Simulation of Solder Fatigue Effects on Typical BGA Package due to Material and Temperature Variations

The past two decades have seen many approaches to solder fatigue and solder joint life published. This subject has proved difficult as various failure mechanisms are proposed and examined. While these theoretical bases are discussed, it often leaves the end developer in a difficult situation as to how to apply an accurate simulation approach to fatigue failure for a particular package. In this paper, one approach advocated by Syed, Sharon, and Darveaux [1] is examined and used to evaluate a “typical” metal lid flip chip BGA (ball grid array) package under variation of a few key parameters under the control of a project via a 2-level DOE (Design of Experiment). The parameters examined in this paper are the in-plane thermal expansion coefficient of the PCB board (a material choice), the lid material (also a material choice), and the temperature range for the BGA environment (a boundary condition and a heat transfer effectiveness choice). Most other parameters that could also affect solder fatigue life are not within areas that are easily changed due to standards, manufacturing methods, cost, or regulations, and hence are not considered. The DOE results show that all three of these individual parameters are significant to the DOE models, but for solder bumps there is an interaction term of the lid CTE (coefficient of thermal expansion) and temperature ΔT, while the solder balls show a significant interaction of the PCB CTE and the temperature ΔT. Bump life cycles to failure ranged from approximately 35,000 to 24 million, while ball life cycles to failure ranged from 48,000 to 24 million.