Predicting the effective thermal conductivity of nanofluids for intensification of heat transfer using artificial neural network

Abstract

Nanoparticles are one of the most promising materials with significant capability to increase the heat transfer coefficient and the thermal conductivity of a base fluid, while their addition to the fluid will decrease viscosity and friction factor. Some of the nanoparticles have high thermal conductivity, which candidate them for practical application as nanoadditives in process industries. This work presents a neuromorphic model for predicting the thermal conductivity of nanofluids, which takes into consideration the effects of size, volume fraction, temperature, and thermal conductivity of nanoparticles as well as the properties of base fluids. The presented model is found to correctly predict the trends observed in experimental data for different combinations of nanoparticles-base fluids with varying concentrations. Twenty six different types of nanofluids, namely, Al2O3-water/EG, CuO-water/TO/EG/MEG/paraffin, Cu-water/EG/oil, Al-water/EG/EO/TO, TiO2-water/EG, ZnO-water/EG, SiO2-water/EG/oil, MWCNT-water/EG/oil/R113, and Ag-water are used to assess the effectiveness of the proposed neuromorphic model. The overall predicted effective thermal conductivities regarding twenty six different nanofluids are in excellent agreement with experimental data with the AAD of 3.06% and R2-value of 0.9309.