High-speed train wheel set bearing analysis: Practical approach to maintenance between end of life and useful life extension assessment

Abstract

Operational safety and reliability in high-speed trains of wheelset bearings are susceptible to wear and damage due to harsh conditions and factors such as lubrication, load variations, high-speed operation, and thermal stresses. Existing predictive maintenance (PdM) strategies often lack the predictive capabilities with a decision support system that is needed to address potential failures effectively. This paper presents Health and Safety, a dynamic risk assessment framework for PdM that integrates real-time health monitoring by using the multiparametric sensory data of vibration, acoustic emission, and speed signals to assess the condition of wheelset bearings. The proposed methodology integrates a vision transformer with a fuzzy logic-based Health and Safety Risk Assessment Matrix (HSRAM) to forecast and identify the health status of bearings based on their degradation stages and predict the remaining useful life, extended useful life, and end-of-life of wheelset bearings. The key findings show that the proposed framework can accurately quantify bearing degradation, enabling timely maintenance interventions and reducing the risk of catastrophic failures. The proposed model outperforms the Paris, Exponential, and Improved Exponential Models by an average of 27.53 %, 18.98 %, and 15.14 %, respectively, which provides a reliable decision-support tool for PdM in high-speed rail applications. The proposed model advances from conventional residual life prediction by determining whether the used life of the bearing can be ended or extended from one stage to another till the next scheduled maintenance under real-time monitoring, which offers a practical approach to bearing health & safety assessment.

Keywords

• High-speed trains
• Wheelset bearing faults
• Dynamic risk assessment
• Health monitoring
• Extended useful life
• End of life