Numerical assessment of the influence of worn wheel tread geometry on rail and wheel deterioration

Current maintenance criteria of railway wheel geometries consider mainly flange thickness, flange height, flange gradient, and depth of tread wear. These geometrical quantities are relatively simple to evaluate, but do not fully describe neither the worn profile nor its influence on subsequent deterioration of wheel and rail. With increased use of automatic wheel profile measurements there is a potential to plan maintenance from measures that provide a more elaborate description of the wheel profile and its operational capabilities. To this end, the paper first outlines a procedure to parametrise worn wheel profiles. This procedure is linked to multibody simulations of a curve negotiating freight wagon to identify parameters of the worn wheel profile with the largest influence on rolling contact fatigue (RCF) and wear. To limit the number of parameter combinations, nearly orthogonal and space filling Latin hypercube sampling is employed. Regression analysis is then utilised to derive simple relationships (meta-models) between geometry parameters and deterioration measures. The derived meta-models of RCF and wear deterioration are shown to produce good agreement to results from fully fledged multibody simulations. The end result is identified geometrical quantities that provide an improved categorisation of the deteriorated wheel tread geometry for improved maintenance decisions.