Creep–fatigue lifetime estimation of SnAgCu solder joints using an artificial neural network approach

Abstract

Reliability assessment of solder joints under thermomechanical cycling is one of the long lasting concerns in the field of electronic materials. To pay attention to this issue, an artificial neural network approach was developed to assess the lifetime of SnAgCu-based solder joints under thermomechanical cycling processes. Using this approach, it is also possible to distinguish the role of creep and fatigue in the damage evolution. The results demonstrate that the model is able to accurately predict the lifetime of solder joints with minimum possible time. It was also revealed that the solder layer thickness plays a crucial role in the interconnection failure, so that an optimum solder thickness is needed for the highest joint lifetime. Moreover, it was determined that the thinner solder layer and hot dwell temperature in the thermal cycling dominate the creep event on the damage evolution, while the thicker solder layers along with higher ramping rates in the thermal cycling are responsible for the fatigue dominance.

Keywords:

• Creep
• fatigue
• neural network
• solder joint