Mei-Ling Wu1, Elviz George2 and Michael Pecht2,3
1MOEMS Reliability Analysis Lab, National Sun Yat-Sen University, Kaohsiung, Taiwan
2Center for Advanced Life Cycle Engineering (CALCE), University of Maryland, College Park, MD-20742, USA
3Prognostics and Health Management Center, City University of Hong Kong
Abstract:
A computerized physics of failure (PoF) based approach to provide rapid estimation of the life expectancy of electronic products is introduced in this paper. Estimation of the life expectancy of electronic products assists equipment manufacturers in determining warranty periods, maintenance schedules, and repair and replacement times. In addition, products are required to meet the reliability and life cycle requirements expected by the customer. Due to increasing competition and technological obsolescence, equipment manufacturers are required to reduce the time to market of new products in a cost-effective manner. Traditional methods of product qualification that involve expensive and time-consuming physical tests can be replaced by computer simulation software that replicates physical tests. These computer simulation programs must take into consideration the responses of a product to anticipated use and test conditions. Based on the responses, failure sites and mechanisms should be identified and time to failure should be estimated. Failure sites and mechanisms in electronic assemblies and commonly used PoF models are briefly described. Simulation-assisted reliability assessment is demonstrated for an electronic assembly module used in communications hardware. The assembly is modelled based on inputs derived from datasheets, material property databases, and experimental measurements of physical properties. The simulation recreates long-haul aircraft loading conditions. Ranking of failure mechanisms reveals that fatigue due to temperature cycling and vibration is the most critical failure mechanism. Components that do not meet the qualification requirements under these loading conditions are identified during the design phase before physical prototype testing, thereby avoiding expensive redesign. The acceleration factor between field and test conditions is determined and can be used to increase the confidence in field life estimations. The life expectancy of the assembly under temperature cycling and vibration loading conditions was also determined.
Keywords: Electronics reliability, physics of failure, simulation, virtual qualification.
Complete article available to CALCE Consortium Members.