Utilization of Shaft Voltage Measured From Induction Motor Rolling Bearings for Early Gear Fault Detection

Condition-based maintenance of electric drivetrains relies on effective fault feature revelation. Instead of mounting extra sensors in the drivetrain, the motor itself can be utilized as a built-in fault detector. The shaft voltage of electric motor is essentially coupled with the drivetrain dynamics, thus, can support rotating machinery fault detection. In this work, the theoretical basis and practical performance of the shaft-voltage-based gear fault detection in induction motor drivetrain are studied. First, the shaft voltage signal is analytically modeled, considering common-mode and differential-mode sources, eccentricity and unbalance, rotor slotting effect, and fault excitation. The contained fault modulation signatures are, thus, revealed and further experimentally validated. Second, the practical performance of the shaft-voltage-based early-stage gear fault detection is quantitatively evaluated based on gear run-to-failure test, and six other types of signals are synchronously measured for comparison. Finally, to get rid of the dependency on mounting extra slip rings at shaft ends in the measurement, a novel scheme of shaft voltage measurement is proposed by utilizing the contact and conductive characteristics of the rolling elements in bearings. These results establish a thorough foundation of the shaft-voltage-based gear fault detection, demonstrating its promising merits of both easy accessibility and early-stage fault detection capability.