Optimization and sensitivity analysis of magnetic fields on nanofluid flow on a wedge with machine learning techniques with joule heating, radiation and viscous dissipation

Abstract

Purpose

In this study, the flow of nanofluids (NFDs), consisting of water and copper nanoparticles over a wedge, is simulated. The analysis considers the effects of a magnetic field (MFD) and Joule heating (JOH). Variables such as nanoparticle volume fraction (NVF), Eckert number (EC), radiation, and wedge angle (BT) are also examined for their impacts on the Nu and C_f.

Design/methodology/approach

The simulation utilizes the similarity method and the Keller box method, implemented through custom coding. Additionally, machine learning techniques are applied for sensitivity analysis and optimization of the results by varying the parameters.

Findings

The findings indicate that increasing the BT, NVF and MFD strength can elevate the average friction coefficient (Cf-m) by up to 42.8 %. Sensitivity analysis reveals that factors like BT and MFD significantly influence the Cf-m and Nu. An increase in MFD strength generally reduces the Nu-m. A larger BT substantially boosts the Nu-m; however, heightened JOH results in a sharp decline in the Nu. An increase in the EC leads to a decrease in the Nu-m. At low radiation parameter (RD) values, increasing this parameter reduces the Nu-m, whereas at higher values, it increases the Nu.

Originality/value

The key contribution of the article is the optimization and sensitivity analysis of NFD flow over a surface, considering the effects of a MFD, JOH, radiation, EC, and BT. This is done to achieve maximum heat transfer and minimum friction loss.