No abstract

To investigate modeling approaches for predicting solder interconnect fatigue life under field thermal conditions, a complex temperature cycle composed of three temperature dwell points was modeled and physically tested. Modeling the solder interconnect life expectancy under these complex conditions is accomplished by segmenting the complex temperature cycle into multiple simple temperature cycles, predicting the life expectancy of the solder interconnect under the simple temperature cycles using the Engelmaier model, and applying Miner's rule to superpose the damage of the segmented cycles and predict solder interconnect fatigue life under the complex temperature cycle. Several modeling strategies based on different segmenting schemes are presented, and the one with the best comparison with the physical test results is identified. A case study is presented for ceramic leadless chip carriers assembled with Sn62Pb36Ag2, Sn96.5Ag3Cu0.5, and SN100C solders. The Engelmaier model constants to model the simple temperature cycles are also provided as a basis for modeling the complex cycle.Index Terms-CLCC, Complex temperature cycling, Engelmaier model, life prediction, linear damage superposition, reliability, vary amplitude and mean.

An experimental study was conducted to determine the effect of rework on the quality and reliability of surface mount solder interconnects. The test vehicle included thin small outline packages (TSOPs) and 2512 resistors assembled onto circuit cards with Sn3.0Ag0.5Cu and Sn63Pb37 solders. The rework process entailed the removal and replacement of the TSOPs and resistors. Reworked assemblies were then subjected to a temperature cycling test. Test results indicated that rework did not degrade the reliability of the TSOP solder attachments, but it decreased the thermal fatigue life of 2512 resistors down by 80% due to the decreased and uneven solder joint heights created by the handsoldered rework process. Solder interconnect material changes caused by the rework process were not observed to be a significant factor in reliability.

No abstract