Updating Multiscale Model of a Long-Span Cable-Stayed Bridge: Fid-Bau Portal

Updating Multiscale Model of a Long-Span Cable-Stayed Bridge

Wang, Feng-Yang / Xu, You-Lin / Sun, Bin / Zhu, Qing

To accurately calculate stress responses and evaluate fatigue damage of the critical structural members/joints of a long-span cable-stayed bridge, a multiscale finite-element (FE) model of the bridge has been developed by using shell/plate elements to simulate the critical structural components (local models) and by using beam/truss elements to simulate the rest part of the bridge (global model). Nevertheless, the multiscale FE model will be updated to best represent the real bridge, and accordingly the multiscale model updating method is required. This paper presents a novel multiscale model updating method for long-span cable-stayed bridges. The many-objective optimization problem with four or more conflicting objective functions is first formulated because global model and local models will be updated simultaneously. The metamodel-assisted multiobjective optimization evolutionary algorithm (MOEA) is then developed by a combination of R2 indicator-based MOEA, kriging metamodel, and evolution control strategy for the multiscale model of the bridge, which is large in size and complex in system. The R2 indicator-based MOEA is used because of its high performance in solving the many-objective optimization problem. The kriging metamodel is used to improve the computational efficiency of the optimization. The evolutionary control strategy is developed to prevent the R2-MOEA from finding false optimal solutions or losing some of the optimal solutions. Finally, the developed method is applied to a long-span cable-stayed bridge in Hong Kong to demonstrate its feasibility and accuracy. The modal frequencies, displacement, and stress influence lines measured by the structural health monitoring (SHM) system installed in the bridge are used to define the multiple objective functions. The updated results show that the proposed updating method is feasible and can improve the accuracy of the multiscale model in global and local structural responses.