A direct method to detect and localise damage using longitudinal data of ends-of-span rotations under live traffic loading

Abstract

Existing work on rotation-based bridge monitoring has focused on indirect methods, such as bridge weigh-in-motion or influence line approaches. However, these approaches require increased instrumentation complexity, and require calibration, necessitating bridge closures. In this paper, we explore the potential of using rotation measurements to create a more practical and cost-effective monitoring system. To this end, we present a damage detection method which directly analyses bridge rotation data measured under live, free-flow traffic loading. We show how the Earth Mover’s Distance, typically used in statistics and image processing, can be applied directly on end-of-span rotation measurement data to achieve effective damage detection and localisation. Numerical simulation results demonstrate the approach’s robustness to the confounding effects of temperature variation and traffic diversity (vehicle type, loading, and velocity). The direct rotation measurement approach is applied to data from an in-service short-span bridge to demonstrate the technique’s capability with free-flow traffic loading.

Detecting Vehicle Loading Events in Bridge Rotation Data Measured with Multi-Axial Accelerometers

“…The purpose of this work is to develop an event detection method which leverages the existing bridge rotation instrumentation (tri-axial MEMS accelerometers) and that can run on low-resource sensor nodes. In Figure 1, the blue blocks represent our existing rotation monitoring approach, described in [9], with the resultant damage detection metrics sent back to the bridge management system to inform operational decision making.…”

“…However, for the vast majority of the bridge stock, their structural condition is only assessed by infrequent, human-visual inspection, and continual, sensor-based monitoring is limited by practical issues such as limited power budget and communications capacity. The ideal, therefore, is to provide autonomous BSHM systems that can enable reliable, long-term rotation monitoring [9] in these resource-constrained environments. For example, BSHM needs only to record and analyse the rotation response when the bridge undergoes external loading as a result of vehicles crossing.…”