Wheel and Rail Interface Probabilistic Optimization Based on Fatigue Life Prediction

January 11, 2016 in

Phd Research by Yuewei Ma


Advisor(s)
Rolf Dollevoet, Valeri Markine

Period
02/01/2012 - ongoing

Theme(s)


Keywords
fatigue life prediction, optimization, probabilistic CSC, wheel/rail contact,

Funding
China Scholarship Council, TU Delft

Link or Download
Not available

Summary

Due to the harsh working environments, nowadays it is not uncommon to find a multitude of wheel-rail interface damage (including wear, squats, head checks et.al), which greatly limit their serving lives and cause considerable economic loss. Thus quite a lot of attention has been drawn on wheel-rail interface for activities, such as unveiling the damage mechanisms, optimising the wheel-rail interface et al. Over the past decades, even though several numerical procedures for wheel-rail interface optimization have been proposed, few of them can take the uncertainty of railway parameters into account, which will greatly influence the result of wheel-rail interface assessment. Moreover, details of wheel-rail interface damage mechanisms are always ignored or simply considered by engineering formulas.

The goal of our research is to develop a comprehensive and robust optimization procedure for overcoming the problems listed above. The variety of contact geometries, friction coefficients, track parameters as well as vehicle properties will be integrated into the developed optimization model. Meanwhile, the interface serving life will be assessed and further employed as a constrain in the optimization process.

Up to now, a three dimensional (3D) explicit finite element (FE) model for accurate determination of stress and strain fields in W/R contact area has been developed and verified with “CONTACT” method. The model is characterized by accounting the realistic W/R contact geometries, non-linear material properties and dynamic behaviour together. In addition, a novel meshing refinement method, aiming to study the micro-scale stress/strain field on the contact surface, is integrated with the 3D FE model. Preliminary results obtained from the numerical simulations are shown in figures below.