Heat transfer enhancement from heat sources using optimal design of combined fins heat-sinks

The rapid movement of heat transferred goes to be the most important target for researchers to deals with the thermal problems of electronic systems. Especially, that has high packing densities in limited space. Consequently, vertical combination between two shapes of fins was considered to improve the thermal performance of heat sink. Here, many models (longitudinal-pin fins) were created as a new approach based on the stable both of material and size. In the analytical solution, Finite element techniques used to solve the mathematical modeling and signum-sinusoidal-signum function was modified to model both the variable heat transfer area and convection perimeter. While in the numerical procedure, ANSYS simulation used as the validation of temperature distribution. In addition, results from previous work were used as second method of validation process. The results show a high level of agreement by maximum difference does not exceed (3.52%). In the calculation process, natural convection, range of Ra number (105-107) were applied for all models of which each one is distinguished by the parameter of area ratio (A*) between hybrid model and longitudinal model. The results show many advantages; a significant drop in temperature profile about (2.7%-8.8%). Also, decrease in thermal resistance by (23%-43%) and increases in heat transferred (29%-78%). Furthermore, all parameters have a significant improve for the models that have A* < 1.48. But this improvement was down to less level at A* > 1.48 as a result of overlapping between the effects of miscellaneous parameters. Irreversibility that associated with the entropy generation was considered. In this step, it becomes important to find the model that achieves maximize of the thermal performance and minimize of the entropy generation. For that, the optimization procedure using genetic algorithm was adopted. It's clearly the best model exist within a close range between 1.509 and 1.518 based on the behavior of the best fitness function. (C) 2020 Beihang University. Production and hosting by Elsevier B.V. on behalf of KeAi.