Digital Twin-Based Lifetime Estimation of SiC Power Modules

This research paper describes the reliability investigation of devices with SiC die on substrates and their lifetime estimation based on digital twin methodology. The most common failure modes that occur in power modules during power cycling and thermal cycling are bond wire degradation, die attach failures or system solder deterioration. The digital twin-based lifetime estimation helps to calculate the lifetime of power devices based on variable input parameters like temperature swing or minimum junction temperature of the chip etc. for example. This method develops a meta model of optimal prognosis file (MOP). It is shown that the inputs can be varied and the corresponding lifetime can be predicted.