Contact mechanics analysis of measured wheel-rail profiles using the finite element method

Abstract A tool has been developed for contact mechanics analysis of the wheel-rail contact. Using measurements of wheel and rail profiles as input, the tool is based on the finite element (FE) code ANSYS. Traditionally, two methods have been used to investigate the rail-wheel contact, namely Hertz's analytical method and Kalker's software program Contact. Both are based on the half-space assumption as well as on a linear-elastic material model. The half-space assumption puts geometrical limitations on the contact. This means that the significant dimensions of the contact area must be small compared with the relative radii of the curvature of each body. Especially in the gauge corner of the rail profile, the half-space assumption is questionable since the contact radius here can be as small as 10 mm. By using the FE method (FEM) the user is not limited by these two assumptions. The profile measurement system Miniprof was used to measure the wheel and rail profiles that were used as input when generating the FE mesh. As a test case, a sharp curve (303 m radius) in a unidirectional commuter train track used by X1 and X10 trains was chosen. The results of two contact cases were compared with the results of the Hertz analytical method and the program Contact. In the first contact case the wheel was in contact with the rail gauge corner. In the second case the wheel was in contact with the rail head. In both contact cases Hertz and Contact presented very similar results for the maximum contact pressure. For the first contact case, a significant difference was found between the FE method and the Hertz method and the program Contact in all of output data. The Hertz and Contact methods both presented a maximum contact pressure that was three times larger (around 3 GPa) than the FE solution. Here, the difference was probably due to the combination of both the half-space assumption and the elastic-plastic material model. For the second contact case, there was no significant difference between the maximum contact pressure results of the three different contact mechanics methods employed.