CPSM Simulation of the Variable Properties’ Role in MHD Non-Newtonian Micropolar Nanofluid Flow Over a Stretching Porous Sheet (Flow Filtration)

Analysing the Exponentially Varying Viscosity of Micropolar Carreau Nanofluid Flow with Variable Fluid Properties in Stretching Porous Sheet

In this study, the researchers assumed a thermal energy system with variable controlling properties, mainly like varying viscosity parameters, and power-law index, which has an impact on the overall procedure. Variable thermo-physical features of induced magnetic field on Carreau flow settled with micropolar nanofluid are explored on account of wide range of applications. The micropolar fluids theory focuses on a type of fluids that have tiny effects resulting from the fluid’s micro-motions. Evaluating an micropolar nanofluid’s electrically conducting flows in magnetohydrodynamic (MHD) by virtue of the thermal device is crucial in present metalworking and metallurgy processes. Therefore, the proposed research came with a novel method of neural network with optimization technique also to calculate the accurate result of varying parameters. The obtained differential equation with partial derivatives is transformed into differential equations with ordinary coefficients using the transformation functions. Consecutively, the differential equations with ordinary coefficients are solved using the solution methods of Adam predictor collector and Runge Kutta Fehlberg methods. The thermal extrusion system includes profiles of angular velocity, velocity, concentration, magnetic field, and temperature, in addition to the governing parameters for each. The effectiveness of values acquired by the solution approach was inadequate to continue the investigation, thus a neural network based quaternion values technique was used in solving differential equations to obtain the optimized values of the novel parameters studied in this research. The Mat Lab software is used to carry out for this research’s execution. The research focuses on the varying parameter of viscosity of the nanofluid, therefore the profiles considered was resultant as that the concentration, temperature, and angular velocity profiles decreases as the values of 0.233886, 0.220491, and 0.107346 in addition to a rise in viscosity parameter. However, the velocity rises with the value of 0.970122 as the viscosity parameter values are steadily increased. The effect of utilizing a genetic algorithm based quaternion neural network to optimise the values of the result is compared to two other optimization strategies (MLP + GA and MLP + GD), moreover to the solved numerical values. The novel optimization technique with neural networks gives a better result than the existing methods and the solved numerical values. As a result, this study examined at the MHD based micropolar Carreau nanofluid’s mass and heat transfer on a permeable stretching surface of an induced magnetic field, and it came up with accurate values optimised by a novel neural network model with a genetic algorithm, which gives less error in training and testing data.