A finite volume and machine learning based investigation of flow dynamics in a vertical duct heated by the sunlight

Thermal and flow features of nanofluid are explored under the impact of an external magnetic field whereas fluid is flowing inside a vertical square duct. It is assumed that the sunlight tends to heat up the duct. Two fins are attached to the duct that receive sunlight of uniform intensity. Two walls of the duct contain a fixed temperature and, are exposed to the direct sunlight under the assumption of a particular thermal boundary condition and developed flow. The governing equations of the problem are solved by the finite volume method. The taller fins not only act as significant obstacles causing reduction in flow velocity and frictional losses but also contribute to enhanced the rate of heat transfer more extensively with the surfaces of fins. Higher Rayleigh number fosters buoyancy-driven natural convection, causing a change in the thermal signature and diminution in thermal distribution over the domain. The magnetic force also tends to impede thermal diffusion, thus hampering heat transfer within the flow. The pressure gradient and the nanoparticle volume fraction significantly reduce fRe. On the other hand, the fin height, Rayleigh number and the magnetic parameter cause an enhancement in the Nusselt number and fRe. Finally, introducing a modern perspective into the research approach, machine learning technique (Neural Fitting app) was incorporated to refine the analysis.