Artificial neural network modeling for prediction of thermal conductivity of functionalized MWCNTs/water nanofluids and a new empirical correlation

Abstract

In this paper, based on experimental data, by employing regression method and artificial neural network, the effects of temperature and nanotubes concentration on thermal conductivity of COOH functionalized Multi Walled Carbon Nano Tubes / water was investigated . A very accurate correlation for thermal conductivity ratio was suggested as a function of temperature and solid volume fraction . Artificial neural network modeling was performed. Temperature and solid volume fraction were employed as input variables and thermal conductivity ratio was used as outputs variable. Optimized ANN by considering minimum prediction error was obtained. Comparisons showed that the ANN can more precisely predict the thermal conductivity ratio of COOH functionalized Multi Walled Carbon Nano Tubes/water nanofluids. The results also revealed that the empirical correlation has an acceptable accuracy.Experimental results showed that the thermal conductivity has a direct and reverse relationshipThe existing correlations in literature were unable to predict viscosity data, hence, a new correlation has been proposed.

Keywords

• modeling
• neural network
• thermal conductivity
• empirical correlation
• nanofluid
• mwcnt