Bearing degradation indicator using cage speed: Extension of BeaFEM to nonstationary operating conditions

The well-known bearing characteristic frequencies (BPFO, BPFI, FTF and BSF) reflect the complex kinematics of operating bearings. Far from the idealised perfect rolling assumptions, the experimental frequencies deviate from their theoretical values. The deterioration of the bearing health is likely to induce changes in the cage rotating speed, materialised by a shift of the angle characteristic frequencies. Monitoring the cage speed using these frequencies can provide a valuable degradation indicator. However, nonstationary operating conditions are also known to influence the bearing motion. This work is a continuation of BeaFEM introduced by Bertoni and André (2023), and assess its potential for monitoring bearings operating under nonstationary conditions. To this end, BeaFEM is reformulated with an original interpretation using cage speed and a short-term adaptation is proposed. The possible causes of cage slip are reminded, and the influence of load is studied on two wind turbine bearings. Based on this knowledge, the influence of operating conditions is mitigated and a strategy to use the cage speed as a valuable degradation indicator is proposed. • A bearing degradation indicator based on cage speed is presented. • The cage slip is estimated from the frequency localisation of fault-related peaks in a spectrum. • The method is designed to process vibration signals operating under variable speed. • The influence of other confounding causes for slippage is studied and mitigated. • The use of cage slip as a bearing degradation indicator is tested on two industrial cases.