TRAINWHEELS · A novel technology for detecting train wheels surface cracks by non destructive testing based on thermography

The European rail network is getting more and more important in the transportation sector: it hosts high speed-trains, higher loads of more passengers or goods, with increased frequency. This determines a higher mechanical load for the moving parts of the carrier involved: therefore, increasing inspection and maintenance are required in order to guarantee safety and well-being of trains. One of the biggest concerns about transportation safety is broken wheels due to internal cracks. Frequent accidents and regular breakdowns have been registered in the latest years due to this failure.Three testing methods have been investigated to identify cracks or problems on the surface or near the surface of the wheel: Magnetic Particle Inspection, Dye Penetrant Inspection, and eddy current testing. These methods show the following drawbacks: MPI and DPI generates high amount of hazardous waste and none of them is automatable, while the eddy current system is sensitive to lift-off variations and probes need to be positioned at a constant distance. In conclusion a new, more efficient, automated and faster method is needed to detect surface cracks.TRAINWHEELS face the challenge of optimizing thermography in order to use it as an automated train wheel crack detection method, meeting all European Safety Standards. Using thermography for train inspection would result in a big improvement compared to previously quoted methods: thermography is an automated and faster approach; it is as well cleaner and cheaper. At the moment, there is no better way of controlling the train wheel surface than with the use of thermography.Our aim is to improve train circulation safety by enhancing the effectiveness of non-destructive tests on wheels, in order to identify cracks as soon as possible. For this reason, we believe that a coupling between an existing method – the Ultrasound testing – and the proposed one, will suit the purpose of delivering a complete and self-consistent detection method