Multi-objective optimization of a mini-channel heat sink with non-uniform fins using genetic algorithm in coupling with CFD models

Abstract

• Genetic algorithm and CFD models are coupled to optimize up to 10 design variables. • Variation of base temperature is effectively controlled by the optimal algorithm. • Optimal solutions reduce pressure drop by 13.62% or thermal resistance by 10.24%. • Front sparse and rear dense fin arrangement is suggested based on optimal geometries. Micro/mini-channel heat sinks (MCHS) have been extensively employed for heat dissipation under high heat flux conditions, and their performances are crucial for safe and stable operation. This study presents a multi-objective optimization work to reduce pressure drop Δ p and thermal resistance θ for the MCHS with non-uniform fins in a staggered arrangement. Multi-objective genetic algorithm and computational fluid dynamics software were coupled to find the Pareto solutions with optimal fin lengths and longitudinal spacings. Compared with the initial MCHS with four uniform fins, Δ p of solution Optimal θ and θ of Optimal Δ p were respectively reduced by 13.62% and 10.24%. Meanwhile, Spearman's rank correlation coefficient was obtained to reveal the relationship between fin configurations and performances, which indicated that the front sparse and rear dense fin arrangements are beneficial in achieving high comprehensive performances. To fully utilize the pumping power, the upstream heat transfer performance was sacrificed, however, all the local maximum base temperatures were excellently controlled to ensure thermal resistance. Furthermore, a multi-criteria decision-making approach was applied to select the best compromise solution Optimal TOPSIS . Besides reducing Δ p by 8.35% and θ by 6.13%, Optimal TOPSIS also reduced the material cost by 10.80% and improved the uniformity of base temperature by 2.18 K.