A Smart Model for the Prediction of Heat Transfer Coefficient during Flow Boiling of Nanofluids in Horizontal Tube

Abstract

The goal of this study is to improve the accuracy and the validity of the prediction of the heat transfer coefficient (HTC) throughout flow boiling of different water-based nanofluids in a horizontal tube by developing an artificial neural network model using Ag/water, Cu/water, CuO/water, Al2O3/water, and TiO2/water nanofluids. The multiple layer perceptron (MLP) neural network was designed and trained by 354 experimental data points that were collected from the literature. Thermal conductivity of nanoparticle, mass flux, volumetric concentration, and heat flux were used to serve as input variables of the model. The heat transfer coefficient (HTC) was used as the output variable. Via the method of the trial-and error, MLP with 8 neurons in the hidden layer was attained as the optimal artificial neural network structure. This developed smart model is more accordant with the experimental data than the correlations of the literature. The accuracy of the developed smart model was validated by the value of mean squared error (MSE=0.042) and the value of determination coefficient (R2= 0.9992 ) for all data.