Contact point accelerations, instantaneous curvature, and physics-based damage detection and location using vehicle-mounted sensors

Abstract

The recent bridge collapses worldwide underscored the perilous state of the global bridge infrastructure. Structural health monitoring (SHM) methods, particularly vibration-based techniques, offer a non-destructive approach to evaluate bridge conditions. However, installing dedicated SHM systems on each of the millions of bridges worldwide is prohibitively expensive. Vehicle scanning methods (VSM) have emerged as a cost-effective alternative, leveraging vehicle-mounted sensors for bridge assessment. Previous VSM research has primarily focused on frequency and mode shape identification through accelerations derived from numerical models. A smaller subset has explored damage detection, either based on modal properties or direct analysis of vehicle accelerations. This study proposes a new damage index that distinguishes itself by establishing a physical connection between the accelerations recorded on the vehicle and the instantaneous curvature of the bridge. This analytical connection provides a straightforward and effective means for damage detection and localization in bridges using VSM, offering the advantage of efficiency and accuracy. As such, the proposed damage index avoids the use of modal properties for damage detection, which are known to be prone to several factors and are not necessarily accurate enough when used in VSM applications, and the over-fitting problems associated with data driven methods. Numerical investigations indicate that the proposed damage index is robust against the potentially adverse effects of several parameters including road roughness, measurement noise, bridge frequency, damage location, variations in boundary conditions, and vehicle speeds.