ICF10B

A decrease in the load level can be caused by microstructural changes, fatigue crack nucleation or cavity formation. Some microstructural changes, for example those involving dislocation multiplication and sub grain formation, result in cyclic hardening. Since crack or void nucleation is associated with a reduction in load-bearing section, an immediate and continuous fall in load is observed unless the defects cease to grow. In the classic model of cyclic softening, microstructural changes produce a drop in load followed by a saturation level until fatigue cracking begins and results in a further load drop [2]. However, in a ductile material such as solder, it is likely that both processes, deformation and cracking, will occur simultaneously to some extent with the former predominant in the early stages. The balance between them will be influenced by the strain range and the ductility of the alloy. It is unlikely that in these alloys, crack initiation will occur immediately. Kariya and Otsuka [3] define three distinct stages in the load drop curve, in the first stage Ø increases rapidly, and crack initiation and propagation occur; the second stage is linear reflecting steady state growth, and the third state shows accelerating growth leading to failure. Most of the fatigue life is spent in the second steady state and the fatigue life is dominated by this steady load drop rate dØ/dN. In this study, since testing was terminated at 20 percent load drop the third state was not usually reached. Jiang et al [4] attributed similar load changes in a Sn37 Pb alloy at room temperature to the predominance of cracking. Microstructural examination indicates progressive fatigue cracking taking place throughout all strain ranges but as yet observation of the very early stages is incomplete. Stress relaxation in the Sn-37Pb alloy is significantly greater than that in the lead-free alloys. This cannot be explained on strength or melting point considerations. Temperature itself has negligible effect on the extent of relaxation in any of the alloys, and the strength of the Sn-0.5 Cu alloy over a range of strain rates is similar to that of the Sn-37Pb and significantly lower than that of the Sn-3.5Ag [5]. The fatigue endurance during continuous cycling is broadly similar for all the alloys examined. Since strength is not a good indicator of high strain fatigue resistance, this is not surprising. Indeed, it has been shown that this similarity extends to many engineering alloys [6]. Under total, as opposed to plastic, strain control some advantage might be expected to accrue to the stronger silver-containing alloy but this is offset by its lower ductility[5]. Endurance in the presence of dwells is reflected by the extent of stress relaxation with the Sn-3.5 Ag alloys exhibiting the longest life. However, the relative insensitivity of the Sn-0.5 Cu alloy to short dwells is difficult to explain. The alloy is ductile but not significantly more so than the Sn-37Pb, and a ductility argument would not account for the loss of relative resistance at longer dwells. It is possible that partitioning of the relaxed strain to different locations or into different mechanisms may be required to account for this apparent anomaly. 5 Conclusions When the solder alloys, Sn-37Pb, Sn-3.5Ag and Sn-0.5Cu, are subjected high strain fatigue at 75˚C and room temperature, it has been found that: • they exhibit cyclic softening to a similar degree which is independent of the strain range, temperature and cycle shape; • their fatigue endurance is broadly similar, with fatigue at 75˚C being slightly more deleterious than at room temperature; • during a dwell period in the cycle, stress relaxation occurs to an extent that is independent of temperature and location of the dwell, although it is significantly higher in the Sn-37Pb alloy; • the endurance of all alloys is usually susceptible to the inclusion of dwells, and lifetimes may fall by up to six times after dwells of 100s duration. 6 References 1. Plumbridge, W. J. and. Bartlett, R. A. (1982) Int. J. Fatigue, 4, p 209. 2. Solomon H. D. (1986). Electronic Packaging - Materials and Processes, ASM p.29. 3. Kariya Y. and Otsuka M. (1998). J. of Electronic Materials, 27, 1229-1237. 4. Jiang, H, Hermann R, and Plumbridge W. J. (1996). J. of Materials Science, 31 p6455-6461. 5. Plumbridge W. J. and Gagg C. R. (1999). J Electronic Materials, 10, 461-468. 6. Plumbridge W. J. (2000) Materials at High Temperatures, ,17, 381-387.

RkJQdWJsaXNoZXIy MjM0NDE=