Experimental study for predicting the specific heat of water based Cu-Al2O3 hybrid nanofluid using artificial neural network and proposing new correlation

In this study, an artificial neural network model has been created in order to estimate the specific heat of Cu-Al2O3/water hybrid nanofluid based on temperature (T) and volume concentration (phi). Specific heat values of the Cu-Al2O3/water hybrid nanofluid prepared in five-volume concentration were measured experimentally in the 20 degrees C to 65 degrees C temperature range. The dataset was reserved into three primary parts, with the inclusion of 901 (70%) for the training, 257 (20%) for the test and 129 (10%) for the validation. As a result of comparison with experimental values, it is concluded that this model predicts specific heat with R-value of 0.99994 and an average relative error of approximately 5.84e-9. In addition, a mathematical correlation has been developed to estimate the specific heat of the Cu-Al2O3/water hybrid nanofluid. The data acquired from the mathematical correlation, developed, were in great correlation with all the experimental values with an average deviation of -0.005%. This result has revealed that the developed mathematical correlation is an ideal design for estimating the specific heat of the Cu-Al2O3/water hybrid nanofluid.