Akademik Veri Yönetim Sistemi
Experimental Heat Transfer, 2025 (SCI-Expanded)
Impingement jets are one of the promising cooling techniques in thermal management. This study investigates the thermal effects of impinging co-axial jets on different operating parameters. This objective is achieved by experimental and machine learning in the cases of confined and unconfined jets with different nozzle-to-plate distances at constant flow rate ratio and different total flow rates. The study is innovative in that it specifically examines the effect of total flow rate variation of coaxial impinging jet on thermal management at constant flow rate ratio for impingement plates with and without fins, in confined and unconfined jet cases. Experiments and machine learning are conducted for dimensionless nozzle-to-plate distances of H/D = 0.8, 1.6, 2.4, 3.2, a flat impingement plate, and impingement plates with fins of 2 and 6 mm length at Q = 60 and 80 LPM total flow rates and flow rate ratio Q* = 0.5. While the experimental results analyze stagnation, average Nusselt numbers, and heat transfer uniformity, machine learning provides highly accurate predictions and enables thermal performance optimization. The results reveal that while fins and a rise in total flow rate increase heat transfer, they reduce heat transfer uniformity. This is the opposite for H/D≥1.6. Heat transfer on a 2 mm fin-length plate improves by 26% and 22% at 80 and 60 LPM, respectively, compared to a flat plate. On the other hand, the heat transfer deviation on the flat plate is 2.16% and 1.79% for 80 and 60 LPM, while it is 9.05% and 7.33% for a 2 mm fin-length plate. Further, as H/D = 0.8 increases to 3.2, the heat transfer deviation for the flat plate at 60 LPM decreases from 2.62% to 1.79%, while for the 2 mm fin-length plate, it decreases from 9.10% to 7.33%, respectively. Finless plates and low total flow rates are recommended for applications requiring uniform heat transfer, while high heat transfer applications benefit from finned plates and high total flow rates. Finally, machine learning predicts the results with high accuracy.