Experimental study on saturated flow boiling heat transfer of R290/R152a binary mixtures in a horizontal tube

Abstract

An experimental study on the saturated flow boiling heat transfer for a binary mixture of R290/R152a at various compositions is conducted at pressures ranging from 0.2 to 0.4 MPa. The heat transfer coefficients are experimentally measured over mass fluxes ranging from 74.1 to 146.5 kg/(m²·s) and heat fluxes ranging from 13.1 to 65.5 kW/m². The influences of different parameters such as quality, saturation pressure, heat flux, and mass flux on the local heat transfer coefficient are discussed. Existing correlations are analyzed. The Gungor-Winterton correlation shows the best fit among experimental data for the two pure refrigerants. A modified correlation for the binary mixture is proposed based on the authors’ previous work on pool boiling heat transfer and the database obtained from this study. The result shows that the total mean deviation is 10.41% for R290/R152a mixtures, with 97.6% of the predictions falling within±30%.

Keywords

flow boiling / heat transfer / binary mixture / R290/R152a

Introduction

Searching for new environmentally friendly refrigerants has been one of the most urgent and important tasks in the refrigeration and air conditioning industry because hydrocarbons containing chlorine (CFCs and HCFCs) have been found to be harmful to the environment. Most hydrocarbons have both an ozone layer depletion effect and a high global warming potential (GWP). Pure propane (R290) and its mixtures are considered good alternative refrigerants for applications in refrigeration, air conditioning, and heat pumps. The 1,1-difluoroethane (R152a) is a good hydrofluorocarbon (HFC) refrigerant because it has the lowest GWP among commercialized HFC refrigerants. It is already widely used in the refrigeration and air conditioning field as a pure refrigerant or as a component in mixed refrigerants. Thus, the new mixture of R290/R152a has a small GWP and no ozone depletion effect. It has been proposed as an alternative to the traditional refrigerants of R22 and R404a [1]. The accurate prediction of the flow boiling heat transfer coefficient in the heat exchanger is one of the most important tasks in alternative refrigerant research. Therefore, it is desirable to study the heat transfer characteristics and predict the flow boiling heat transfer coefficients of R290/R152a.

Limited data on the local flow boiling heat transfer coefficients of the mixture with different compositions are available in open published literature. Jung et al. [2] have evaluated the flow boiling heat transfer coefficients of the azeotropic mixture R12/R152a in a horizontal stainless steel tube with a 9-mm inner diameter. In this paper, saturated flow boiling heat transfer coefficients in a horizontal tube of the new binary azeotropic mixture R290/R152a are measured. Based on measured data, a new correlation is proposed for pure refrigerants and binary mixtures.

Experimental apparatus and procedure

The schematic diagram of the experimental apparatus is shown in Fig. 1. The circulation of the tested fluid is driven by a self-designed magnetic pump. A Coriolis mass flow meter is installed behind the circulation pump to measure the mass flow rate. An electrical preheater is fixed before the inlet of the test section to obtain the required inlet quality. The test section has five test units. Each test unit is a 120 mm long copper tube with an inner diameter of 8 mm and an outer diameter of 40 mm. The five copper tubes are connected by four 60 mm long stainless steel tubes. The inner diameter of each stainless steel tube is 8 mm, similar to the testing copper tube, while the thickness is 1 mm to eliminate axial heat conduction. The copper tubes and stainless steel tubes are well welded together by vacuum brazing to ensure that the inner flow passage at the connection acts as one whole tube. The DC electric heating wire coil is tightly twisted on the outer surface of the copper tube, which provides the Joule heating effect, to obtain the required heat flux by adjusting the DC voltage. Two quartz glass tubes with an inner diameter of 8 mm are installed at the inlet and outlet of the test section for flow pattern visualization.

Table 1 summarizes the measuring ranges and uncertainties of the measurements. The uncertainties of heat transfer coefficients range from 7.6% to 15.4%, which were calculated using the guidelines suggested by the NIST [3]. All instruments were carefully calibrated.

The local flow boiling heat transfer coefficient is defined as

(1)h={\displaystyle \frac{q}{T_{\mathrm{w}}-T_{\mathrm{s}\mathrm{a}\mathrm{t}}}},

where q indicates the inner wall heat flux of the tube and can be obtained from the measured total heat input and heat loss. Heat losses to the environment and the stainless steel tube were determined through a special series of experiments in which there was no fluid in the test section. T_w is the local inner wall temperature and is calculated from the measured wall temperature by applying the one-dimensional, radial, steady-state heat conduction equation for a hollow cylinder. T_sat is the saturation temperature of the fluid and is assumed to be the average thermodynamic equilibrium temperature. Details of this experimental apparatus and heat transfer data reduction can be found in Ref. [4].

