Experimental investigation of thermal conductivity and its ANN modeling for glycol-based Ag/ZnO hybrid nanofluids with low concentration

A novel glycol-based silver-coated zinc oxide (Ag/ZnO) hybrid nanofluids has been synthesized to investigate its thermo-physical properties with low volume concentrations (phi =0.05% to phi =0.2%) in comparison with ZnO nanofluids. The Ag/ZnO nanoparticles of size 25-40 nm are synthesized by chemical precursor method. The synthesized nanoparticles were characterized by scanning electron microscopy, transmission electron microscopy and X-ray diffraction analysis. The stability of the dispersed Ag/ZnO hybrid nanoparticles in glycol was observed visually and by zeta potential values. The thermal conductivity of the prepared nanofluids was measured experimentally by KD2 Pro, and maximum enhancement of 15.66% and 20.53% is reported at 25 degrees C and 55 degrees C, respectively, due to the coating of higher thermal conductivity Ag on the ZnO nanoparticles. An empirical correlation has been developed by the curve fitting technique to predict the thermal conductivity enhancement at different concentrations and temperatures. Also, experimentally measured results have been compared with the model developed by an artificial neural network. The correctness of the model has been evaluated by the R-squared value, mean square error, and average absolute relative deviation percent. Based on an experimental data test set of 104 samples, the ANN model consisting of 2 hidden layers with 10 neurons in each layer was the most accurate fit.