Multi-objective optimization of a microchannel pin-fin hybrid heat sink based on CFD and NSGA-II genetic algorithm

High-power compact electronic devices often generate concentrated heat in small areas, leading to serious cooling challenges. A hybrid heat sink that combines microchannels with pin-fins is a practical solution to handle these hotspots more effectively. However, most existing studies focus on regular fin geometries or uniform heating conditions and rarely consider hotspot-oriented shape co-design. To identify fin shapes that better address local hotspot issues, 50 pin-fin shape designs using CFD simulations are tested to evaluate their thermal resistance and pumping power. Then, a correlation model is built to show how geometry affects performance and a multi-objective genetic algorithm (NSGA-II) is used to find the best design. The fin shape of the TOPSIS solution is considered optimal, as it effectively enhances local boundary layer disruption near the hotspot region, thereby enhancing localized convective heat transfer. At the same time, the fin design maintains smooth flow channels, minimizing additional pressure losses. Under the same pumping power, the optimized hybrid heat sink achieves a 16.5% reduction in total thermal resistance. Conversely, when thermal resistance is kept constant, the required pumping power is reduced by 62.9% compared to a traditional microchannel heat sink. These results show that optimizing the shape of pin-fins can greatly improve the thermal-hydraulic performance of a microchannel pin-fin hybrid heat sink. • Explores non-regular pin-fin cross-sectional shapes instead of conventional fixed geometries. • The cross-sectional shape of pin-fins is parameterized using six geometric variables. • A correlation between fin cross-section and thermal-hydraulic performance is established, while the pin-fin shape is optimized using NSGA-II. • The optimized design achieves 16.5% lower thermal resistance and 62.9% less pumping power compared to traditional designs.