AI-driven modeling of bioconvective nanofluid flow: An ANN approach to anisotropic slip and heat transfer in 3D systems

Abstract

Understanding the dynamics of nanofluids with microorganisms is crucial for biotechnology and thermal engineering applications, as they significantly influence heat and mass transfer processes. This study aims to analyze the behavior of nanofluids containing microorganisms over a three-dimensional moving surface, incorporating the effects of bioconvection, anisotropic slip, and convective boundary conditions. The governing equations are transformed into ordinary differential equations using similarity variables and solved numerically. Additionally, an Artificial Neural Network (ANN) is trained on numerical simulation data to develop a predictive model, enabling rapid and accurate predictions without the need for computational simulations. The results indicate that increasing the Prandtl number (Pr) from 3.2 to 6.2 leads to a 27.4 % reduction in the temperature profile (θ), while the concentration profile (φ) exhibits an inverse trend, increasing by 18.6 % within the boundary layer. Additionally, an increase in the Lewis number (Le) from 2 to 5 enhances thermal diffusivity, resulting in a 14.8 % increase in the thickness of the thermal boundary layer. The close agreement between numerical and ANN predictions validates the model's accuracy, demonstrating the effectiveness of machine learning in capturing complex fluid dynamics. These findings contribute to the optimization of heat and mass transfer processes in nanofluid applications, reducing computational costs while maintaining high precision.