Issue link: https://iconnect007.uberflip.com/i/271756
36 The PCB Magazine • March 2014 laminate material, process thermal history and package construction. The nature of the test produces comparisons at extreme or boundary case conditions but the informa- tion provides insight into more standard con- ditions as well and may be widely applicable. The expansion of the test program to begin to establish the relationship to package stiffness may enable a first level model of susceptibil- ity to this type of mechanical damage around BGA devices. The fact that the rank order of laminate materials did not change with alter- nate package construction leads us to believe that a more comprehensive model than the current is possible. We would with increased amounts of data be able to expand on the cur- rent, board thickness and strain rate model to include laminate material and package stiff- ness. Observations • The dominant failure mode for Pb-free compliant materials under flexure is pad crater • The process strain limit model developed for IPC 9704 when interfacial fracture was the dominant failure mode can be modified to pro- duce more realistic results for Pb-free compliant materials • Under the spherical bend test geometry Pb-free-compliant materials have been shown to have a different strain rate dependency for mechanical failure under load • Significant differences in laminate material performance are evident under this test method • Laminate material susceptibility to this de- fect increases with thermal exposure Acknowledgements The authors would like to thank LSI and IBM for their contributions of material. PCB References 1. Nadimpalli, S., Spelt, J., A Geometry and Size Independent Failure Criterion for Fracture Prediction in Lead-Free Solder Joints, Interna- tional Conference on Solderability, SMTA, To- ronto 2010. 2. Nadimpalli, S., Spelt, J., Characterization and Prediction of Pad-Crater Fracture in Lead-Free Compatible PCBs, International Conference on Soldering Reliability, SMTA, Toronto, May, 2011. 3. Roggeman B, Borgesen P, Li J, Godbole G, SPHERICAL BEND TESTING continues Figure 12: 95% cc for doe3 distributions.