Time-Domain Finite Element Model Updating for Operational Monitoring and Damage Identification of Bridges

Abstract

The well-known limitations of modal system identification methods have led to a broad exploration of alternative solutions for operational monitoring and damage diagnosis of structures. This study presents a time-domain Bayesian finite element model updating approach to jointly identify the vehicular loads and finite element modeling parameters of bridges using the vibration data and the location of vehicles traversing the bridge as input. A Bayesian model updating is devised and verified through a series of case studies based on numerically simulated data from a prestressed reinforced concrete box-girder bridge model. Damage states are defined for concrete degradation and delamination, steel corrosion, and loss of prestressing force. Ten different damage scenarios, encompassing the range from minor localized to major distributed damage, are examined. The responses of the damaged bridge are simulated under random traffic scenarios. The acceleration responses, along with the location of the vehicles on the bridge, are used for jointly estimating the model parameters and vehicular loads. The estimated model parameters are then used to infer the location and extent of damage within the bridge. The results show the successful performance of the proposed approach in a numerically simulated environment.