Probabilistic modeling of solder joint thermal fatigue with Bayesian method

Solder joint thermal fatigue failure is a major cause for the failure of the electronic packaging. It is influenced by the device geometry, material fatigue properties, temperature stress and other parameters. All of these parameters contain uncertainties, whereas Coffin-Manson model which is widely used to evaluate fatigue life does not take uncertainties of model parameters into consideration. In this paper, a probabilistic physics of failure (PPoF) of solder joint thermal fatigue using Bayesian theory to update parameters is put forward. Comprehensively considering the influences of uncertainties of all parameters on solder joint failure, and using Monte Carlo method to solve PPoF model, probability density function for single point failure can be obtained. Through establishing the relationship between the probability of failure and time, stress, structure and materials, this method provides a new way for reliability prediction.