Novel artificial neural network approach for entropy generation in Jeffery–Hamel blood flow with magnetized hybrid nanofluid in non-uniform arteries

The main goal of this research is to investigate the artificial neural network (ANN)-based solution for the Jeffery–Hamel model, specifically focussing on the behaviour of non-Newtonian Sisko fluid in the magnetohydrodynamic (MHD) flow of hybrid nanofluid through vertically converging channels. Gold and iron oxide are nanoparticles that use blood as their base fluid. The Sisko fluid model more accurately represents the rheological properties of blood due to its shear-thinning and shear-thickening characteristics. The vascular segment is classified as a convergent channel. A slip boundary condition is implemented to eradicate the viscous effect at the arterial wall area. An examination of transport phenomena is favoured by utilising heat transfer during melting, applicable to the melting of fats or plaques at the vessel walls. Heat transfer procedures incorporate the benefits of quadratic thermal radiation in cancer treatment, hyperthermia, and tumour therapy. The governing equations are transformed to dimensionless form using the similarity transformation. The non-dimensional governing equations are numerically solved using the bvp4c method along with the artificial neural network (ANN). The velocity profile, temperature, and entropy generation are examined graphically for various parameters and also compared with ANN results. The skin friction and heat transfer rate are also looked at and presented in tables. The graphical outcomes demonstrate that the velocity profile declined and the temperature profile enhanced when the magnetic field and Sisko fluid parameter enhanced. The tabular results show that the skin friction coefficient decayed for large values of volume fraction for both and nanoparticles and the Nusselt number enhanced for large values of radiation parameter and magnetic parameter.