Acceleration Based Condition Monitoring of Steel Bridges

Abstract

This paper presents a frequency domain-based method for continuous health monitoring of steel truss bridge, based on the power spectral density and frequency analysis of the recorded acceleration response time histories at the selected node points of a bridge under regular operational conditions. This method is also implemented on a railway steel truss bridge connecting the mainland India to Rameswaram, named the Pamban bridge. The measured acceleration response time histories at some selected nodes on the bridge were parameterized into 4 power spectral density and frequency-based parameters. Also, the frequency analysis of the acceleration response time histories was done. The continuous variation of these parameters with measurement points was observed for about a year to investigate the suitability of these parameters to detect, locate damages in a railway steel truss bridge, and find their severity. Then, the consistency of damages detected by the variation of these parameters and the damages expected based on strain measurements and visual inspections made on the bridge was examined to evaluate the suitability of these parameters for the level 3 damage detection. The same is documented and performance of the parameters discussed.