Bridges

In this study, a statistical pattern recognition technique was developed based on time series analysis to localise cracks in steel reinforced concrete structures using vibration data. The technique has been developed for the Sydney Harbour Bridge – one of Australia’s iconic structures and the world’s largest steel arch bridge. Measurements are collected from tri-axial accelerometers (MEMS type), which are integrated into sensor nodes developed at the National ICT Australia (NICTA). The sensors are located on the underside of the bus lane, near the eastern side of the bridge deck. Sensor nodes monitor vibration levels whenever a vehicle passes over the sensors (passive approach) and data are collected as the vibration level exceeds a pre-set threshold. Our approach uses two-stage Auto-Regressive (AR) and Auto-Regressive with eXogenous inputs (ARX) prediction models to extract damage-sensitive features from vibration data in the time domain. Variation between the residual errors obtained from the intact and damaged states were used to identify unusual behaviour due to structural damage. The results demonstrated the effectiveness of the approach in handling large volumes of data and the ability to identify the presence of cracking. After the initial research project, the laboratory testing and the onsite trial, the technique is being developed as a tool to aid in the management of the asset.