Multi-objective Optimization Study of Stacked BGA Solder Joints under Coupled Bending-Vibration Load and Crosstalk Effects

• Published in: IEEE transactions on components, packaging, and manufacturing technology (2011) packaging, and manufacturing technology (2011), 2024-12, p.1-1
• Main Authors: Wu, Liye, Huang, Chunyue, Wei, Jisheng, Liang, Ying, Gao, Chao, Liu, Xianjia, Cao, Zhiqin
• Format: Article
• Language: English
• Subjects: Bending; Bending vibration coupling; Crosstalk; Crosstalk analysis; Entropy weight TOPSIS decision making; Finite element analysis; Load modeling; Loading; Multi-objective optimization; Reliability; Soldering; Stacked solder joints; Strain; Stress; Vibrations
• ISSN: 2156-3950
• DOI: 10.1109/TCPMT.2024.3520978

A finite element analysis model and crosstalk analysis simulation model of stacked BGA(Ball Grid Array Package) solder joint interconnection structure were established, and the stress-strain finite element analysis under bending-random vibration coupled loading conditions and the crosstalk analysis between stacked BGA solder joints under high frequency conditions were conducted, along with the completion of the bending vibration coupled strain measurement experiment. Based on the response surface method, a multi-objective optimization mathematical model of stacked BGA solder joints was constructed, and the structural parameters were optimized using the MOEA/D(Multi-Objective Evolutionary Algorithm based on Decomposition) combined with the entropy weight TOPSIS decision method. The results show that the maximum bending vibration coupled stress and strain of the stacked BGA solder joints are both located on the lower surface of the contact between the lower solder joint and the PCB(Printed Circuit Board), and on the outer edge of the contact surface. As the signal frequency increases, near-end crosstalk and far-end crosstalk gradually increase, and near-end crosstalk is greater than far-end crosstalk. Among the 17 sets of Pareto optimal solutions obtained, the comprehensive optimization ratios of the structure parameter combinations in the first two orders reached 9.56% and 9.63% respectively, with the optimization effects of each target relatively average, achieving the purpose of multi-objective optimization.