Theoretical Research on Suspension Bridge Cable Damage Assessment Based on Vehicle-Induced Cable Force

Abstract

As one of the most critical load-bearing components in suspension bridges, cables require accurate damage assessments. Contemporary research methodologies primarily rely on cross-validation across multiple cables, which can present challenges in identifying damage under certain conditions. To address this limitation, this study proposes an evaluation method utilizing the cable force of individual cables. A damage index is introduced by the ratio of vehicle-induced cable tension (defined as the ratio of vehicle-induced cable force without weight to its baseline value), and the finite element model of a suspension bridge is used to verify this index. Initially, the finite element model of a suspension bridge is established, and a large number of datasets are generated using this model. These datasets include vehicle weight and vehicle-induced cable force. Subsequently, Gaussian Mixture Model (GMM) clustering is employed to categorize the dataset according to lanes, thereby establishing baseline values for each lane. Finally, damage assessments are conducted using data from individual cables and are validated against the outcomes obtained from the upstream–downstream cable force ratio method. The results show that the data trend of the damage index is clearly visible in six lanes, with the most pronounced trend occurring in the lane farthest from the cable (the sixth lane). The robustness of the index is also verified by adding 2% Gaussian white noise data, providing a research basis for related engineering projects.