A simulation-based multi-objective design optimization of electronic packages under thermal cycling and bending

Abstract

In this study, a simulation-based multi-objective design optimization methodology was developed for improving electronic packaging reliability. It was demonstrated using a generic model of an electronic package on a printed wiring board. The objective for the optimization was to improve the reliability of solder joints under both thermal cycling and bending by optimizing a group of design parameters. A parametric finite element model was developed using ANSYS for both load conditions. To improve the numerical efficiency of the optimization, a multi-quadric response surface method was implemented to approximate the response of finite element simulations for each loading condition. Subsequently, the multi-objective optimization of solder joint reliability was implemented using a Minmax principle on all response surfaces and a differential evolution algorithm as optimal search engine, which is capable of finding global minimum when local minima exist. Our study demonstrated that the reliability of the solder joints is significantly improved for this given generic model of electronic package. The proposed methodology can be effectively used in improving the reliability of electronic packages.