Abstract
The effect of different operational loading scenarios on predicted crack growth direction for a propagating inclined railhead crack is assessed by 2D finite element simulations. Studied load scenarios include a moving Hertzian contact load, a temperature drop, rail bending due to a passing wheelset, and combinations thereof. The direction of the unbiased crack propagation is predicted using an accumulative vector crack tip displacement criterion. The numerical model is validated for the individual load scenarios. Restraints due to crack face locking are imposed by a threshold parameter, whose influence is also assessed. For combinations of thermal and contact loads, the predicted crack path is found to diverge gradually from transverse growth, corresponding to pure thermal loading, to shallow growth, corresponding to a pure contact load. For combined bending and contact loading, there is a discrete jump in the predicted crack direction as the contact load increased while the bending load is kept constant. These results are well aligned with empirical experience.
Highlights
Rolling contact fatigue (RCF) of railway wheels and rails is a pervasive, costly and complex phenomenon [1,2]
In order to evaluate the effect of mesh size on crack path predictions, a sensitivity analysis was performed on three different FEdiscretisations for bending load analyses
The predictions have been validated against analyses featuring a previously validated in-house code [12] for a moving Hertzian contact load, and qualitatively validated for the thermal and bending loads; for a purely tensile load cycle, the crack deviated into mode I growth as expected
Summary
Rolling contact fatigue (RCF) of railway wheels and rails is a pervasive, costly and complex phenomenon [1,2]. RCF crack predictions are complicated by the fact that (frictional) rolling contact conditions cause a non-proportional multiaxial stress/strain. The Vector Crack Tip Displacement (VCTD) criterion yielded promising results in four fatigue tests simulated using a linear elastic material model [9] It was shown in [9] that modelling the cyclic elastic–plastic material response does not improve the accuracy of the predicted crack growth directions by VCTD. To be able to identify the operational conditions that cause transverse crack growth, a previously developed crack growth direction criterion in [9], expanded from [11] to account for non-proportional loading, is utilised in the current study. The novelty of the research lies in the introduction of unbiased crack propagation, the analysis of crack growth under combined loading, the introduction of a restraint parameter to account for crack face locking and the assessment on its influence on the prediction
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have