People’s living today is much enriched by electronic products: computers, mobile phones, LCD TVs, digital cameras and many other “intelligent” household appliances. These electronic products, no matter whether simple or complex, are created by connecting functional devices (like silicon chips), assistant components (like passives) together with circuit boards. There are not many materials on earth suitable to serve this particular connection. Solder is found to be the most suitable, despite its being soft and subject to creep, even at room temperature. The life of solder joints often determines the reliability of electronic products. The reliability of solder joints, therefore, is always a critical design parameter....[ Read more ]

People’s living today is much enriched by electronic products: computers, mobile phones, LCD TVs, digital cameras and many other “intelligent” household appliances. These electronic products, no matter whether simple or complex, are created by connecting functional devices (like silicon chips), assistant components (like passives) together with circuit boards. There are not many materials on earth suitable to serve this particular connection. Solder is found to be the most suitable, despite its being soft and subject to creep, even at room temperature. The life of solder joints often determines the reliability of electronic products. The reliability of solder joints, therefore, is always a critical design parameter.

Fatigue failure is one of the major reliability issues with solder joints. Since electronic components are made of different materials, they expand to different extents under temperature loading. Serving as interconnecting material, solder joints are subject to thermal fatigue when an electronic product heats up and cools down. These changes in the thermal status of the product happen when the product is in operation, and/or when the environmental condition changes. Thermal fatigue of solder joints is unavoidable. Nevertheless, its influence can be minimized with thorough understanding of the issue.

Investigation of thermal fatigue in solder joints began more than half-a-century ago. The material behavior of solder has been well-studied and the failure behavior of solder joints has been well-characterized. As computers become more advanced, people have started to use finite element (FE) methods to quantify the thermal fatigue damage of solder joints. This benefits the evaluation of solder joint fatigue life under different packaging configurations and use conditions. There are several modeling methodologies developed for the purpose. Yet, they bear deficiencies and hence, a lack of predictive accuracy. This thesis has identified two major modeling problems: 1) the prevailing modeling approaches provide insufficient guidance on the acquisition of a stabilized solution; 2) there is no case-independent definition of the “critical volume” for the calculation of solder joint thermal fatigue damage. The outcome from these two problems is that the fatigue damage evaluated becomes user-dependent, even when the analyst or engineer is experienced. Solutions to these two problems are proposed in the thesis. In addition, they are demonstrated through the fatigue damage evaluation of a PBGA, a QFN and a CBGA.

It typically takes years for solder joints to fail under thermal fatigue. To test solder joint reliability in a short time, accelerated temperature cycling (ATC) tests are implemented by industry. Although “accelerated”, ATC tests often still take months to accomplish. The time-to-failure of solder joints depends on the test conditions. The JEDEC and IPC industrial standards do provide certain guidelines in this regard. However, there are no explicit rules. The choice of the test parameters often rely on the engineer’s experience and previous reliability data. This has been an issue when people did lead-tin soldering in the past. Now industry is migrating toward lead-free soldering, it is time to re-assess the solder joint reliability of most new products and to reconsider the optimization of ATC test conditions. Dwell time is one of the test parameters governing the time-to-failure of solder joints. An analytical approach is investigated in the thesis to determine the optimum dwell time for the ATC test. 3-D FE models of a PBGA, a QFN and a CBGA are used to validate this approach. Satisfactory prediction results have been achieved.