Since 2022, the Bane Nor Measurement train is equipped with a TDL (Track Dynamics logger) designed by SD&M - France. The TDL captures, by means of 4 triaxial accelerometers on the bearings of a bogie and 4 water resistant microphones, both the forces and noise emission generated by the complete Norwegian railway network, automatically and without need for any user interaction. The acquired data (0-20kHz) is processed towards parameters usable for noise mitigation and maintenance planning, and both raw and processed data are transferred through a 4G/5G connection to a dedicated cloud repository. In Norway, about 4000 kilometers of tracks are covered by the TDL, recording parameters for protection of the infrastructure (rails, sleepers, ballast), for troubleshooting annoyance of citizens (noise, vibrations), and for verification of maintenance activities (grinding, tamping, switches, etc.). Roughness spectra (1-50 cm wavelengths), squat detection, and railpad damping properties are some examples of parameters that can be extracted from the data. Here we will focus mainly on detection or rail defects such as squats, studs, and weldings, by applying advanced signal processing such as band filtering, wavelets, and specific issue-detection algorithms. A comparison can be made with other measurement techniques, e.g., laser scanning or high-speed photography.
Read full abstract