A methodology was developed to correlate field life with observed degradation for electronics modules. This procedure was developed by identifying common deterioration characteristics in field units, modeling observed trends and then developing a model to predict future deterioration trends. This particular method focused on the deterioration of solder joint strength due to solder fatigue and comparing these values with a threshold based on known electrical failures. A conditional probability density function was formulated and quantified for both random shear strength and the minimum shear strength within a module. The conditional probability density function characterized both the changing mean and variance for a normally distributed random shear strength. With this methodology, time or mileage (life) prediction is based on the probability that the minimum performance response is less than a defined failure threshold. The methodology described herein promises to be an effective product development tool as the effect of design changes on product life can be more quickly and easily evaluated. While the technique developed herein is applicable to all electronic designs, the method was developed with a particular focus on understanding these relationships for automotive (harsh environment) applications.

No abstract

No abstract