Comparisons of single-phase and two-phase models for numerical predictions of Al2O3/water nanofluids convective heat transfer

Abstract

This paper numerically studies Al2O3/water nanofluids that convectively flow inside the laminar regime of a tube with a constant wall heat flux with various models. Eight different models are utilized to predict the heat transfer behavior of the nanofluids, and the predictions are compared with the available experimental data from literature and the values of the conventional correlation. Comparisons show that all eight models are suitable for the prediction of Al2O3 nanofluids with relatively small particle concentrations (0.25 wt% and 0.5 wt%, mass fractions) of Al2O3/water, within the maximum deviation between the predictions of all models and corresponding experimental data less than 20%. While comparison results with experiments of relatively large particle concentrations (0.6%, 1.0% and 1.6%, volume fractions) show that Mixture model overestimates the heat transfer performance. Discrete phase model increases the prediction accuracy about 10% for two-phase models and agrees well with the classical Shah Equation within the maximum error of 5.5%. The error of Nusselt number between the predictions of discrete phase model and experimental data falls off with the increase of Reynolds numbers and axial direction position. The discrete phase model, Xuan-Roetzel dispersion model, and Talieh-Abbas dispersion model are precise approaches to predict the laminar convectional heat transfer behavior of Al2O3/water nanofluids in the scope of 0–1.6% nanoparticles. The Xuan-Roetzel dispersion model and Talieh-Abbas dispersion model are suggested for applications where the calibration data are available.