Reliability based design optimization for fine pitch ball grid array: Modeling construction and DOE analysis

This paper presents a reliability based optimization modeling approach demonstrated for the design of fine pitch ball grid array (fpBGA) structure. In this study the focus is on the strategy for handling the uncertainties in the package design inputs and their implementation into the design optimization modeling framework. The analysis of thermo-mechanical behavior of the package is carried out to predict the warpage behavior of the fpBGA package and the die surface stress. A 3D non-linear finite element model including appropriate information of packaging material properties is constructed to predict the thermal-mechanical behavior of fpBGA after molding process. Real samples are fabricated and undergone shadow moire¿ inspection and analysis to validate the finite element model. A screening experiment based on orthogonal DOE scheme procedure is carried out in order to find out the correlation between the response and input factors (including interactions) and defining their significance. Finally the most essential/significant input factors can be selected to conduct further analysis (response surface reconstruction, interpolation and optimization) with adding more simulation runs.