Modeling the Effect of Microstructural Variability on Solder Joint Lifetime

Abstract

The objective of this thesis was to develop a microstructure-based FE modeling technique to be used in solder joints for semiconductor packaging applications. This technique is capable of generating random solder joint microstructures featuring β − Sn grains and grain boundaries by means of 3D Voronoi tessellations. The anisotropic material behavior of β − Sn grains is described by the Garofalo-Hill creep model, which combines the Garofalo creep equation with the anisotropic Hill equivalent stress definition . The β − Sn grains also feature anisotropic elastic behavior. The grain boundaries are implemented in the FE models as interface elements and they are fitted with a custom-developed constitutive model. This constitutive model combines isotropic creep and isotropic elasticity. The granular microstructure of SAC solder joints is generated via random Voronoi tessellations. The tessellations are used to automatically generate large amounts of unique solder joints with 5 to 9 grains each, with each grain having a random material orientation. The modeling technique was used to qualitatively estimate the stochastic variability that the microstructural differences introduce in the creep response of the solder joints.
Multiple simulation campaigns were performed to analyze this variability on a single solder joint level as well as on a product level. A number of sample products with random combinations of unique solder joints were developed for the latter campaign.