Experimental and machine learning study on the influence of nanoparticle size and pulsating flow on heat transfer performance in nanofluid-jet impingement cooling

Abstract

Maximizing heat transfer efficiency is crucial for enhancing performance and durability in diverse engineering applications, including fuel cells, EV batteries, and solar PV/T systems, thereby advancing sustainable energy innovation. This study investigates thermal dissipation from a simulated heat sink aligned with a PV cell’s back plate via jet impingement cooling. Specifically, it examines the impacts of pulsatile cooling and nanoparticle size in hybrid nanofluids, comprising combinations of Al2O3 and MWCNT in water, with varied nanofluid volume fraction (0.05 vol% ≤ ɸ ≤ 0.3 vol%) and flow Reynolds number (15000 < Re < 40000). Key findings reveal significant influences of nanoparticle size, nanofluid concentration, and pulsating flow on heat transfer performance. Notably, sample D demonstrated the highest heat transfer enhancement, achieving approximately 52.94 % and 79.06 % improvement in continuous and pulsating jet cooling compared to de-ionized water under continuous jet cooling. Machine learning classifiers were employed to identify critical thermal performance parameters, with Reynolds number identified as the most significant factor influencing heat transfer. Random Forest and Gradient Boosting classifiers showed notable accuracy in predicting Nu, emphasizing the role of machine learning techniques in optimizing thermal management strategies for improved heat dissipation from solar PV cell backplates.