Hybrid Model for Estimating Remaining Useful Life of SiC MOSFETs

This technical report presents a hybrid model aimed at estimating the remaining useful life (RUL) of silicon carbide (SiC) metal-oxide semiconductor field-effect transistors (MOSFETs) subjected to power cycling experiments. The model focuses on predicting the RUL by utilizing historical data on average junction temperature swings and temperature-compensated drain-source ON-state resistance. The primary failure mode considered is bond-wire lift-off, which is analyzed through a damage accumulation scheme based on Paris’ crack law. The report details the experimental setup, including the power cycling process applied to 25 SiC MOSFETs, and discusses the significance of accurate RUL predictions in enhancing the reliability of power electronic systems. Furthermore, it outlines the potential for developing more sophisticated prognostic schemes using physics-informed neural networks (PINNs) in future work. The findings aim to contribute to the field of Prognostics and Health Management (PHM) for power electronic devices.