Results and discussion

Figure 2 presents the vapor-liquid phase diagram of the R290/R152a mixture. This binary mixture shows the azeotropic phase behavior at a certain concentration, at which the mixture behaves like a pure substance. Therefore, it is likely to exhibit very little heat transfer degradation.

Heat transfer coefficients versus different qualities, saturation pressures, mass fluxes, and heat fluxes are illustrated in Figs. 3 to 6. The results were obtained over the Reynolds number ranging from 2215 to 6863. Figure 3 demonstrates the dependence of the heat transfer coefficient on quality at eight different compositions. There is a general trend of increasing heat transfer coefficient with increasing quality for all pure substances and mixtures. The heat transfer coefficients of R290 and R152a reached higher values than those of the mixtures. The heat transfer coefficients of the mixtures at mole fraction 0.74 of R290 are higher than those of other compositions because the composition is close to the azeotropic point.

The local heat transfer coefficients of the azeotropic mixture R290/R152a are plotted as a function of the mole fraction of R290 for three heat fluxes in Fig. 4. The heat transfer coefficients of the mixtures are significantly lower compared with those of the pure components. The heat transfer coefficients of R290/R152a mixtures decrease if the concentration difference of the liquid and vapor increases (Fig. 4). At the same time, the heat transfer degradation becomes larger at higher heat flux. This phenomenon has also been observed in other mixture measurements, which is mainly caused by mass transfer resistance [2]. In the azeotropic composition range, the values of temperature glide and concentration glide are quite small where the mass transfer effect is greatly reduced. The mixture only behaves as a pure refrigerant. However, the heat transfer coefficients of the R290/R152a mixtures near the azeotropic point are slightly higher than those of pure R290 and R152a at high heat flux.

The influence of saturation pressure on the local heat transfer coefficient is presented in Fig. 5 at a mass flux of 96.2 kg/(m²·s) and a heat flux of 39.3 kW/m². The different saturation pressure causes the change not only of the mixture property, but also of the equilibrium characteristic. The flow boiling heat transfer coefficients for pure substances and mixtures increase with increasing pressure.

Figure 6 illustrates the variation of the local heat transfer coefficient as a function of mass flux at a constant saturation pressure of 0.4 MPa and a heat flux of 39.3 kW/m². The dashed curves in Fig. 6 represent the ideal heat transfer coefficient h_i, which is defined using the linear mole fraction method as

(2)h_i={\displaystyle \frac{\mathrm{1}}{x/h_{\mathrm{1}}+(\mathrm{1}-x)/h_{\mathrm{2}}}}={\displaystyle \frac{h_{\mathrm{1}}{h}_{\mathrm{2}}}{(\mathrm{1}-x)h_{\mathrm{1}}+x{h}_{\mathrm{2}}}}.

The flow boiling heat transfer coefficients for the R290/R152a mixture are lower than the ideal values, except for the azeotrope, at which they are very close to the ideal value. The relative heat transfer coefficient ratios h_tp/h_i are 0.81, 0.85, and 0.87, respectively, when mass flux G = 96.2, 110.6, and 129.4 kg/(m²·s). Heat transfer degradation decreases with increasing mass flux. For an azeotropic mixture, it is reasonable to assume that there would be no mixture effect on heat transfer at the azeotropic concentration. However, for other concentrations apart from the azeotrope, the mixture effects would play an important role in heat transfer degradation similar to that in typical non-azeotropic mixtures. Thus, the heat transfer coefficient of the binary mixtures in the region of non-azeotropic composition is significantly affected by mass flux.

The flow patterns of the fluid were obtained from direct visual observations using a sight glass located at the end of the horizontal test section and then recording by a camera. The flow patterns observed in the experiment are wavy flow, semi-annular flow, and annular flow. Figures 7 (a)-(c) respectively show photographs that are representative of the observed flow patterns. The flow patterns are plotted in the coordinates of G and x for pure substance R290 in Fig. 8 (a), and for the R290/R152a mixture in Fig. 8 (b). The flow pattern would be wavy, semi-annular, and annular in the direction of increasing G and x. For the R290/R152a mixture, the transitions happen at higher G and x compared with those for R290. Thus, at low mass flux, the flow pattern of pure substance could always become annular while the pattern of mixture is still wavy under the same conditions.

Correlation development

The experimental heat transfer coefficients of the two pure refrigerants were compared with the predicted results using well-known correlations such as Gungor-Winterton [5], Liu-Winterton [6], Shah [7], Kandlikar [8], and Kew-Cornwell [9]. The detailed results of comparisons are shown in Fig. 9. The mean absolute deviations (MAD) of the correlations and the experimental data are summarized in Table 2, which is defined as

(3)\mathrm{M}\mathrm{A}\mathrm{D}={\displaystyle \frac{\mathrm{1}}{N}}{\displaystyle \sum {\displaystyle \frac{\left|h_{\mathrm{p}\mathrm{r}\mathrm{e}}-h_{\exp }\right|}{h_{\exp }}}}\times \mathrm{100}\%.

