Does the addition of nanoparticles always lead to heat transfer enhancement? Numerical solution enriched by artificial neural network

Abstract

In this study, the effects of loading CuO nanoparticles on heat transfer and pumping power of CuO/liquid paraffin nanofluid were discussed. Although the presence of CuO nanoparticles can increase the thermal conductivity but by increasing the viscosity, the heat transfer may be reduced. Also, the pumping power may not change significantly, depending on the application of the nanofluid and the material for base fluid. Numerical analysis revealed that for NTU greater than five, there is no chance for heat transfer between fluid and tube. In this study, a new figure of merit (FOM) is introduced, and the effect of the presence of nanoparticles in laminar and turbulent regimes at the constant volume/mass flow rate has been examined for CuO/Paraffin nanofluid in tubes. The presence of nanoparticles boosts the density (positive effect) and simultaneously diminishes the specific heat (negative effect). Under the constant mass flow rate, the negative effect of CuO nanoparticles is predominant, and therefore the addition of CuO nanoparticles reduces heat transfer up to 4.8%. A general negative effect of nanoparticles is to increase pumping power. Based on calculations, in the laminar flow regime, the sensitivity of the pumping power to the presence of CuO nanoparticles is higher, so in laminar flow, the pumping power increased more than in the turbulent regime. Defining temperature, mass fraction, and the ratio of nanofluid properties to the base fluid as input variables to the artificial neural network (ANN), it can estimate the FOM to affirm whether adding nanoparticles is useful or not by a maximum error of less than 1%.