Vibration-based damage assessment of the Boirs viaduct with a multistage model updating approach

Damage qualification is generally of concern in vibration-based damage assessment of bridges and structures, in which the model-based method, such as the finite-element (FE) model updating, is often a favorable choice in terms of efficacy and accuracy. However, the numerical problems to be solved are often ill-conditioned and sometimes even ill-posed, due to the limited number of the measurements and the large number of potential parameters to be updated in the FE model. Furthermore, in real-world applications, the effects of experimental and modeling errors. need to be accounted for From the practical point of view, there is often a trade-off between the cost of the vibration measurements and the precision of the experimental results, and another trade-off between the computational efficiency of the simplified FE model and the numerical fidelity of the refined FE model. The satisfactory solution of these problems is crucial to the success of the damage assessment in real-world applications, which remains scarce. In this paper, a multistage model updating approach is proposed for damage quantification of a prestressed concrete girder bridge by using the results of an extensive experimental campaign. Although severe corrosion damage had been found by visual inspection before the measurements, the vibration-based damage assessment was performed without a priori knowledge of the existence and location of damage. The appealing feature of this work is the scale and complexity of the studied structure. By updating a simplified orthotropic plate model, the damage was satisfactorily identified in terms of its extent and severity for both a simulated damage scenario and the real structure. A good balance between the accuracy and efficiency has been reached by the proposed multistage model updating approach. The developed methodology can be applied to similar multigirder bridges, which are common in the regional highway network.