V. Vinay Kumar

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.

This research examines the optimization of thermal enhancement in viscous fluids using statistical and numerical methods. The present study considers the impact of thermophoretic particle deposition on the swirling flow generated by the cylinder’s torsional motion. It is presumed that the cylinder rotates continuously along its axis and that the expansion rate of the cylinder wall is proportional to the position along the axis from the origin. Similarity variables are employed to convert partial differential equations (PDEs) into ordinary differential equations (ODEs), which are subsequently solved using the Runge-Kutta-Fehlberg-4 ^th –5 ^th (RKF-45) order numerical technique. Furthermore, Taguchi-based optimization, in conjunction with sensitivity analysis, is employed to assess the thermal transfer rate. The graphical representation indicates that with a rise in the Reynolds number, both the velocity and thermal profiles decrease, while the concentration profile increases. The Taguchi technique indicates that the third level of the heat source/sink parameter, the second level of the Dufour number, and the first level of the Reynolds number provide optimal circumstances for optimizing the Nusselt number. The analysis forecasts its industrial significance by attaining a maximum heat transfer rate of 4.916537. The Reynolds number (94.76%) has the largest effect on the Nusselt number, while the heat source/sink parameter (1.53%) has the least.