremendously compared with other rating methods. This systematic rating approach can provide essential information for future bridge maintenance and replacement prioritization. Additionally, a more accurate posting sign is recommended for future bridge load limits.