Estimation of pool boiling heat transfer coefficient of alumina water-based nanofluids by various artificial intelligence (AI) approaches

Abstract

Quick transportation of the latent heat during the boiling leads to efficient application of this process in various heat transfer equipments. Boiling heat transfer coefficient is one of the most affecting parameters on the amount of transferred energy by boiling process. Recently, nanofluids have been widely applied for improving the performance of boiling process. Despite the large amount of experimental studies on the pool boiling heat transfer coefficient (PBHTC) of nanofluids, there’s developed no accurate mathematical/empirical approaches for estimating this parameter up to now. Since the Al2O3 nanoparticle is a safe material, can be simply produced in large scale, and its dispersion in base fluids often have an excellent stability, alumina-based nanofluids have received high attention in various thermal processes. Therefore, in this study, pool boiling heat transfer coefficient of water-alumina nanofluid is tried to be predicted using artificial neural networks. Correlation matrix analyses confirm that diameter of nanoparticles, its weight concentration in base fluid, excess temperature (wall superheat), and operating pressure are the best independent variables for estimating the considered parameter.Different types of the artificial neural networks are designed and their predictive accuracies are evaluated by experimental data. Thereafter, the best model is selected through comparing the predictive accuracy of various intelligent approaches.The results show that a feedforward multi-layer perceptron (MLP) network with the structure of 4-12-1 (i.e. twelve hidden neurons) is the best model for estimation of the PBHTC of water-alumina nanofluid. This model was able to predict the considered coefficient with overall MSE=4.17, AARD%=9.53, R2=0.9929, RMSE=2.042. Our proposed approach can be simply employed to control boiling-based equipments in a variety of industrial applications, and optimize their operations.