AN EXPERIMENTAL METHOD TO INVESTIGATE TRUE STRESS DURING THE FATIGUE OF LEAD-FREE SOLDERS

摘要

Measurement of stress in fatigue experiments fails to account for reducing solder joint area. Here we look at measuring lead-free solder joint cross-sectional area realtime and in-situ. There is an urgent need to predict lead-free solder joint reliability (I.e. life-time) under a wide range of service conditions. Hence for robust high reliability electronic designs it is necessary to develop a model that captures actual material properties at the relevant scale, and predicts life-time. This model must include certain degradation mechanisms, such as structural cracking, joint integrity, and the allowable damage up to joint failure. This maximum allowable limit is often expressed as an ultimate true stress within a solder joint, before the onset of mechanical and electrical failure. Techniques are developed to identify these true stresses during fatigue as a function of temperature and strain. The experimental method described uses a precise real-time resistance measurement during mechanical cyclic fatigue testing to estimate the degree of cracking in a solder joint. The experimental solder joint under test has typical assembly dimensions. Crack growth in this standard solder joint specimen is confirmed by microscopy, both optical and ultrasonic. Knowing how the true stress develops and the actual values will significantly improve reliability predictions. Furthermore FEA models that include degradation can be refined to replicate the observed behaviour and hence improve life-time modelling in a wider sense. The recent data show that true stress values for fatigued lead-free solder joints are reaching well above 30 Mpa which is significantly more when compared to estimated nominal stresses based on unaffected solder joint cross-sectional area.