Accelerated aging experiments for capacitor health monitoring and prognostics

This paper discusses experimental setups for health monitoring and prognostics of electrolytic capacitors under nominal operation and accelerated aging conditions. Electrolytic capacitors have higher failure rates than other components in electronic systems like power drives, power converters etc. Our current work focuses on developing first-principles-based degradation models for electrolytic capacitors under varying electrical and thermal stress conditions. Prognostics and health management for electronic systems aims to predict the onset of faults, study causes for system degradation, and accurately compute remaining useful life. Accelerated life test methods are often used in prognostics research as a way to model multiple causes and assess the effects of the degradation process through time. It also allows for the identification and study of different failure mechanisms and their relationships under different operating conditions. Experiments are designed for aging of the capacitors such that the degradation pattern induced by the aging can be monitored and analyzed. Experimental setups and data collection methods are presented to demonstrate this approach.