Thermal stress analysis of PBGA under natural convection using a fluid-solid coupling method

Previous researchers usually set a constant heat transfer coefficient (HTC) for the surface heat exchange according to experience or empirical formulas to investigate the distribution of temperature and stress in their thermo-mechanical analysis, but this method does not satisfy the real condition since HTC may not be the same as people assume especially when the size is very small. Moreover, HTC is not constant but a function of temperature, location and complex structure of the package. Therefore the results obtained by setting constant HTC may not be accurate. In this paper, we demonstrate that HTC setting has a big influence on the distribution of temperature and stress, a slight change of HTC has great effect on the precision of final results. A more accurate method is necessary. In order to realize the above target, the fluid-solid coupling method was presented. To verify the validity of this method, an experiment was conducted. And the temperature difference between the simulation and experiment is less than 10%, which indicates that the fluid-solid coupling method is valid. We calculated the temperature and stress distributions by using the constant HTC based on empirical equation and by the present coupling method, the comparison shows that there is difference when using these two methods. This is due to that the HTC is different at every location of packaging surface, using an averaging and constant HTC will produce errors. To obtain the accurate result, fluid-solid coupling method must be utilized, especially when the empirical formula is not available in many application cases.