Comparative Study on Non-Hertzian Rolling Contact Models Considering Yaw Angle
To investigate the influence of a yaw angle of wheelset on wheel-rail contact behaviors and assess the accuracy of fast non-Hertzian wheel-rail contact methods,three fast non-Hertzian methods considering the yaw angle were chosen:extended Kik-Piotrowski method(EKP),modified Kik-Piotrowski method(MKP)and modified semi-Hertzian method(MSHM)to analyze the normal and tangential contact behaviors between the wheel and rail.Based on the research ob-jects of S1002CN wheel tread and CHN60 rail,and the reference results calculated by exact theory CONTACT,the ac-curacy of microscopic contact results including the shape of the contact patch,contact stress and shear stress calculated by fast non-Hertzian methods was compared.The vehicle-track coupled dynamic model was established by UM software to obtain the wheel-rail motion parameters in the simulation,as the input parameters of the wheel-rail contact models to calculate the wheel rail contact solution.Based on cumulative error method in statistics,the calculation accuracy and sta-bility of different contact models were assessed under actual vehicle operating conditions.The results show that a yaw an-gle can result in the asymmetry of the contact patch and contact stress.For normal contact problems,the calculation ac-curacy of MKP is higher than that of EKP and MSHM in most contact conditions.However,when the curvatures between the wheel and rail contact patch vary significantly,in the case of a lateral displacement of 6 mm,the relative error of the maximum pressure calculated by MKP reaches 47％.Under this contact condition,MSHM has better estimation about the pressure,with relative error of 27％.For tangential contact problem,EKP+FaStrip and MKP+FaStrip methods have bet-ter accuracy than MSHM+FASTSIM.For dynamic simulation,MKP+FaStrip model has higher calculation accuracy and stability,because both EKP+FaStrip and MSHM+FASTSIM consider the wheel-rail curvatures which are usually discon-tinuous.
万方数据
维普
