Multi-factor model for lifetime prediction of polymers used as insulation material in high frequency electrical equipment

Abstract

High frequency (HF) electrical equipment fed by power electronic converters, such as HF transformer and rotating machine, are subjected to intense electrical and thermal stresses. For research on aging behavior of polymers under the influence of multi-factors, quantitative characterization of the coupling effect between factors is always a challenge. Therefore, a design of experiment (DOE) method has been used to conduct accelerated aging tests on polyimide (PI) films under different conditions of voltage frequency, amplitude and temperature. The influence of these three factors on insulation life of PI film is investigated in this paper, along with the interactions between them. Based on the investigation, a generalized multi-factor model for lifetime prediction is proposed, which is validated by conducting additional experiments. The results show that an inverse power law exists between insulation life and electrical stress (voltage frequency and amplitude), and an exponential function can describe well the effects of temperature on lifetime. Influence of coefficients of frequency and amplitude are very similar and are much larger than that of temperature. In addition, voltage amplitude and temperature have the largest coupling effect, while frequency and temperature have the least. Reliability verification of the lifetime prediction model proves the effectiveness of DOE method in studying aging behavior or other properties of polymers under the synergistic effect of multiple factors or stresses.