Viscosity of CuO nanofluids: Experimental investigation and modelling with FFBP-ANN

Abstract

The current work examines the viscosity of nanofluids as a function of nanoparticle concentration, size, and temperature. CuO nanoparticles with an average size of 15 nm, 45 nm, and 75 nm have been used in this study. Viscosity measurements of volume concentrations of 1% to 4% are carried out in the temperature range of 293 K to 353 K. The results show a maximum increase of 23% in effective viscosity due to the addition of nanoparticles. However, the viscosity decreases by about 80% as the temperature increases from 293 K to 353 K. Furthermore, a maximum increase of 0.5% in viscosity is observed with the particle size increment. Based on present experimental data and literature data, a correlation is developed to predict the effective viscosity. Furthermore, ANN modelling is carried out to predict the effective viscosity of nanofluids and compared with the developed correlation. The results demonstrate that ANN modelling outperforms hypothesised correlation in estimating the effective viscosity. The R-squared values for the ANN model and correlation model are 0.95 and 0.90, respectively.