Comparison of FEA Modeling Techniques for Plastic Ball Grid Array Assemblies

The reliability study of solder joints is one of the primary issues in lead free electronic assemblies, and life testing of electronic assemblies is often performed using time consuming thermal cycling life tests. Finite element analysis often provides a more time efficient way to obtain and understanding of the behavior of the assembly solder joints if the material properties have been accurately characterized. Lead free solder joints typically experience yielding and highly nonlinear material behavior when subject to thermal cycling, and the Anand viscoplastic model is widely used to describe the nonlinear behavior of lead free solders. Predictions of the FEA models are typically combined with energy-based fatigue models to obtain reliability predictions. In this paper, we have compared several different FEA approaches for Plastic Ball Grid Array (PBGA) assemblies to compare their advantages and disadvantages, especially in the areas of simulation accuracy and efficiency. The examined approaches included different types of 2D and 3D models (2D plane strain, 3D slice, and 3D quarter model), use of Multi-Point Constraints (MPCs), and use of Submodeling. For the various approaches, models were developed for an example PBGA assembly with 15 × 15 mm body size and 324 area array lead free solder joints on a 0.8 mm pitch. For each model type, the predicted plastic work per cycles was calculated in the critical solder joint, and suggestions were made to improve simulation accuracy and reduce simulation time.