Deep learning-based prediction of heat transfer in Newtonian fluid flow over convective cone and wedge surfaces

Abstract

Problem statement: The absence of a detailed thermal study of the hybrid nanofluid flow over wedge and cone surfaces, including the Christov heat flux model, inclined magnetic field, porous effects, and convective boundary conditions, and the accurate prediction with Deep Neural Networks, inspires the current study of relevant heat transfer characteristics. Aim and objective: The purpose of this work is to investigate and improve heat transmission in scaled thermal systems using a hybrid nanofluid made of copper and alumina nanoparticles floating in water.This investigation looked at the thermal behaviour of flow over wedge and cone geometries under influence of the Christov heat flux model, inclined magnetic field, porous medium, and convective boundary conditions. Moreover, Deep Neural Networks are also applied for the accurate prediction of heat transfer behaviour. The work aims at making contributions to better thermal management, sustainable energy and industrial use in line with the global sustainable development objectives. Methodology: Governing equations are solved through Lobatto IIIa with the BVP5C solver in MATLAB software and results are represented through graphs and table values. Main findings and conclusion: Both cone and wedge geometries are characterised by a rise in temperature profile as the parameters of the heat source and increased thermal energy creation within the fluid and heat transmission, which is further bolstered by radiative heat transfer, cause radiative heat transfer to grow. Physically, this results in an increase in the fluid's particle internal energy, and a thicker thermal layer raises the temperature distributions in both. Through the Deep Neural Network model the Nusselt number was predicted and Optimisation have been done through the Response Surface Methodology.