MONITORING OF RAIL-TRACKS BASED ON MEASURED ACCELERATION DATA

Abstract

Railway tracks are used as mass transportation system for transporting large number of people and goods from place-to-place to keep the economy running smoothly. Hence it is inevitable to keep the tracks healthy for safe and on-time movement of trains. Traintracks are complex systems that contain ballast, sleepers, fasteners and rails. Therefore, monitoring only one/two elements (e.g., ballast/train-track) will not provide enough information to understand the overall performance of the railway tracks. To tackle such issue, herein a sensor fusion i.e., accelerometers, fiber-optic sensors, strategy is adopted and sensors are placed at different locations of a real rail-track. In order to measure the vibration signal four accelerometers are employed, first one is placed on the rail (between two sleepers), second one is installed on the rail but above the sleeper, third one is exactly on the sleeper, and last one is on the precast railway trough. In a first step, the investigation has focused into accelerometers data only. The tests are performed for the following loading conditions: (i) shaking the track via an APS400 type shaker, (ii) hitting the track by an impact hammer, and (iii) by passing a real train on the track. The time-series data are analyzed and the frequencies and spectrums are estimated via the use of fast Fourier transform (FFT). The changes of frequencies of the tested rail-track at different locations due to the various loading conditions are observed. In a later step, an autoregressive type time-series model has been developed and validated where the initially obtained results show good agreement with the measured data. The current findings will assist to monitor the rail-track for any further changes.