Detailed wheel/rail geometry processing with the conformal contact approach

Abstract

This paper proposes a new way of considering wheel–rail contact in multibody systems simulation that goes beyond the traditional planar constraint and elastic approaches. In this approach, wheel–rail interaction is modelled as a force element with pressures and shear stresses distributed over a contact area that may be curved, supporting conformal contact situations. This by-passes the selection of the contact reference location and reference angle, which are delicate aspects of planar contact approaches. The idea is worked out introducing the curved reference surface as the new backbone for the computations, instead of the tangent plane used previously in planar contact approaches. The steps are described by which the curved reference is constructed in CONTACT, using generic facilities for markers, grids, and coordinate transformations, by which generic wheel/rail configurations can be analyzed in a fully automated way. Numerical results show the capabilities of the new method for measured, worn profiles, suppressing discontinuities in the forces when multiple contact patches split or merge. A further application concerns the evaluation of strategies used in planar contact approaches. There we find that the tangent plane’s inclination is of the biggest importance. This should be defined in an averaged way to achieve maximum correspondence to the more detailed curved contact approach.

Challenges and progress in the understanding and modelling of the wheel–rail creep forces

“…The overall procedure is generalised using a curved reference surface. This by-passes the main subtleties of planar contact approaches and extends to conformal contact configurations [17,63,64]. New perspectives are given also on the creepages ξ , η, φ, after a discussion on their effects on asymmetric contact patches [61,65,66].…”

“…Multibody simulation is used at the scale of wheelsets, vehicles, and trains, e.g. for vehicle dynamics and virtual homologation [19]; boundary element methods are used primarily to assess the stresses as needed for profile evolution [17,20]; finite element methods are used for detailed studies involving plasticity and material damage [21]. A complete mathematical-physical model would honour the interactions between the different scales involved in the problem, such as between the overall flexible deformation of the contacting bodies and the local frictional forces.…”

“…The full theory provided by CONTACT is generally considered too slow for day-to-day work in vehicle dynamics where the overall motion doesn't depend too critically on the creep forces and where millions or billions of contact problems may need to be solved. With improved numerical solvers [51,52], it needs around 25 ms per case for a practical resolution [17]. This is acceptable for specific cases, e.g.…”

“…A new perspective on the overall three-step procedure is given by Vollebregt [16,17], viewing the tangent plane as an approach to manage the complexity of wheel-rail interactions. This tangent plane is introduced such that normal and tangential contact problems can be studied in isolation, supporting modular programming.…”

“…These creepages may be viewed as artificial values connecting the scales of contact mechanics and overall dynamic motion. The true motion of the surfaces is approximated by a linear function, w = [ξ − φy, η + φx] T , which appears sensitive to the reference point that is used for linearisation [17].…”