Development of a neural architecture to predict the thermal conductivity of nanofluids

Abstract

Abstract The present study proposes an artificial neural network (ANN) model for correctly estimating the thermal conductivity property of nanofluids. The ANN model was designed based on using 800 existing experimental data containing spherical nanoparticles of TiO2, ZnO, CuO, Al2O3, ZrO2, Fe2O3, Fe3O4, SiO2, CeO2, MgO, Fe, Al, Cu, Ag, Sic and diamond in various fluids of oil, ethylene glycol, water, and radiator cooling. Here, effective parameters of thermal conductivity of the base fluid and dispersed nanoparticle, nanoparticles volume fraction (0.4 − 0.4%), temperature (10 − 80 ℃), and particle diameter (4 − 150 nm) were considered as input variables, while the thermal conductivity of nanofluid was defined as the target variable. The Levenberg-Marquardt (L-M) back-propagation algorithm was used to design this model. According to the results, the best R and lowest MSE using 5-13-1 topology were founded to be about 0.9975 and 0.000238, respectively, indicating good fitting between predicted results and target points. Also, the results of the comparison between the ANN model and experimental points indicated successful validation of the presented model for estimating the thermal conductivity of nanofluids.