Thermal performance and flow characteristics of methanol based Fe 3 O 4 -CoFe 2 O 4 -Co 3 O 4 nanofluid systems: numerical and ANN approaches

The study examines the thermal transport and fluid flow characteristics of Fe3O4, CoFe2O4 and Co3O4 nanoparticles suspended in methanol within convergent-divergent channels. This paper investigates two different flow regions: Region I, which considers the presence of radiation and heat generation/absorption mechanisms along with wall expansion-contraction phenomena, and Region II, which incorporates the additional effects of magnetohydrodynamics and porous media. The partial differential equations describing the system are transformed into ordinary differential equations by introducing similarity variables. Numerical solutions are obtained using the Runge-Kutta method, while complex nanofluid dynamic predictions are generated using a hybrid Artificial Neural Network framework. Results demonstrate that when the Reynolds number is doubled, flow velocity increases within converging channels while decreasing in diverging channels, particularly at elevated Hartmann numbers (Ha=700). Increased heat sinks (Q=−6) reduce the maximum temperature, while higher thermal radiation (Rd=0.2) decreases the temperature profiles. Viscous heating causes the Brinkman number (Br=2) to increase the temperature in both channels. The ANN model shows greater than 98% accuracy compared with numerical solutions. This paper provides fundamental insights into optimizing hybrid nanofluid thermal performance for applications in microfluidic systems, electronic cooling and energy conversion devices.