A Statistical Prediction of Electrical Discharge Initiation and Semi-Analytical Transient Mixed Lubrication Model of a Rolling Element

Abstract Rolling element bearings are an integral component of electric vehicles, supporting radial and axial loads in powertrain components such as electric motor shafts and wheel bearings. Fast-switching inverters enable precise, variable-control of motor performance at the cost of possible stray current leakage into mechanical components. These currents naturally seek to cross the insulating fluid film in rolling element bearings. In doing so, a destructive arc may form and cause irreversible damage to metallic bearing surfaces. A unique contribution of the work is that it provides a method to use the statistical height distribution to predict the likelihood of electrical breakdown and arcing. To predict film thickness it uses a closed-form elasto-hydrodynamic lubrication (EHL) models to present a semi-analytical model of this arcing phenomenon. Existing EHL models are modified for mixed lubrication and electrical contacts by incorporating a solid rough surface asperity contact model and a flow factor modified lubrication model. The model accounts for transient effects and considers changes in speed and other parameters during operation. The resulting model predicts the likelihood of surface damage and electrical properties of the bearings through the statistical asperity height above a critical value calculation. The damaged regions predicted by the model are in qualitative agreement with the experimental tests.