Self-powered fault diagnosis of rolling bearings based on triboelectric effect

In this study, self-powered fault diagnosis of rolling bearing (SP-FDRB) is conducted based on the triboelectric effect. Flexible interdigital electrodes are glued to the outer ring of a rolling bearing to form a rolling-type free standing mode triboelectric nanogenerator (RF-TENG). The RF-TENG has adequate service life because it is designed to avoid direct contact between the flexible electrode and the rolling element and maintain the structural integrity of the bearing. A prototype of the RF-TENG structure is fabricated and the power generation mechanism of the RF-TENG is analyzed and verified by electrostatic field simulation. The electrical output test of the RF-TENG is conducted. The variations in the output current amplitude and frequency with change in the speed of rotation are analyzed, and the maximum output power of the RF-TENG corresponding to the optimal load is obtained accordingly. The reliability and bearing capacity of the RF-TENG are illustrated through long-term continuous operation and various bearing load tests. Based on the output current signal model, the fault characteristic frequency of the RF-TENG output current is obtained when the outer ring, inner ring, and rolling ball have localized faults. SP-FDRB is performed based on the RF-TENG output current signal by utilizing several machine learning algorithms. The results show that the classification accuracy can exceed 92%, which is equivalent to the accuracy based on the vibration signal. The proposed RF-TENG based SP-FDRB is feasible and has excellent application prospect. • SP-FDRB is conducted based on the triboelectric effect. • A rolling-type free standing mode triboelectric nanogenerator is fabricated. • Variations in output current amplitude and frequency with speed are analyzed. • SP-FDRB is performed based on RF-TENG output current by utilizing ML algorithms. • The RF-TENG based SP-FDRB is feasible and has excellent application prospect.