Efficient cooling capability in microchannel heat sink reinforced with Y-shaped fins: Based on artificial neural network, genetic algorithm, Pareto front, and numerical simulation
Abstract
Microchannel heat sinks play a vital role in modern technology due to the increasing demand for efficient thermal management in compact electronic devices. These systems enhance heat dissipation and maintain optimal operating temperatures, yet conventional heat sinks often fail to meet the stringent cooling demands of modern technologies. To address this, a novel microchannel heat sink reinforced with Y-shaped fins was introduced as an advanced cooling solution. Unlike traditional straight fins, Y-shaped fins improve flow distribution, reduce hot spots, and enhance temperature uniformity across the system's surface. Artificial neural network models were developed to evaluate the impact of fin geometry on performance. Key geometric parameters, including fin attack angle, vertical spacing, and horizontal spacing, were used as input variables, while the Nusselt number and pressure drop were selected as performance outputs. The results revealed that the fin attack angle was the most influential parameter affecting the outputs. The applied cost functions demonstrated the high accuracy of the models in predicting system performance. A genetic algorithm was employed for single-objective optimization targeting three criteria: maximizing total efficiency, minimizing pressure drop, and maximizing the Nusselt number. Two optimized designs were proposed. Design 1 (with an attack angle of 60°, vertical spacing of 120 μm, and horizontal spacing of 400 μm) was optimal for maximizing total efficiency and minimizing the pressure drop. This design achieved a 69.26 % increase in the Nusselt number and a 42.9 % improvement in total efficiency compared to the finless design. Design 2 (with an attack angle of 120°, vertical spacing of 175.650 μm, and horizontal spacing of 200 μm) focused on only maximizing heat transfer, resulting in a 109.28 % increase in the Nusselt number and a 27.1 % improvement in total efficiency. A multi-objective optimization process was conducted in response to the need to balance these multiple objectives. The TOPSIS analysis and Pareto fronts for the Nusselt number and pressure drop were generated to provide a comprehensive framework for designing an efficient and practical microchannel heat sink.