Impact of Fatigue Modeling on 2>sup<nd>/sup<Level Joint Reliability of BGA Packages with SnAgCu Solder Balls

Abstract

This paper compares the application of different constitutive material and fatigue models for lead-free (SnAg4.0CuO.5) solder proposed by Schubert, Wiese, Dudek and Syed. The application of these models is demonstrated and results are compared with the experimental findings for different type of Ball Grid Array packages, which are subjected to a controlled temperature cycling condition. In addition, a semi-empirical approach was used to fit the finite element modeling results with experimental temperature cycling test data for 2nd level solder joint fatigue failures.

Fatigue Reliability Analysis of Sn–Ag–Cu Solder Joints Subject to Thermal Cycling

“…At 25 • C, most of stress (more than 50%) relaxes rapidly in less than 120 s, and then the stress relaxation is insignificant over a period of 1 day. Some researchers [30]- [32] used 25 • C as a reference temperature in the FE simulation for investigating BGA solder joint fatigue life subjected to TC loading, and they achieved a good agreement between life prediction and experimental results. Pyland et al [37] used underfill cure temperature, 150 • C, as a stress-free temperature in the FE simulation for an underfilled BGA package to predict a consistent fatigue life compared with reliability test results.…”

“…2 shows TC profile used in the FE simulation. Room temperature (25 • C) was considered as a stressfree reference temperature, which was used in many others' work [30]- [32]. [25], [29] VI.…”

“…Usually, the top interface layer of component/solder is a critical area in terms of failure site. The thickness of the interface layer elements is 25 μm for all FE models, which was used for BGA solder joints [13], [21], [32], [33]. Different mesh densities along the radial direction and across the thickness of the 25 μm interface layer were implemented as shown in Fig.…”