Investigation of double-layered wavy microchannel heatsinks utilizing porous ribs with artificial neural networks

Abstract

Microchannel heatsinks with porous media, wavy and multi-layered configurations introduce compact designs and high efficiencies in cooling applications. Current Computational Fluid Dynamics (CFD) based study bridges the concept by introducing a novel double layered microchannel heatsink with wavy up-down and porous ribs alongside each other. Since Artificial Neural Network (ANN) trained by CFD results reveals designs with better thermal and hydraulic performance with an insignificant computational time, ANN is employed for finding the best performers and effects of waviness on Nusselt number, pressure drop, and maximum temperature difference at the bottom of the microchannel for a Reynold number range of 100 < Re < 800 by defining a thermal efficiency factor. Even though the pressure drop increases in the wavy design, the thermal performance is affected positively. Nusselt number increases more than 55% for Re = 800. The thermal performance increases with the introduction of Dean vortices, especially at higher Re. The best design provides maximum bottom temperature difference of 1.2 °C, with a Thermal Efficiency Factor of 1.34. However, it is observed that better thermal efficiency factors are attained by using a wavy channel at the top layer and keeping the bottom layer straight for the final design.