Influence of the quality of material models on warpage and lifetime prediction by finite element simulation

Abstract

The ever-increasing demand towards sustainability and reduction of the ecological footprint results, as a direct impact on the electronics industry, in reliability simulation and lifetime prediction of printed circuit boards (PCBs) and electronic modules becoming more and more inevitable. Accurate prediction, however, requires sophisticated material models of all base materials that are capable of representing the actual material behavior under the given loading situations during manufacturing and application. This study addresses the highly problematic lack of available material data of dielectric and conductive base materials commonly used in PCB manufacturing and its influence on warpage and reliability simulation. An M2X communication module PCB comprising coupons for reliability testing is used to compare measured data to simulated results obtained by using material models with different degrees of comprehensiveness, beginning from data sheet values provided by material suppliers to the fully characterized temperature-dependent and orthotropic material behavior of fiber-reinforced dielectrics, resins and copper. Depending on the quality of the implemented material data, significant differences were observed when evaluating the simulated results. Comparison to the measured warpage of the module and the lifetime determined by hot oil temperature cycle testing makes it obvious that insufficient material data lead to inconclusive results or inaccurate predictions, with, if being nevertheless relied on, potentially serious consequences during the application of the real products. Based on these findings, suggestions for best-practice material models of dielectrics and copper are presented, as it could be proven that the comprehensive understanding of material behavior is the key to reliable warpage and lifetime predictions of PCBs and electronic modules by simulation.