Bearing degradation dynamic modelling and fault analysis for a reusable liquid rocket engine turbopump

The scope of this study is to develop an advanced dynamic model for ball bearings, with a specific focus on simulating the vibration behavior of bearings in reusable liquid rocket engine turbopumps under conditions of surface degradation. The methods employed underwent significant enhancements, including the integration of various degradation profiles. Detailed analysis of vibration signals under different defect types, such as outer and inner race, or ball defects, was conducted. Additionally, more realistic defect shapes, including piecewise and half sine, were introduced. The study also explored scenarios involving multiple defected bearings, covering both single and combined types of defects. Finally, the influence of defect size was examined under both stationary and non-stationary defect profiles. Following the generation of vibration signals, a signal processing method was proposed with the main aim of detecting defect types in combined defected bearings, as well as diagnosing defect size and number. The results confirm the capability of the designed 5 Degree of Freedom nonlinear system model to generate vibration signals under various conditions, including stationary and non-stationary profiles. The efficiency of the proposed signal processing method was demonstrated through synthetic data. This comprehensive approach provides valuable insights, enhancing the understanding of bearing behavior under degradation conditions. In the realm of reusable rocket propulsion, the improved signals are crucial for training predictive algorithms aimed at estimating the Remaining Useful Life of turbopumps bearings. Additionally, the proposed method contributes to the advancement of bearing degradation diagnosis, ultimately promoting the performance of aerospace systems.