Finite element modelling of the thermal effects of the manufacturing process on the quality of electronics interconnection

This thesis demonstrates the feasibility of using finite element analysis to model the thenno-mechanica1 effects of manufacturing processes on electronics interconnection. The thesis has two significant complementary parts. The first of these explores the modelling of heat transfer in a commercial solder reflow furnace to contribute to the understanding of the non-unifonn distribution of heat on a printed circuit board within the furnace. The second part of the thesis identifies the thermal stresses in a die attach assembly using an electrically conductive adhesive, typically cured in a reflow furnace, thereby indicating potential points of failure. The first part of the thesis reports the calibrated modelling of a reflow furnace of commercial complexity as part of the demonstration of the feasibility of creating a concurrent engineering design tool for electronic interconnection. The application of a technique using area proportional thermal conductances to model the convective heat transfer between the furnace and board is demonstrated. The model explores the effects of the mechanisms of heat transfer within the oven to show that this is more uniform with the addition of convection and that edge heaters appear to have little effect. In the second part of the thesis the effect of the material properties of glass transition temperature, Youngs Modulus and coefficient of thermal expansion were quantified using finite element modelling, to show that these properties significantly affect the structural integrity of the interconnections constructed using conductive adhesives. Electronic Speckle Pattern Interferometry (ESPI) methods are also shown to be an effective method of validating such finite element models. This work required the measurement of the viscoelastic properties of typical materials using novel specimen preparation techniques. It is anticipated that the conclusions presented are conservative because of the assumptions and sensitivity analyses made during the modelling activity. The combined results of the two parts of this work demonstrates the feasibility of modelling the solder reflow process and highlights the potential of electrically conductive adhesives as replacement for solder in high temperature applications. Promising avenues for further work include improved non-linear modelling of interconnection systems and understanding of the effect of rework, including reliability assessment. In addition, further work remains to be carried out in the determination of relevant material properties in representative configurations.