Finite element simulation of double-layer BGA solder joints under torsional loading conditions and optimization of solder joint morphology parameters

Abstract

In this paper, a prototype Ethernet chip RU82566DM manufactured by Intel Corporation is used to build double-bump BGA Solder Joints. Based on the packaging method in its data sheet, the finite element model of the stacked solder joints was built using ANSYS software, and the effect of the morphological parameters of the stacked joints on their maximum stress under torsional loading conditions was analyzed. Firstly, the three parameters of solder ball diameter, solder joint height, and pad diameter were selected as the control factors, and the stacked solder joint stress was used as the target value. The Box-Behnken method was used to obtain 17 sets of horizontal combinations of different morphological parameters of the stacked solder joints. After that, the torsional stresses were simulated for each of the 17 finite element models established. Finally, the response surface method was used to fit the stress values to the morphological parameters of the solder joints in the 17 sets of simulations to obtain the regression equations and to perform the simulation verification of the optimal combination of morphological parameters of the stacked solder joints. The results showed that: under torsional loading conditions, the force distribution within the stacked solder joint array and the force distribution of individual stacked solder joints are not uniform. The key solder joint in the overall model of the stacked solder joint is located at the farthest corner from the array of solder joints, and the maximum force value occurs at the outermost point of contact between the stacked solder joint and the PCB; According to the analysis results, the diameter of the solder ball, the height of the solder joint, and the diameter of the pad have significant effects on the torsional stress of the stacked solder joint; The response surface model is a good fit, when the solder ball diameter is 0.61 mm, the solder joint height is 0.50 mm and the pad diameter is 0.48 mm, the torsional equivalent force of the stacked solder joint is the smallest, 25.56 Mpa. To verify the accuracy of the regression model, finite element simulations were performed using this optimal combination of solder joint morphology parameters, and the simulation results showed a maximum stress of 26.3 Mpa, with an error of only 0.38% between the two, so the model based on the response surface method can be considered accurate.