Contour Analysis for Heat Transfer Rate in a Wedge Geometry with Non-Uniform Shapes Nanofluid: Gradient Descent Machine Learning Technique

Abstract

The flow of boundary layer over porous wedge nanofluids, hybrid nanofluids, and ternary hybrid nanofluids under constant transverse magnetic field and heat radiation have all been examined. The following cases are taken into consideration: Zirconium oxide + A A7072 + Polyethylene Glycol-water (Case-2); Magnesium oxide + Polyethylene Glycol-water + AA7072 + Zirconium oxide (Case-3); Polyethylene-water as base fluid; nanofluid, hybrid nanofluid, ternary hybrid nanofluid cases. The governing equations are solved using the shooting technique with the fourth-order Runge-Kutta method. The approximate relationship between temperature, velocity, heat transfer rate, and friction factor coefficient at the wedge is shown graphically for a range of values of the pertinent parameters. The flow of boundary layer over porous wedge nanofluids, hybrid nanofluids, and ternary hybrid nanofluids under constant transverse magnetic field and heat radiation have all been examined. Cases containing Polyethylene Glycol-Water + AA7072 (Case-1), Zirconium oxide + A A7072 + Polyethylene Glycol-water (Case-2), and Magnesium oxide + Polyethylene Glycol-water + AA7072 + Zirconium oxide (Case-3) are taken into consideration. Polyethylene water is used as the base fluid. The governing equations are solved using the shooting approach and Runge-Kutta fourth order. The approximate relationship between temperature, velocity, heat transfer rate, and friction factor coefficient at the wedge is shown graphically for a range of values of the pertinent parameters. Comparing Ternary hybrid nanofluid to nanofluid and hybrid nanofluid situations, it is discovered that the skin friction coefficient is higher. It has also been noted that ternary hybrid nanofluids have a higher Nusselt number than hybrid and nanofluids. When applied to specific dimensionless parameters, the basic linear regression machine learning technique using the gradient descent method accurately predicts the truth values. Finding the optimisation conditions values based on the key factors influencing the Response Function is easier with the RSM method's help. When applied to specific dimensionless parameters, simple linear regression machine learning using the gradient descent method accurately predicts the truth values.