Appraising influence of COOH-MWCNTs on thermal conductivity of antifreeze using curve fitting and neural network

Curve fitting and neural network modeling are suitable methods for modeling the complex relationship between various parameters in engineering problems. In this study, at the first, a curved fitting was performed on experimental data related to nano-antifreeze containing carbon nanotubes, which led to the presentation of a two-variable correlation to predict its thermal conductivity. After that, an artificial neural network was designed to evaluation of the effects of temperature and solid volume fraction on the thermal conductivity of nano-antifreeze. For modeling, the volume fraction and temperature were applied as input variables. By selecting 9 neurons for the hidden layer, the output of the neural network, which was thermal conductivity ratio, was obtained. The results showed that the proposed equation has good accuracy for engineering applications. However, comparative results showed that the neural network has a more accurate prediction than curve fitting for the thermal conductivity of the antifreeze containing multi walled carbon nanotubes (MWCNTs).