Effects of varying amplitudes on the fatigue life of lead free solder joints

Realistic manufacturing, test, handling and service conditions tend to involve significant variations in loading of microelectronics assemblies. However, assessments of reliability are commonly based on accelerated cycling tests with fixed amplitudes and effects of variations are estimated assuming a damage accumulation rule such as Miner's rule. These rules may overestimate the life of lead free solder joints under realistic conditions by orders of magnitude. The severity of such errors is not likely to be captured in accelerated testing with combinations of amplitudes selected at random or based on convenience. Brief exposure to one or more high strain rate loading cycles may significantly reduce the effective stiffness of typical lead free solder joints in subsequent cycling at lower amplitudes and strain rates. This may greatly complicate modeling and generalization of test results whether failure occurs through the solder or the intermetallic bond to the contact pad, or by pad cratering. Systematic studies of individual SAC305 joints in isothermal fatigue led to an approach to the determination of `worst case' scenarios and the prediction of effects of amplitude variations on solder failure. Preliminary results on lower-Ag alloys and the MaxRel™ alloy are presented as well.