Analysis for enhanced heat transfer in natural convection modified nanofluid with shaped optimized nanoparticles

Abstract

Specific shapes and sizes of nanoparticles can be acquired for numerous important applications. The shape of nanoparticles is prominent in all fields of trace metal detection, for molecular labeling, bio-molecular assays and nano-technology utilization. The current study presents the shape effect of nanoparticles in natural convection (Cu-Al2O3-TiO2)/H2O modified nanofluid flow in a lid driven square cavity. The water (H2O) based modified nanofluid contained three categories of nanoparticles (Cu-Al2O3 and TiO2) with spherical (sphere) and non-spherical (column & lamina) shaped. The equations of current model are handled with the finite element method (FEM). For comparisons and accuracy of the study the convolutional neural network (CNN), and convolutional neural network (CNN) with bidirectional gated recurrent unit (BiGRU) model are applied. The characteristics of modified nanofluid are illustrated and explained in terms of streamlines, isotherm plots, velocity and temperature, kinetic energy and heat transfer. The biggest accuracy in the comparison is observed. The outcomes show that at phi = 0.02 the heat transfer rate is highest in the existence of lamina-shaped nanoparticles. The lamina-shaped nanoparticles outperform spheres by 50.71 % and columns by 32.34 %, demonstrating a significantly higher heat transfer rate.