Characterization of Test Board Design and Evaluation of High-Performance Solder Alloys Under Vibration Testing
View/ Open
Date
2024-07-23Type of Degree
PhD DissertationDepartment
Industrial and Systems Engineering
Restriction Status
EMBARGOEDRestriction Type
FullDate Available
07-23-2029Metadata
Show full item recordAbstract
In the service life of electronic products, there are many applications where they experience vibration loading. However, experimental vibration durability is less investigated due to the complexity involved. Previous research on electronic packages has been conducted predominantly to study the effects of low cycle fatigue. Very little research has been done in the field of high-cycle fatigue. This is critical for aerospace as well as automotive industries. Solder joints connect components to PCB. They form the primary mechanical, electrical, and thermal interconnects. Therefore, reliability of solder joints is crucial for performance of electronic assemblies, especially in critical environments. Vibration loading induces stress that is of low amplitude but high frequency. The focus of this study is to evaluate lead-free solder alloys in a high-frequency environment. This study involved two printed circuit board designs. The first board design had failures predominantly in the form of trace fractures, which is common in a vibration test and hence they are not optimal to study solder fatigue. Challenges involved with board design for vibration test are observed. To overcome these challenges, a new test board was designed for high cycling vibration test. The frequencies used in this test are above 400 Hz. Various alloys are tested and their performance is evaluated with eutectic SnPb alloy as a benchmark. In this work, a sinusoidal vibration test is employed to evaluate the durability of various solder alloys. These alloys were developed for harsh environment applications. Test matrix comprises of lead-free solder alloys along with the conventional SAC305. Unlike previous studies, solder joints created here are homogenous, that is, solder paste alloy was matched to the solder ball type. A 0.8mm pitch BGA208 component with ii 18 mil solder spheres was assembled to a board with OSP surface finish. Test boards were designed to monitor the corner joints separately, from the rest of the solder joints in the package. The assembled board was then secured to a fixture on top of an electrodynamic shaker using JEDEC standards. A pilot test was conducted to verify if the new board was suitable for vibration fatigue. Two accelerometers were mounted, one on the shaker and the other one on the test board. The test starts with running a sine sweep to determine the natural frequency. The natural frequency is determined with the response of both the accelerometers. PCBs are then subjected to a resonance dwell test until failure. The electrodynamic shaker is connected to a vibration control system and a data logger to monitor and log the resistance data. Data was acquired from the data logger to compare characteristic lifetimes. Lifetime reliability was evaluated using a Weibull chart. Failure modes of all the alloys were investigated by cross-section analysis with the help of SEM and optical microscope. Dye and pry analysis was also performed. Fatigue cracks were seen at the solder IMC interface consistently. This is because the IMC is the most brittle part of the solder joint, and they are susceptible to failure during high cycle fatigue tests. Therefore, new test board design was found to be suitable for high cycling vibration test and further testing is performed on this board. For evaluating solder alloys after isothermal aging, eutectic SnPb, SAC305, Innolot and SABIX were selected. These boards were aged for 3 months at 125 ̊C and tested at the same vibration amplitude and their performances were compared. Similar to as-assembled boards, fatigue cracks were observed near the IMC interfacial layer. Another challenge faced during the vibration test was due to resistance monitoring technique. A two-wire sensing was used for the pilot test and lot of fatigue cracks were undetected, only to be iii observed during cross-section analysis. An alternative method of four-wire sensing technique is compared to results acquired from the two-wire sensing technique and the four-wire method is found to provide accurate measurements for high frequency vibration testing. Crack growth rate using these resistance measurements were analyzed and Innolot was found to have the slowest growth rate in as assembled alloys and SAC305 had the slowest growth rate among aged samples.