Multi-objective optimization of a non-uniform sinusoidal mini-channel heat sink by coupling genetic algorithm and CFD model

• Published in: Applied thermal engineering Vol. 257; p. 124198
• Main Authors: Ge, Ya, Yin, Weixing, Lin, Yousheng, He, Kui, He, Qing, Huang, Si-Min
• Format: Journal Article
• Language: English
• Published: Elsevier Ltd 01.12.2024
• Subjects: Heat sink; Heat transfer enhancement; Multi-criteria decision-making; Multi-objective genetic algorithm; Non-uniform wavy channel; Heat transfer enhancement; Multi-objective genetic algorithm; Non-uniform wavy channel; Heat sink; Multi-criteria decision-making
• ISSN: 1359-4311
• DOI: 10.1016/j.applthermaleng.2024.124198

[Display omitted] •CFD models are coupled in the evolutionary process to find optimal solutions.•Maximum base temperature is precisely controlled by the genetic algorithm.•Pareto solutions reduce thermal resistance θ by 27.74% or pressure drop Δp by 59.51%.•Base temperature is more uniform as its standard deviation is reduced by 76.98%.•Raising downstream disturbances appropriately is an effective strategy for a low θ. Wavy mini-channel heat sinks (MCHS) can disrupt the development of the thermal boundary layer, resulting in superior performance compared to straight MCHS. For complete utilization of the coolant, this study proposed and optimized a non-uniform sinusoidal MCHS with varying wavelength or varying amplitude along the flow direction. The optimization for five design variables was accomplished through a multi-objective genetic algorithm, where the thermal resistance θ or the pressure drop Δp were defined as two objectives. The computational fluid dynamics (CFD) software was employed to solve all direct problems encountered during the evolutionary process. Compared to the straight MCHS with different channel widths, the optimal designs achieved reductions of 27.74 % in θ or 59.51 % in Δp. Simultaneously, all the optimal values of the wavelength ratio and amplitude ratio indicate that enhancing the heat transfer performance downstream is more efficient. It was observed that among the five design variables, the number of wave cycles is the most relevant parameter with a Spearman’s correlation up to 0.983. Subsequently, a multiple-criteria decision-making approach was employed to ascertain the best compromise solution to balance the two objectives. Although the θ of the best compromise solution could be higher by 38.57 % than that of the lowest θ solution, the Δp presents a more substantial reduction of 92.45 % after the trade-off. Compared to the straight MCHS with the same Δp, the best compromise solution with varying wavelength reduces the θ and the standard temperature deviation by 26.20 % and 76.98 % respectively, while lowering the average base temperature by 3.07 K. Therefore, the optimal non-uniform wavy MCHS, along with the optimization approach presented, is meaningful for practical applications.