Wheel bearings are critical components for the operational efficiency and safety of railway and on-road vehicles. With the continuous advancement of condition monitoring technology, vibration signal-based fault diagnosis methods have become an effective approach for detecting early faults in wheel bearings. However, the vibration response of wheel bearings typically exhibits strong non-stationary characteristics and is highly susceptible to interference from mechanical and environmental noise, making it difficult to identify fault features accurately. To address this issue, this study focuses on exploring the emerging On-Rotor Sensing (ORS) technology for more accurately capturing and processing the non-stationary vibration responses from wheel bearings. The ORS technology consists mainly of a triaxial micro-electro-mechanical system (MEMS) accelerometer, micro-processor and wireless module that can be mounted to a rotating shaft to obtain high quality signals allowing high accuracy monitoring of the bearing supporting with inherent instantaneous speed estimation and order-tracking analysis. In particular, the study assesses the ORS's usability and the corresponding data analysis methods for ultimately monitoring the axel box bearings of railway vehicles The study firstly establishes a 6 degree-of-freedom (DOF) wheel bearing dynamic model that particularly incorporates internal bearing clearances and external excitations. Numerical simulations are then conducted to gain the vibration characteristics of the bearing under different operating and fault conditions, especially, the characteristics can be increased by clearances but weaken by the irregular excitations. Subsequently, the ORS accelerometer is configured to fix on the shaft end of a full-scale axle bearing in a dedicated inhouse test platform for evaluating the performance of bearing monitoring and diagnosis. The monitoring and diagnostic results shows the ORS sensor outperforms conventional sensing methods in distinguish changes in bearing clearances and diagnosing common bearings such as defects in outer race, inner race and rolling elements when the system operates under stationary conditions. Finally, for evaluating the technology in practical applications with nonstationary operations, the ORS sensor is installed on an automotive drive shaft to collect vibration response data from the wheel bearing. Based on the orthogonal outputs of ORS the instantaneous wheel speed is calculated and order tacking spectrum is subsequently obtained to show the characteristic amplitude at ball-pass-frequency-of outer race (BPFO) for bearing wear assessment along with the amplitudes at shaft orders for tyre wears and wheel imbalances, and providing strong technical support for bearing condition monitoring and fault diagnosis.
| Date of Award | 16 Sept 2025 |
|---|
| Original language | English |
|---|
| Sponsors | Beijing Institute of Technology |
|---|
| Supervisor | Fengshou Gu (Main Supervisor) & Helen Miao (Co-Supervisor) |
|---|