Artificial intelligence and numerical simulation based assessment of trihybrid structured flow over a curved geometry: Thermalized case analysis

Abstract

Trihybrid nanofluids exhibit superior heat transfer capabilities as compared to single or binary nanofluids. The synergistic interaction between the various types of nanoparticles results in an increased thermal conductivity as well as heat transfer rate. Trihybrid nanofluids are ideal for heat exchange applications such as solar thermal systems, power plants, heat exchangers, and electronic cooling systems. Our concern in this paper is to incorporate the artificial intelligence (AI) and numerical simulation technique to assess the thermal attributes of trihybrid structured flow over a curved geometry. The nano-composition of gold (Au), single-walled carbon nanotubes (SWCNTs) and, aluminium oxide (Al2O3) make amalgamation in the base fluid H2O to prepare the trihybrid mixture. Two methods are employed to model and analyze the governing system. The first one is Quasi-linearization method (QLM), a numerical technique used to linearize and solve non-linear differential equations. The second method is Bayesian regression neural network (BRNN), a machine learning approach that integrates Bayesian statistics with neural networks to predict outcomes. The results are compared, under limiting conditions, with the earlier ones to validate the model. For the higher curvature parameter, higher will be the velocity and lower will be the temperature in the flow regime. The AI-driven numerical simulation for trihybrid structured flows over different geometries opens new avenues in the engineering applications involving aerodynamics, biomedical devices, and industrial processes.

Keywords

• Bayesian regression neural network
• Curved geometry
• Partial differential equations
• Trihybrid fluid
• Quasi-linearization method
• Numerical results