| Summary: | 碩士 === 國立中正大學 === 機械系 === 87 === The reliability of surface mounting electronic packaging such as PQFP and BGA could depend on a number of parameters that related to material, configuration, and manufacturing. One of the main factors led to failure is the coefficient of thermal expansion (CTE) mismatch between two layers of packaging which will cause notably thermal stresses concentration in their interconnections. Therefore, it is important to establish the experimental method to examine the thermal strain distribution inside packaging. It is also important to develop numerical analysis model to predict the thermal strain and stress during the change of the packaging configuration.
The aim of this research is to develop the experimental methods and numerical model that can be use as design tools to study the thermal deformation of electronic packaging. In this research, a whole field optical experimental method Phase Shifting Moire Interferometry are set up first to detect thermal displacement contour within the packages cut plane with high sensitivity up to 0.03164μm and the effects of temperature change in the range from 55 to 95 degree. Full field strain maps can also be measured on the packages cut plane. Two types of testing methods is used to study the thermal deformation of BGA packaging. One is uniform heat up the whole specimen to simulate reflow process, the other is embedded wire type resistance on die surface then input electrical power to heat up from die surface to whole packaging. This method is tempted to simulate the thermal deformation for electronic packging in real time work condition. Next, three-dimensional finite element models of PQFP and BGA are built and analyzed with the material parameters and the CTE values. The thermal deformation predict by this hybrid FE model are re-compare with Moire testing data. Once the FE Model that compare well with Moire testing data; then it can be used to analyze the thermal stress and strain effected by structural modification so that better configuration of the packaging structure can be predicted.
|
|---|
