Research On The Maintenance Strategy For The Wheel Of High-Speed EMU

The maintenance procedures applied for the wheel of High-speed EMU nowadays are formulated by combining the predict repair with the preventive maintenance, which regards the safety as the top priority. Using the maintenance methods and experiences of the advanced Western countries for reference, it’s aimed at extending the expected lifetime of wheel and enhancing its safety and reliability in service. However, it may not adapt to the situations of the high-speed train in China due to different operating and maintenance conditions compared with others. Therefore, the method of determining an optimal maintenance, which is able to satisfy the requirements of safety and reliability in China, has become a significant issue.The main works of the thesis is as follows:Firstly, the ultrasonic detection defect data as well as the shape detection data are obtained during the study on two failure modes of the wheel of the high-speed EMU. According to the defect data in the different vehicle types and in the different wheel parts, the impact of the vehicle type on the occurrence of the extreme defects is analyzed based on the Bayesian survival analysis theory. As a result, the vehicle type affects less on the driving distance before getting some extreme defect.Secondly, based on the Semi-Markov theory, the degradation process of the wheel is investigated, and the multi-state degradation model of each part and the whole wheel’s multi-state degradation model are proposed. Then, the K-S testing is applied to verify the Weibull distribution assumption and the rationality of grouping states for the conditions of each part. As a result, under the each running distance point, the state probabilities of each part, the ones of the wheel, and the wheel reliability function are deduced and are calculated by Matlab.Thirdly, in order to obtain the situation of the wheel abrasion, the statistical analysis is carried out on the wheel shape detection data, which includes the wheel diameter, the flange height and the flange thickness. The daily wear rates as well as the average losses after the planned profiling are also statistically calculated.Lastly, the determine methods of the ultrasonic defection period are discussed by taking into account respectively the reliability demand, the no extreme defect criteria, and the economic requirement. Then on the consideration of the two failure modes, a wheel maintenance model based on the Monte-Carlo simulation concept is provided in order to find out the expected lifetime and the number of profiling under different maintenance periods. In conclusion, the ultrasonic defection and the planned profiling are proposed to be execute respectively every 210000 km and 250000 km in order to maximize the wheel service lifetime. The service lifetime is supposed to be 4252156 km, and the number of profiling is about 17.02, which lengthen the wheel service life, decrease the new wheel changes as well as save the maintenance cost.