Material simulation-based electronic device prognosis

Abstract

Electronic systems, such as power supplies, are complex multilayered devices consisting of different materials with inherent variability. Thermal gradient cycling occurs during system operation which eventually results in thermo-mechanical fatigue induced failure. Such failures can result in immediate electronic system shutdown with no advanced fault or warning signals; thus preventing the use of conventional fault-to-failure detection approaches as a means of predicting maintenance need. This NAVAIR sponsored SBIR project uses state-of-the-art material modeling to predict degradation of circuit board elements as a means for "simulated fault detection." This effort has been focused on the specific aspect of solder fracture and fatigue since electronic industry statistics have attributed this failure issue as a driving factor in system reliability. This project demonstrates feasibility for using conventional, off-the-shelf sensing, to predict solder degradation due to thermal cycling as a means to prognosticate electronic device reliability