Blood rheology-based analytical framework for time-dependent radiative Jeffrey nanofluid flow modelling using Laplace transform

Purpose Non-Newtonian fluids are widely used in biomedical and industrial domains, including blood rheology, lubrication technologies and shock-absorbing devices, because of their adaptive flow properties. Motivated by these wide-ranging applications, this study aims to investigate the Jeffery hybrid nanofluid flow over a porous vertical surface. The analysis considered the dispersion of three nanoparticles, ferric oxide (Fe3O4), cadmium telluride (CdTe) and magnesium oxides (MgO) in blood as a base fluid. Design/methodology/approach The principle of scaling analysis is used to reformulate the governing equations into a dimensionless partial differential equations (PDEs). Moreover, the PDEs are analytically solved by the application of Laplace transformation, and the impact of physical parameters on velocity and thermal profiles are demonstrated graphically. In addition, the artificial neural network (ANN) was trained using the Levenberg–Marquardt optimization algorithm to predict the heat transfer rate at the surface of the plate. Findings The outcomes of the present analysis stated that the yield stress of Jeffery fluid plays a critical role in suppressing the flow profile, and the enhancement of nanoparticles in base fluid which significantly raises the fluid temperature. However, the results of neural networks show better prediction accuracy, as indicated by a low mean square error value. Originality/value The present study provides an exact solution of Jeffery hybrid nanofluid flow past an upright plate applying Laplace transformation. The ANN framework is implemented to predict the thermal transfer rate with high accuracy. The predictive outcomes of ANN demonstrate excellent agreement, confirming the validity of the analytical and machine learning approach.