| Summary: | 碩士 === 中華大學 === 機械工程學系碩士班 === 102 === Electronic Packaging reliability is of great concern to semiconductor and electronic product manufacturers. From the first to the fourth generation computer, the tremendous growth of computers and 3C products, and its significant impact on our lives is the invention of the microprocessor. The Wafer Level Chip Scale Package (WLCSP) is still available as one of the promising packages since 1998. The reliability issue for any particular packaging should be investigated before mass production.
The purpose of this research is to design and analysis of the reliability performance for Ankor type copper pillar WLCSP packaging in board level. The outline procedures include: (1) Solder joint reliability in TCT loadings is performed by finite element method; (2) Ankor copper pillar structure optimization by Taguchi method; and (3) Based on the optimal structure, drop test and simulation are performed. In finite element method, the loading of the packaging is simulated under JESD22-A104-B conditions.
To provide reliability issues, the thermal-mechanical behavior using 3-D 1/4 finite element model by ANSYS was performed. Solder joint with SAC305 Anand constitute is considered under the temperature thermal cycling temperature range of -40 to 125°C and the plastic work energy is selected as the performance index. Due to the dimension scale of different components, the submodel technique is used.
The Optimal cupper pillar infrastructure is analyzed using eight factors with L18(21×37) orthogonal table by Taguchi method. Analysis of variance is performed the consequence of the factors. The copper pillar height and epoxy material are the most critical factors to the solder joint reliability. The life prediction of optimal design is 43 % better than that of the original design.
Under the optimal design, peeling stresses of each component around the package side are investigated by changing the dimensions of the most two important factors. In drop test, JEDEC standard JESD22-B111 conditions B is performed by drop test apparatus in own Lab. Two finite element models, Input-D and Support Excitation, are performed and compared the dynamic responses with experimental method. The contribution of this study is an guidance for the manufacturer.
|
|---|
