DAMAGE DETECTION OF RAILWAY BRIDGES USING STRAIN CORRELATION

摘要

Railway networks are among the most critical components of our transportation systems and maintenance of these networks is of the utmost importance. Bridges are vital links within the railway network and most of these bridges are approaching their design life. Also considering that axle loads have been increased several times during the past decades, developing effective Structural Health Monitoring (SHM) techniques for railway bridges is expected to help the infrastructure owners with their decision making strategies. In this context, developing specialized data analysis methodologies for railway bridges is the main objective of this study. In this paper, preliminary results of our ongoing efforts to develop a strain correlation based damage detection method for railway bridges are presented. This method offers an easier and faster way to locate the damage. First, a detailed finite element model (FEM) of a railway bridge is created. Then, this model is used to simulate and collect strain data under moving train loading for different structural conditions (healthy and damaged), which is then analyzed using a correlation based damage detection method previously introduced by the authors. The structural change (e.g. damage) in the bridge is assessed by comparing the correlation matrices. The comparison shows the changes in the correlation coefficients for each measurement location, so that structural changes can be detected and localized. Tracking the condition of the bridge continuously in this manner and early detection of potential structural changes are deemed very valuable for the infrastructure owners for developing more economical maintenance strategies and address the problems before they become too costly.