A Direct Integration Approach to Drive-by Damage Monitoring of Railway Tracks

Railway tracks can be monitored by visual inspection or, more recently, indirectly using inertial sensors installed in a passing vehicle. Defects in the track such as depressions in the profile or points of low stiffness (e.g. hanging sleepers) interact dynamically with passing vehicles and can be detected with accelerometers and gyrometers. This can be achieved using special purpose track recording vehicles or through instrumentation of regular trains in service. It has been shown in previous research that an optimisation procedure can be applied to back-calculate track profiles from vehicle-mounted sensor data. This involves finding the profile that gives a best fit to the measured data. In this paper, a new direct integration approach is introduced to find the same track profile in a fraction of the computing time. The Newmark-Beta method is used. Compared with the profile calculated using the optimisation algorithm, the results are similar. However, direct integration is much more efficient than optimisation and allows the calculation to be completed in a fraction of the time. The calculated track profile can be used to estimate points of low stiffness.