Wear of wheels and rails is a problem of substantial magnitude for railways throughout the world. Efforts to control wear by using new wheels profiled to match worn wheels have met with a measure of success. Purely empirical approaches to this problem require years of effort before results can be assessed. This paper describes a computational method for predicting worn profiles. The method is applied here to freight cars running on tangent track, and it includes the capability to handle both stable and hunting running. The method has been evaluated with four different wear models, and it was found that the predicted worn wheel profiles are relatively insensitive to the selection of wear model. Results include predictions from AAR and CNA wheels running on new rail, and from a simulated unit-train operation.

This content is only available via PDF.
You do not currently have access to this content.