Thermal Energy Optimization of Nanofluid Movement in Spinning Porous Tubes Through Neural Networks

Abstract

Objective: The current study combines artificial intelligence (AI) with mathematical modeling to analyze nanofluid flow behavior, convective heat transfer, and thermal radiation in a rotating permeable channel. Methods/Analysis: Governing equations are derived for a CuO-water nanofluid using its thermo-physical properties, bulk conditions, and nondimensional variables; the partial differential equations are reduced to ordinary differential equations (ODEs) through similarity transformations, and the resulting boundary value problem (BVP) is solved using a shooting technique coupled with the Runge-Kutta (RK4) method to generate the reference dataset. The Feedforward Multilayer Perceptron (MLP) Artificial Neural Network (ANN), trained with the Levenberg-Marquardt algorithm (LMA), is then developed as an optimized surrogate model. Findings: The effects of the Reynolds number, magnetic parameter, suction/injection, rotation parameter, and Grashof number on the velocity and temperature distributions are interpreted through computational simulations and graphical analysis. The ANN model attains very low mean squared error, near-unity regression, and small gradients, confirming convergence and high predictive accuracy. Novelty: By integrating an ANN surrogate with the rotating permeable-channel nanofluid model, this study contributes to AI-computational methods in fluid mechanics and applied mathematics, offering value to both academic research and engineering practice.