Abstract: Aiming at the dynamic changes in wheel profiles, the optimal rail top/side Friction Coefficient Combinations (FCCs) for wheels with worn flange and worn tread profiles in different service stages were explored separately.Wheel profile wear data was tracked along two subway lines as the distance traveled increased.A dynamic model of subway vehicle was constructed to analyze the effects of rail top/side FCCs on wheel-rail wear properties,attack angle of wheelset,surface fatigue index and derailment coefficient,with the dynamic changes in wheel profiles taken into account. Then a multi-objective optimization model was developed based on the entropy-weighted TOPSIS method,which includes the four performance indicators mentioned above as the criteria. Finally,the optimal FCCs were obtained.The results suggest that as wheel profiles dynamically change,the wheel-rail contact points show sudden changes in position and tend to concentrate to varying degrees,resulting in poorer match between wheels and rails.The comprehensive performance of worn flange profiles is affected primarily by rail top friction coefficient during initial operation. As the distance traveled increases,the rail side friction coefficient predominates.The comprehensive performance of worn tread profiles is influenced mainly by rail top friction coefficient in all stages of wheel service.During the stages of initial operation,early wear,middle wear,and late wear,the optimal rail top/side FCCs for wheels with a worn flange profile are 0.30/0.10,0.25/0.10,0.20/0.10,and 0.20/0.10,respectively. For wheels with a worn tread profile,these are 0.20/0.15,0.20/0.30,0.20/0.35,and 0.30/0.15,respectively.

Key words: rail top/side friction coefficient combination, worn flange profile, worn tread profile, entropy-weighted TOPSIS method