Experimental Investigation with ANN Modeling of Thermal Conductivity and Viscosity of a Ternary Nanofluid at Different Mixing Ratios and Volume Concentrations

An experimental investigation of thermal conductivity and viscosity of a new ternary nanofluid with four different mixing ratios and three different volume concentrations is conducted for different temperatures. Graphene nanoplatelets, copper oxide and alumina nanoparticles are used in the preparation of the ternary nanofluid with water-ethylene glycol mixture as the base fluid. The thermal conductivity results reveal the predominant contribution of graphene nanoplatelets in the particle mix. The maximum enhancement in thermal conductivity is found to be 19% at 0.5% volume concentration. Increasing the nanofluid concentration beyond 0.3% gives only a marginal enhancement in terms of figure of merit. Artificial neural network is also designed to predict the thermal conductivity and the viscosity. Seventy-eight sets of experimental data are utilised to build the ANN model and an overall regression data of 0.99496 is obtained indicating good agreement between the ANN model and the present experimental data.