Intertial standards development for track maintenance

Methods of using inertial measurements for track geometry variation, either directly or in setting standards, have been reviewed by the authors. It was found that the normal approach is still to use chord-based measurements, even though, in most cases, inertial measurement is the instrumentation method. Acceleration measures are generally converted to displacement using double integration and then chord equations are used to convert measurements to traditional formats. The possible value of including the additional information obtained from inertial measurement was evaluated by simulating five different vehicles in scenarios of P1 level defects as defined in AS7635 [1] and ARTC Section 5 [2] standards. Derailment criteria were evaluated using 800 simulations including variations in defect length. Correlation between axle inertial data and wheel unloading was evaluated. It was found that there is weak correlation with wheel unloading in loaded freight wagons indicating good suspension performance and, conversely, strong correlation with wheel unloading in empty freight wagons indicating poor suspension performance. The implications of this for measurements systems are discussed. Various approaches to signal processing were explored, but a principal difficulty of all data processing approaches is the final conversion of displacement data to measurements relating to the track site. Irrespective of the processing, the act of windowing data introduces a ‘chord type calculation’, and this step introduces the problem of relative measurement and vanishing measurements at cyclic defect wavelengths. A new approach to processing track displacement data has been proposed that utilises a Local Minima approach. The method was demonstrated for isolated defects, cyclic defects and track spectra.