Using Perceptron Feed-Forward Artificial Neural Network (ANN) for Predicting the Thermal Conductivity of Graphene oxide-Al2O3/Water-Ethylene Glycol Hybrid Nanofluid

Shaopeng Tian, Noreen Izza Arshad, Davood Toghraie, S. Ali Eftekhari, Maboud Hekmatifar
2021 Case Studies in Thermal Engineering  
A B S T R A C T In this paper, Artificial Neural Network (ANN) was used to investigate the influence of temperature and volume fraction of nanoparticles on the thermal conductivity of Graphene oxide-Al 2 O 3 / Water-Ethylene glycol hybrid nanofluid. Nanofluids were prepared with the volume fraction of nanoparticles 0.1, 0.2, 0.4, 0.8, and 1.6% in the temperature range of 25-55 • C. The nanofluid's thermal conductivity results were extracted from six different volume fractions of nanoparticles
more » ... d seven different temperatures. Then, to generalize the data and obtain a function, the Perceptron feed-forward ANN was used, simulating the output parameter. The outcomes show that the ANN is well trained using the trainbr algorithm and has an average of 1.67e-6 for MSE and a correlation coefficient of 0.999 for thermal conductivity. Finally, we conclude that the effect of increasing the temperature of nanofluid is less against the volume fraction of nanoparticles, especially in low concentrations. This effect is negligible and in the absence of nanoparticles, increasing the temperature from 20 • C to 55 • C leads to an enhance in thermal conductivity of about 6%. However, at high concentrations of nanoparticles, increasing the temperature leads to further thermal conductivity. At volume fraction nanoparticles 1.6%, increasing the temperature from 20 • C to 55 • C increases the thermal conductivity from 0.45 to 0.54 W/m.K.
doi:10.1016/j.csite.2021.101055 fatcat:v3cl3hrksbecjc4y2xda4ebubq