Thermal conductivity of non-Newtonian nanofluids: Experimental data and modeling using neural network

Three different types of nanofluids were prepared by dispersing γ-Al2O3, TiO2 and CuO nanoparticles in a 0.5 wt% of carboxymethyl cellulose (CMC) aqueous solution. Thermal conductivity of the base fluid and nanofluids with various nanoparticle loadings at different temperatures were measured experimentally. Results show that the thermal conductivity of nanofluids is higher than the one of the base fluid and the increase in the thermal conductivity varies exponentially with the nanoparticle concentration. In addition to increase with the nanoparticle concentration, the thermal conductivity of nanofluids increases with the temperature. Neural network models were proposed to represent the thermal conductivity as a function of the temperature, nanoparticle concentration and the thermal conductivity of the nanoparticles. These models were in good agreement with the experimental data. On the other hand, the Hamilton Crosser model was only satisfactory for low nanoparticle concentrations.