Arc-curved microchannels engraved on segmented circular heat sink for heat transfer augmentation; ANN-based performance optimization

Autor: Noureddine Elboughdiri, Sinan Q. Salih, Bhupendra Singh Chauhan, Aliashim Albani, Tirumala Uday Kumar Nutakki, Fahad Alturise, Salem Alkhalaf, Saifeldin M. Siddeeg, Ahmed Deifalla
Jazyk: angličtina
Rok vydání: 2024
Předmět:
Zdroj: Case Studies in Thermal Engineering, Vol 53, Iss , Pp 103837- (2024)
Druh dokumentu: article
ISSN: 2214-157X
DOI: 10.1016/j.csite.2023.103837
Popis: The proficient management and supervision of thermal conditions in microelectronic equipment are becoming progressively challenging owing to the imposition of heightened heat fluxes. The exploration of recently developed microchannel heat sinks (MCHSs) is currently underway to effectively tackle this matter. Modifying the geometric characteristics of these gadgets has proven to be highly effective in augmenting their comprehensive performance. In the present study, a novel arc-curved microchannels engraved on segmented circular heat sink and introduced configurations comprising of six discrete segments of spiral microchannels. The study focused on examining the influence of individual partitions on the Nusselt number (Nu) and pressure drop (ΔP) of the MCHS. The obtained results demonstrate that all the proposed configurations exhibit considerably higher values of the Nu in comparison to the reference MCHS consisting of 21 rectangular channels. Nevertheless, the notable augmentation in the value of the parameter Nu occurred simultaneously with an escalated ΔP. Hence, the relative efficiency index was employed as a more precise measure for evaluations. All of the designs suggested in this investigation were found to be effective when subjected to volume flow rates surpassing 47.303 ml/min. In an illustrative case, the segmented circular MCHS encompassing 10 spiral microchannels exhibited an enhanced overall efficiency of 25.9 % when operated at a volume flow rate of 66.225 ml/min, as compared to the benchmark design. In addition, artificial neural networks (ANN) were utilized to comprehensively scrutinize the influence of variables and construct a prognostic model to ascertain the overall effectiveness of the MCHS, obviating the necessity for convoluted, time-consuming, and costly experimental procedures and computations. The presentation of the Pareto optimal points effectively portrayed the optimal and unfavorable outcomes under consideration.
Databáze: Directory of Open Access Journals