Predicting vibration-induced fretting in land grid array sockets in simulated field scenarios

Abstract

Electrical contacts provide means for a separable connection between two current carrying conductors. Sockets containing these contacts could be exposed to mechanical vibration due to shipping, which can result in micromotion between mating surfaces. Repeated micromotion/fretting could lead to wearout of the protective gold layer and expose the base metal that can oxidize. Traditional laboratory based fretting experiments may not reflect the field reliability risks. In this study, a predictive capability is developed to investigate contact fretting in a socket due to random vibration using a finite element approach. Considering the degrees of freedom involved in the analysis and the resolution needed, a multiscale modeling approach utilizing global models with substructures to track high risk areas and local models to monitor fretting wear on the highest risk contact is employed. This approach allowed the study of micromotion at the contact interface, capturing stick-slip phenomenon, which can influence fretting wear. Predictions are validated through extensive experimentation, which includes matching of fretting risk areas, matching dynamic responses, and matching of fretting wipe lengths and location through image processing techniques.