Modelling and analysis of heat transfer in porous and non-porous interfaces in a backward step channel with ternary hybrid nanofluids

Heat exchange between porous and non-porous materials is widely studied in engineering and geothermal applications because of their demonstrated capability to improve thermal performance. Applications include gas turbine blades, chemical reactors, seepage heat transfer, and electronic cooling components. Additionally, the use of backward step channels has been thoroughly investigated due to their structural advantages in a range of engineering systems. In the present study, magnetohydrodynamic flow through a backward step channel containing porous material (concrete) up to half the height of the downstream region is investigated. The study focuses on the interaction at the junction between porous and solid regions, where the non-porous section is analyzed using ternary hybrid nanofluids consisting of gold, silver, and copper nanoparticles dispersed in water as the base fluid. A two-dimensional model is employed, solved using COMSOL Multiphysics 6.2, which incorporates the Navier–Stokes equations, Brinkman equation, and two-dimensional heat equation. The investigation includes a parametric analysis to explore the influence of Reynolds number (Re), Darcy number (Da), porous medium porosity, total nanoparticle volume fraction in the ternary hybrid nanofluids, Hartmann number (Ha), and the geometry of the nanoparticles . The results are validated against available literature. The research investigates how temperature and Nusselt number behave at the boundary separating porous and non-porous regions. Findings indicate that interface temperature rises with increased Darcy and Reynolds numbers, while it decreases as the magnetic field intensity (represented by the Hartmann number) and the concentration of ternary hybrid nanofluids are increased. The study identified that optimal thermal conditions for achieving the highest interface temperature occur when Da = 1, Ha = 5, Re = 1000, ε = 0.2, ϕ = 1%, and n = 3.7. Furthermore, the study highlights findings related to the local Nusselt number, showing that the highest average Nusselt number of 2.7838 was achieved under the conditions Re = 1000, Ha = 25, Da =0.01, ε = 0.2, ϕ = 1%, and n = 3.