Updating probabilities of bridge reinforcement corrosion using health monitoring data

Abstract In this paper a methodology is presented to update estimated level of corrosion damage using routine short-term health monitoring measurement data. The case study bridge is a reinforced concrete (RC) slab. It is subject to traffic loading from a weigh-in-motion database and time-dependent deterioration is modelled through corrosion. Different damage indicators based on different virtual sensor measurements (strain, deflection and rotation) are investigated. A method of comparing the performance of the damage indicators is presented and their data is used to achieve a Bayesian update of prior estimates of bridge reinforcement corrosion. It is shown how, with few and uncertain data, the estimation of deterioration level can be significantly improved, particularly with damage indicators based on rotation.