Cluster computing‐aided model updating for a high‐fidelity finite element model of a long‐span cable‐stayed bridge
Kaiqi Lin,You‐Lin Xu,Xinzheng Lu,Zhongguo Guan,Jianzhong Li
DOI: https://doi.org/10.1002/eqe.3270
2020-03-07
Abstract:<p>Accurate and high‐fidelity finite element (FE) models are in great demand in the design, performance assessment, and life‐cycle maintenance of long‐span cable‐stayed bridges. The structural system of a long‐span cable‐stayed bridge is often huge in size and complex with many components connected and various materials constituted. Therefore, the FE model of a long‐span cable‐stayed bridge involves a large number of elements and nodes with many uncertainties. The model updating of the FE model to best represent a real bridge is necessary but very challenging. One of the challenging issues is that the numerical computation needed for searching the global optimum of a large set of structural parameters is so extensive that the existing FE (not surrogate) model‐based updating methods cannot fulfill this task. In this study, a cluster computing‐aided FE model updating framework is proposed for the high‐performance FE model updating of large and complex structures. In the framework, several computer software packages, including MSC.Marc, Python, and MATLAB, are interconnected for making use of their respective functions of strength. The shake table test of a scaled physical structure of the Sutong cable‐stayed bridge in China is used to validate the accuracy and efficiency of the proposed framework. The simulated bridge responses based on the updated FE model are in good agreement with the measured ones from the shake table test. The successful application of the proposed framework provides a reference for the model updating of other types of large and complex structures.</p>
engineering, civil, geological