Cable-stayed bridge all-round morphology identification based on 3D point cloud model

Abstract

Traditional inspection methods, that is, total station, can be labor-intensive, time-consuming, and efficiently challenging, especially for long-span cable-supported bridges. Recently, Terrestrial Laser Scanning (TLS) has been attractive in bridge digitization because of its non-contact and high-efficiency features. It can provide a comprehensive and all-round morphology assessment through establishing the bridge’s 3D point cloud model. Based on our previous research on point cloud registration of long-span bridges, this article focuses on spatial form identification using the registered point cloud model, specifically the identification of cable sag and girder spatial volume. In detail, precisely monitoring cable sag is critical for cable force identification, while accurate girder volume is essential bridge information, which should be tracked regularly and used for long-term performance evaluation and maintenance guidance. To evaluate the precision of the proposed TLS-based inspection solution, a case study is conducted on a 152 + 370 + 152 m cable-stayed bridge. Results show the identifying error of TLS-based solution remains at 2% for cable sag identification and 5% for girder volume recognition. The traffic-induced sag variance can also be captured by the point cloud data. Therefore, the TLS-based inspection method is thoroughly evaluated and demonstrated as a highly efficient and trustworthy solution.