ANFIS modelling of effective thermal conductivity of ethylene glycol and water nanofluids for low temperature heat transfer application

Abstract Nanofluids have been proven to be promising coolants for heat transfer application due to their superior thermal properties. In the present work, experimental studies were carried out to investigate the thermal conductivity of ethylene glycol (EG) and water mixture (35:65 v/v in ratio) based CuO, Al2O3 and TiO2 nanofluids. The study has considered nanofluid containing nanoparticles in the concentration range of 0.2 to 2 wt%. The effect of nanoparticle concentration on thermal conductivity was investigated in terms of effective thermal conductivity (ratio of thermal conductivity of nanofluid to that of base fluid) at low temperatures (5 to 25 °C). Results indicated significant improvement in thermal conductivity at lower temperature of nanofluids. Enhancement of 6.34%, 4.87% and 3.59% in thermal conductivity was obtained for 2 wt% of concentration at 5 °C for CuO, Al2O3 and TiO2 nanoparticles respectively compared to EG:Water. Correlations for effective thermal conductivity were developed by regression considering two input (temperature, concentration) and three input (temperature, concentration and nanoparticles thermal conductivity) parameters using experimental results. An intelligent model, adaptive neuro-fuzzy inference system (ANFIS) approach was used to model the effective thermal conductivity of nanofluids. ANFIS model performed better than the correlations developed.