Data for: A Single Multilayer Perceptron Model (ANN) for Different Oxide/EG Nanofluid’s Viscosity from the Experimental Study
Description
In this work, different oxide nanofluids Al2O3 (13nm and 50nm), CeO2 (25 nm), CuO (50 nm) of volumetric concentration 0.2%, 0.5%, 0.8%, 1% and 1.5% in ethylene glycol were prepared.The surface morphology test of nanoparticles was done using FESEM and TEM. CTAB in 1:10 of the weight of nanoparticle was used to stabilize the CuO nanofluids and no surfactant was added to stabilize for alumina and cerium oxide nanofluids. Viscosity tests were performed on Anton Paar SVM 3000 Stabinger Viscometer in the increasing temperature of 20 ℃-80 ℃ in 5℃ of temperature interval. The exponential decrease in viscosity and a linear decrease in density were found for all nanofluids. Since these oxide nanofluids have the same tendency of change in viscosity and density, therefore, a single multilayer perceptron artificial neural network modeling was utilized to predict the different oxide based nanofluid viscosity. Volume concentration, temperature, particle type, and size were taken as the input parameter. R2 for training and testing data were 0.9998 and 0.9999. The model was validated by predicting the viscosity results given by the other scientists.