Akademik Veri Yönetim Sistemi
AIP Advances, cilt.16, sa.2, 2026 (SCI-Expanded, Scopus)
Novel cooling methods are needed to increase the efficiency of energy systems and to develop compact engineering systems. The present work proposes a novel bifurcating channel cooling system, which contains a sinusoidal wavy wall, a rotational circular cylinder at the junction, and an inclined magnetic field during forced convection of a ternary nanofluid. Numerous values of the Reynolds number (between 200 and 1000), non-dimensional rotating speed (Ω between −4000 and 4000), Hartmann number (Ha between 0 and 20), amplitude of the wavy wall (Af between 0.01 and 0.3), and wavy number of the wavy wall (Nf between 1 and 8) are all subjected to numerical simulations. It is observed that at the highest Re, the average Nusselt number (Nu) increment factor for the upper and lower hot walls with the cylinder configuration is 1.96 and 2.77, respectively. Using rotating cylinders at maximum speed for the upper and lower hot walls may improve cooling performance by 19.3% and 34%. For N-C (no-cylinder) and W-C (with cylinder) scenarios, employing a magnetic field at maximum strength improves the cooling performance of the upper wall by around 44.5% and 26%, whereas using W-C degrades the cooling performance of the lower wall by 20%. Enhancements in cooling performance are largely influenced by the wave number and corrugation amplitude, and the amount of the increment depends on the cylinder’s presence. When compared to the reference case (no-cylinder, flat wall, Ha = 0), the cooling performance improvements for the best case of the upper (wavy, W-C, Ha = 20, Ω = 4000) and lower hot wall (wavy, W-C, Ha = 0, Ω = −4000) are 175.4% and 190.1%, respectively. Using feed-forward neural networks, successful estimation results are produced for the cooling performance of the T-channel’s top and lower hot walls.