Title :
Reliability, lifetime prediction and accelerated testing of prospective alternatives to lead based solders
Author :
Mavoori, Hareesh ; Chin, Jason
Author_Institution :
Dept. of Mater. Sci. & Eng., Northwestern Univ., Evanston, IL, USA
Abstract :
In this paper, the mechanical properties, especially those related to fatigue, of two lead-free solder systems: 96.5 Sn-3.5 Ag and 91 Sn-9 Zn have been compared with those of 63 Sn-37 Pb solder to assess the reliability of lead-free solders as replacements for lead-tin solders in second level surface mount technology (SMT) interconnects. The strategy adopted to reduce the time required for generating a mechanical property database is discussed along with models to predict fatigue life from short time tests and early test data. The performance of single-connection feed-forward back-propagation neural nets in fatigue lifetime prediction is tested and compared with that of a semi-empirical model
Keywords :
backpropagation; circuit reliability; fatigue; feedforward neural nets; life testing; printed circuit testing; production testing; silver alloys; surface mount technology; tin alloys; zinc alloys; SnAg; SnZn; accelerated testing; fatigue; lifetime prediction; reliability; second level surface mount technology; semi-empirical model; short time tests; single-connection feed-forward back-propagation neural nets; Databases; Environmentally friendly manufacturing techniques; Fatigue; Lead; Life estimation; Life testing; Mechanical factors; Predictive models; Surface-mount technology; Zinc;
Conference_Titel :
Electronic Components and Technology Conference, 1995. Proceedings., 45th
Conference_Location :
Las Vegas, NV
Print_ISBN :
0-7803-2736-5
DOI :
10.1109/ECTC.1995.517811
