Microstructural effects on constitutive and fatigue fracture behavior of TinSilverCopper solder

As microelectronic package construction becomes more diverse and complex, the need for accurate, geometry-independent material constitutive and failure models increases. Evaluations of packages based on accelerated environmental tests (such as accelerated thermal cycling or power cycling) only provide package-dependent reliability information. In addition, extrapolations of such test data to life predictions under field conditions are often empirical. Besides geometry, accelerated environmental test data must account for microstructural factors such as alloy composition or isothermal aging condition, resulting in expensive experimental variation. In this work, displacement-controlled, creep, and fatigue lap shear tests are conducted on specially designed SnAgCu test specimens with microstructures representative to those found in commercial microelectronic packages. The data are used to develop constitutive and fatigue fracture material models capable of describing deformation and fracture behavior for the relevant temperature and strain rate ranges. Furthermore, insight is provided into the microstructural variation of solder joints and the subsequent effect on material behavior. These models are appropriate for application to packages of any geometrical construction. The first focus of the thesis is on Pb-mixed SnAgCu solder alloys. During the transition from Pb-containing solders to Pb-free solders, joints composed of a mixture of SnPb and SnAgCu often result from either mixed assemblies or rework. Three alloys of 1, 5 and 20 weight percent Pb were selected so as to represent reasonable ranges of Pb contamination expected from different 63Sn37Pb components mixed with Sn3.0Ag0.5Cu. Displacement-controlled (constant strain rate) and creep tests were performed at temperatures of 25°C, 75°C, and 125°C using a double lap shear test setup that ensures a nearly homogeneous state of plastic strain at the joint interface. Rate-dependent constitutive models for Pb-contaminated SnAgCu solder alloys ranging from the traditional time-hardening creep model to the viscoplastic Anand model are described. The second focus of the thesis is on fatigue damage accumulation in SnAgCu solder alloys. While, typical fatigue fracture models are empirical, recently a non-empirical model termed Maximum Entropy Fracture Model (MEFM) was proposed. MEFM is a thermodynamically consistent and information theory inspired damage accumulation theory for ductile solids. This model has been validated recently for Sn3.8Ag0.7Cu solder alloy, and uses a single damage accumulation parameter to relate the probability of fracture to accumulated entropic dissipation. Isothermal cycling fatigue tests on Sn3.0Ag0.5Cu and mixed SnPb/Sn3.0Ag0.5Cu solder alloys at varying strain rates and temperatures are conducted using a custom-built microscale mechanical tester capable of submicron displacement resolution. MEFM is applied here in conjunction with the Anand viscoplasticity model to predict the softening occurring over successive cycles as a result of damage accumulation. The damage accumulation parameters for Sn3.0Ag0.5Cu in different aged states are related to a microstructural parameter which quantitatively describes the state of coarsening. In addition, damage accumulation parameters for the three mixed solder alloys are reported. This approach allows for a non-empirical prediction of both constitutive and fracture behavior of packages of different geometries and different microstructural states under thermo-mechanical fatigue. Approaches to solder joint reliability predictions from materials science and mechanics perspectives differ dramatically. Materials science methods identify key failure mechanisms, but most models cannot predict failure. In contrast, mechanics approaches often provide estimates of joint lifetime, but fail to provide insight into microstructural influences. This work attempts to connect the two fields by relating constitutive behavior and fatigue fracture models for different alloys and aging conditions to one or more microstructural parameters