No abstract

Regions of extreme low-adhesion between the wheel and rail can cause critical problems in traction and braking. This can manifest in operational issues such as signals being passed at danger, or pessimistic network wide responses to mitigate for localised issues. Poor traction conditions can be caused by oil contaminants, rain, ice, condensation of water droplets (micro-wetting) or leaves on the line, where compressed leaf contamination can cause a rapid decrease in adhesion. The complexity of the problem arises as a result of the inability to directly measure and monitor all the factors involved. There remains a lack of real-time information regarding the state and location of low-adhesion areas across rail networks. On-board low adhesion detection technology installed to in-service vehicles is a suggested method to capture up-to-date adhesion information network wide and minimise significant disruptions and cancellations in railway schedules. This paper extends a principle of a model-based estimation technique previously developed in straight track running for operating in a curving scenario. The vehicle model of focus here will be a high-fidelity, multi-body physics representation of a full-vehicle.

No abstract