Erratum: “Water-Heated Pool Boiling of Different Refrigerants on the Outside Surface of a Horizontal Smooth Tube” [ASME Journal of Heat Transfer, 2012, 134, 021502]

This discussion aims to provide the boiling heat transfer coefficients when the Petukhov and Popov correlation is used to calculate the water-side convection heat transfer coefficients. For each refrigerant, the water-side flow conditions are kept nearly the same during the experiments, which covers a range of Reynolds numbers of 30,000 < Re < 58,000 and Prandtl numbers of 5.6 < Pr < 9.5; under these flow conditions, the predictions from the Petukhov and Popov correlation are 10.3–14.3% higher than those from the Sieder-Tate correlation. As a result, the boiling heat transfer coefficients derived by using the Petukhov and Popov correlation are lower than those derived by using the Sieder-Tate correlation by a degree that depends on the refrigerants due to their different boiling heat transfer coefficients on the outside surface. Figure 1 shows the comparison for the five refrigerants, where h_{o_PP} and h_{o_ST} are, respectively, the boiling heat transfer coefficients when the Petukhov and Popov correlation and the Sieder-Tate correlation are used. It is noted that the largest disparities, approximately 22%, are for R22 or R134a at the highest heat fluxes. For refrigerants R236fa and R245fa, the disparities are much smaller due to their lower boiling heat transfer coefficients in relation to those of R22 and R134a. For the refrigerant R123 that has the lowest boiling heat transfer coefficients, the disparity is less than 5% due to the dominance of overall heat transfer resistance on the refrigerant side.

Figure 2 shows the boiling heat transfer coefficients of the five refrigerants when the Petukhov and Popov correlation is used to calculate the water-side convection heat transfer coefficients. Also plotted in the figure are the predictions from the correlation proposed on the basis of the electrically heated boiling data by Jung et al. [(5)]. A comparison indicates that the electrically heated boiling correlation underpredicts the water-heated boiling performances by 25–40% depending on the refrigerant and the heat flux. Figure 3 is plotted to compare with the predictions from the Cooper correlation [(6)]. Except for R123, the Cooper correlation predicts reasonably well the boiling heat transfer coefficients for other refrigerants (+3% to +24% for R245fa, −3% to −5% for R134a, −13% to +13% for R236fa, and +10% to +18% for R22). For R123, the Cooper correlation overpredicts by 17–77%; larger disparities occur at lower heat fluxes. Table 1 shows the disparities between the predictions from four other commonly used boiling correlations [(7,8,9,10)] and the present measurements. When the Sieder-Tate correlation was used, the predictions from the Gorenflo correlation and the measurements have a consistent disparity of 30% over entire heat fluxes for all the five refrigerants. In the case when the Petukhov and Popov correlation is used, the magnitude of disparities at higher heat fluxes for R134a and R22 are less than 30% (Table 1). As in Ref. [1], an attempt is also made here to multiply the predictions from the Gorenflo correlation by a constant to fit the measurements. A trial-and-error procedure led to a multiplier constant of 1.31. Figure 4 shows the predictions from the modified Gorenflo correlation (that is, the Gorenflo correlation multiplied by 1.31) and the measurements, with the disparities given in Fig. 5 for the five refrigerants. In Fig. 5, h_{o_pred} denotes the boiling heat transfer coefficients predicted by the modified Gorenflo correlation, and h_{o_meas} denotes the measured boiling heat transfer coefficients. The predictions from the modified Gorenflo correlation agree with the present measurements within ±20%.