Railway Wheel Flats. Martensite Formation, Residual Stresses, and Crack Propagation

Doctoral thesis, 1998

A railway wheel flat is a flat spot on the rolling surface of a wheel caused by its unintentional sliding on the rail. The reason for the sliding may be poorly adjusted, frozen or defect brakes, or too high braking forces in relation to the available wheel/rail adhesion. Contaminations on the rail such as leaves, grease, frost and snow aggravate the problem. The temperature rise caused by the sliding, and the rapid cooling into the adjacent material when the wheel starts rolling again, may lead to formation of brittle martensite around the flat. The thermal impact and the phase transformations will generate large residual stresses. The residual stress field will interact with the rolling contact stresses and it will promote the formation and growth of cracks. Another consequence of wheel flats on a rolling wheelset is impact loading on bearings and suspensions and on rails, sleepers, bridges, etc. A numerical model for the prediction of martensite formation is proposed. The model is based on a commercial FE-code, enhanced by subroutines to handle the phase transformations and describe the thermal load. Features and limitations of the model are examined. A parametric study is performed in order to determine those combinations of thermal load and brake locking time which lead to considerable martensite formation. The effect of modifications of material parameters is investigated. A constitutive model is developed for the calculation of stresses in a material going through phase transformations. Transformation plasticity and plastic hardening memory loss during phase transformations are handled. The model is implemented in a commercial FE-code. Simulations indicate large residual stresses beneath the martensitic area of the wheel flat. Extensive full-scale field experiments have been designed, performed and evaluated. More than 200 wheel flats were formed under controlled conditions involving different wheel loads, train speeds and sliding durations. Also the friction coefficient between the wheel and the rail was varied (and indirectly measured). Samples extracted from the tested wheels were metallographically examined with respect to phase transformations and cracks around the flat. The numerical model for prediction of martensite formation was qualitatively verified and quantitatively calibrated. In the experiments, martensite was found beneath all flats, and cracks were observed in most cases. It is concluded that the risk for future spalling in the wheel tread should be considered for all wheelsets with flats. A damaged wheelset should be taken out of service as quickly as possible. When reprofiling the wheels, all martensite and also an additional layer of several millimeters should be machined off.