Thermal cycling reliability of aged PBGA assemblies - comparison of Weibull failure data and finite element model predictions

Isothermal aging causes detrimental changes in the microstructure, mechanical response, and failure behavior of lead free solder joints in electronic assemblies. These material changes also degrade the reliability of solder joints in assemblies subjected to aging prior to field use. In the current work, we have extended our previous research on the effects of aging on lead free solder material behavior to explore the effects of prior aging on solder joint (board level) reliability in actual assemblies. Our overall objective was to develop new reliability prediction procedures that incorporate aging effects, and then to validate the new approaches through correlation with thermal cycling accelerated life testing experimental data for pre-aged assemblies. Traditional finite element based predictions for solder joint reliability during thermal cycling accelerated life testing are based on solder constitutive equations (e.g. Anand viscoplastic model) and failure models (e.g. energy dissipation per cycle model) that do not evolve with material aging. Thus, there will be significant errors in the calculations with lead free SAC alloys that illustrate dramatic aging phenomena. This work has implemented a theoretical framework for correcting this limitation and including aging effects in the reliability modeling. The developed approach involved the use of: (1) a revised set off Anand viscoplastic stress-strain relations for solder that included material parameters that evolve with the thermal history of the solder material, and (2) a revised solder joint failure criterion that included aging effects. The effects of aging on the nine Anand model parameters were determined experimentally for SAC305 lead free solder as a function of aging temperature and aging time. The revised Anand constitutive equations for solder with aging effects were then incorporated into standard finite element codes. The applied aging-aware failure criterion was based on the Morrow-Darveaux (dissipated energy based, DeltaW) approach, with both the fatigue criterion for crack initiation and the crack growth law incorporating material constants that depend on the prior aging of the solder material. Fatigue data were measured for SAC solder using uniaxial and shear test specimens that were aged for various durations and temperatures prior to cycling. The developed reliability modeling procedure has been applied to a family of assembled PBGA components. In the simulations, the packages were subjected to isothermal aging followed by thermal cycling accelerated life testing. The model predictions were correlated with solder joint reliability test data for the same components. The experimental test vehicle incorporated several sizes (5, 10, 15, 19 mm) of BGA daisy chain components with 0.4 and 0.8 mm solder joint pitches (SAC305). PCB test boards with 3 different surface finishes (ImAg, ENIG and ENEPIG) were utilized. Before thermal cycling began, the assembled test boards were divided up into test groups that were subjected to several sets of aging conditions (preconditioning) including 0, 6, and 12 months aging at T = 125 oC. After aging, the assemblies were subjected to thermal cycling (-40 to +125 oC) until failure occurred. The failure data for each test group were fit with the two parameter Weibull model, and the failure plots have demonstrated that the thermal cycling reliabilities of pre-aged assemblies were significantly less than those of analogous non-aged assemblies with degradations of up to 53% for one year of prior aging. Finite element modeling using the modified Anand model for solder was performed for the four different components sizes to predict the stress-strain histories of both non-aged PBGA assemblies and aged assemblies that had been subjected to constant temperature exposures for various times before being subjected to thermal cycling. The plastic work (DeltaW) per cycle results from the finite element calculations were then combined with the aging aware fatigue and crack growth models to estimate the reliability (cycles to failure) for the aged and non-aged assemblies. As expected, the predictions showed significant degradations in the solder joint life for assemblies that had been pre-aged before accelerated life testing. The coefficients in the aging aware crack growth model were selected to reflect the board surface finish and SAC solder combination. With this approach, good correlation was obtained between the new reliability modeling procedure that includes aging and the entire set of measured solder joint reliability data that includes multiple component sizes, prior aging conditions, and board surface finishes.