Life Prediction of Ground-Wall Insulation Material in Electric Motors Based on Accelerated Degradation Test under Different Failure Mechanisms

The aging and breakdown of ground-wall (GW) insulation in electric motors may lead to serious faults and safety accidents, so predicting the life of GW insulation materials is important. In this article, high temperature accelerated degradation tests are performed on polyimide (PI) film, a kind of GW insulation material widely used in electric motors, to obtain its four key insulation indexes under accelerated stress, including insulation resistance, insulation capacitance, tangent of dielectric loss angle and maximum partial discharge. Experiment results show that a single model cannot accurately simulate the evolution of maximum partial discharge, due to the different failure mechanisms under different temperature stresses. Therefore, a Wiener-inverse Gaussian (IG) fusion model is proposed to predict the maximum partial discharge of PI film, and together with the Arrhenius model, the life of PI film under different temperatures can be predicted. Further experimental verification has been performed, and the accuracy of the proposed fusion model is demonstrated from multiple perspectives by using the methods of Kolmogorov–Smirnov test, Akaike information criterion, and area ratio based on the data from experimental verification. The proposed Wiener-IG fusion model can better fit the degradation data under different failure mechanisms, and then can be used to predict the life of GW insulation material.