System identification of a historic swing truss bridge using a wireless sensor network employing orientation correction

System identification was performed on the swing span of a steel truss bridge dating from 1896 using acceleration data collected from a wireless sensor network (WSN). The swing span can rotate 360 degrees to allow river traffic to pass through the locks located under the bridge. The WSN installed on the swing span consists of 23 nodes that measure synchronized tri-axial acceleration. Five days of measured data were segmented according to three functional positions of the span (locked facing downstream, locked facing upstream, swung for river traffic) and load conditions. Subsequently, the modal parameters corresponding to the bridge's three functional positions were obtained using frequency domain decomposition method. The initial system identification of the bridge resulted in significant anomalies, in comparison with the finite element (FE) model. To improve the accuracy of the overall identification results, the sensor orientation correction technique is proposed for the measured data from the WSN. The improvement has been verified by comparing the orientation-corrected mode shapes with non-corrected mode shapes and FE mode shapes using the modal assurance criterion values. Analysis of the modal parameters inferred that the boundary conditions for the bridge positions are different because of the interaction of the locking mechanism with the abutments in those positions. The proposed approach for system identification, including sensor orientation correction, is to be used as part of the comprehensive structural heath monitoring strategy for this historic structure. Copyright (c) 2014 John Wiley & Sons, Ltd.