Modeling of thermal diffusivity of nanofluids using artificial neural network
Fakhri Yousefi and Nadieh Parsazadeh
An artificial neural network model has successfully been utilized to compute the thermal diffusivity of nanofluids such as Al2O3/Water, TiO2/ Water, CuO/Water, CNT/Water, Au/ Water, Au/ EG and Au/ Ethanol as a function of the temperature, thermal diffusivity of nanoparticle, volume fraction of nanoparticle, diameter of nanoparticle and the thermal diffusivity of base fluids. The mean square error (MSE) was chosen as a judge of the performance of the network. The trained network presents good agreement between the predicted and the literatures with AAD = 1.13% and high coefficient of determinations, R2 = 0.996. Besides, this work is compared with other models.
Keywords: Artificial neural networks, Thermal diffusivity, Nanofluid, Nanoparticle