Optimal design of a wavy Micro-Channel based on Multi-Objective genetic algorithm

Abstract

With the increase of heat generation in electronic equipment, microchannel heat sinks (MCHS) have been widely adopted as a solution for heat dissipation under high heat flux conditions. This study aims to optimize the thermal resistance (θ) and the pressure drop (Δp) of a wavy MCHS through a multi-objective optimization method. The channel shape of the MCHS is controlled by adjusting the coordinates of seven nodes (Si) evenly distributed along the flow direction. The genetic algorithm and CFD software COMSOL Multiphysics are combined to determine the Pareto front with the optimal θ and Δp. The Pareto front suggests that gradually increasing the channel width in the inlet section can enhance the comprehensive performance, by appropriately introducing perturbations in the middle and rear sections. The Spearman’s correlation is adopted to reveal the relevance between Si and objective functions, which also indicates that downstream nodes have a higher correlation. Furthermore, a multi-objective decision-making algorithm is employed to identify OptimalTOPSIS as the best compromise solution. Compared to the straight-through channel, OptimalTOPSIS not only reduces Δp by 40.8% with almost unchanged θ, but also improves the uniformity of the substrate temperature by 4.2%.