A new heat transfer coefficient correlation for pure refrigerants and near-azeotropic refrigerant mixtures flow boiling within horizontal microfin tubes

A new correlation is presented to predict the heat transfer coefficients (HTCs) of pure refrigerants and near-azeotropic refrigerant mixtures undergoing flow boiling within horizontal microfin tubes. This is accomplished by first putting together a 2622-point experimental database from 25 sources. The data includes CO 2, R12, R1234yf, R1234ze(E), R134a, R22, R32, R404A, and R410A, 2.1-14.85 mm fin root diameter tubes, -20 degrees C to 39.9 degrees C saturation temperatures, vapor qualities from 0 to 1, reduced pressures from 0.03 to 0.78, and heat and mass fluxes ranging from 1 to 58.7 kW/m(2) and 25 to 820 kg/s m(2), respectively. The entire database was randomly divided into two subsets, 90% of the data being used to develop the new correlation, with the remaining 10% used to validate it. The correlation was developed in two steps. Thirty-eight unique dimensionless parameters pertinent to flow boiling in microfin tubes were first selected. Multi-variable regression analysis was then applied to identify the most significant variables influencing the flow boiling Nusselt number. The new correlation was evaluated and compared with six extant correlations on an overall basis as well as for several bins of parameters. Overall evaluation for the entire database, for the 10% test data, and for points with experimental uncertainties less than +/- 17% shows that the new correlation is significantly better than any of the extant correlations. For these three overall assessments, the new correlation predicts 71.5-80.5% of the data within +/- 30% error bands, with a mean absolute deviation of 21.5 to 25.2%. Additionally, the distribution of the entire database and the performance of the new correlation was examined relative to refrigerant, and for various bins of fin root diameter, D-r, mass flux, G, all-liquid Reynolds number, Re-al, reduced pressure, P-R, saturation temperature, T-sat, heat flux q, and vapor quality, x. In general, the new correlation shows reasonably good predictions, which are better than those of the existing correlations for most parameter bins, with MAD values generally smaller than 35%. Based on the bin analysis, parameter ranges in which one of the extant correlations gives better predictions than the new correlation are also identified. With these exceptions in mind, the new correlation can be used as a reasonably reliable predictive tool for a large variety of refrigerants under different operating conditions of practical interest. (c) 2017 Elsevier Ltd and IIR. All rights reserved.