Among the results, the Gungor-Winterton correlation [5] showed the best agreement. The total MAD is 13.91%. Acceptable results were also obtained with the Liu-Winterton [6] and Kew-Cornwell [9] correlations. The Gungor-Winterton [5] correlation can be expressed as

(4)h_{\mathrm{t}\mathrm{p}}=E{h}_{\mathrm{s}\mathrm{p}}+S{h}_{\mathrm{n}\mathrm{b}}.

The heat transfer coefficient is expressed as the sum of a nucleate boiling term with a suppression factor S and of a convective evaporation term with an enhancement factor E. h_sp has been given by the Dittus-Boelter equation for convective flowing in the channel.

(5)h_{\mathrm{s}\mathrm{p}}=\mathrm{0.023}{(\mathit{R}{\mathit{e}}_{\mathrm{l}})}^{\mathrm{0.8}}{(\mathit{P}{\mathit{r}}_{\mathrm{l}})}^{\mathrm{0.4}}{\displaystyle \frac{k_{\mathrm{l}}}{d_{\mathrm{h}}}},

(6)\mathit{R}{\mathit{e}}_{\mathrm{l}}={\displaystyle \frac{G{d}_{\mathrm{h}}(\mathrm{1}-x)}{{\mu }_{\mathrm{l}}}},

(7)E=\mathrm{1}+\mathrm{24000}B{o}^{\mathrm{1.16}}+\mathrm{1.37}{X}_{\mathrm{t}\mathrm{t}}^{-\mathrm{0.86}},

(8)X_{\mathrm{t}\mathrm{t}}={\left({\displaystyle \frac{\mathrm{1}-x}{x}}\right)}^{\mathrm{0.9}}{\left({\displaystyle \frac{{\rho }_{\mathrm{l}}}{{\rho }_{\mathrm{v}}}}\right)}^{\mathrm{0.5}}{\left({\displaystyle \frac{{\mu }_{\mathrm{l}}}{{\mu }_{\mathrm{v}}}}\right)}^{\mathrm{0.1}},

(9)h_{\mathrm{n}\mathrm{b}}=\mathrm{55}{p}_{\mathrm{r}}^{\mathrm{0.12}}{(-{\log }_{\mathrm{10}}{p}_{\mathrm{r}})}^{-\mathrm{0.55}}{M}^{-\mathrm{0.5}}{q}^{\mathrm{0.67}},

(10)S={(\mathrm{1}+\mathrm{1.15}\times {\mathrm{10}}^{-\mathrm{6}}{E}^{\mathrm{2}}\mathit{R}{\mathit{e}}_{\mathrm{l}}^{\mathrm{1.17}})}^{-\mathrm{1}}.

E is the forced convective heat transfer enhancement factor and S is the suppression factor. X_tt is the Lockhart-Martinelli parameter. h_nb is the nucleate boiling coefficient and calculated from Copper’s pool boiling correlation [10]. p_r is the reduced pressure and M is the molecular weight in Eq. (9).

A number of predictive methods or correlations have been developed to predict the heat transfer coefficient of mixtures. In this study, the previous research of our group [11] on pool boiling heat transfer was employed and modified. The equations of the pool boiling heat transfer coefficient of the relative refrigerant mixtures can be expressed as follows:

(11)K={\displaystyle \frac{h_m}{h_i}}={\left\{\mathrm{1}+{\displaystyle \frac{\mathrm{\Delta }{T}_{bp}}{\mathrm{\Delta }{T}_{id}}}{\left|y-x\right|}^{C_{\mathrm{1}}}{\left({\displaystyle \frac{p}{{\mathrm{10}}^{\mathrm{5}}}}\right)}^{C_{\mathrm{2}}}\left[\mathrm{1}+C_{\mathrm{3}}\exp \left(-{\displaystyle \frac{q}{\mathrm{3}\times {\mathrm{10}}^{\mathrm{5}}}}\right)\right]\right\}}^{-\mathrm{1}},

(12)\mathrm{\Delta }{T}_{db}=T_d-T_b,

(13)\mathrm{\Delta }{T}_{id}={\displaystyle \frac{q}{h_i}},

where ΔT_id is the ideal temperature difference and ΔT_db is the boiling range. h_i is the ideal heat transfer coefficient of the refrigerant mixtures, which is defined in Eq. (6). h₁ and h₂ are the heat transfer coefficients for R290 and R152a, respectively, which are calculated from the Gungor-Winterton correlation [5]. The values of C₁-C₃ in Eq. (11) are empirical constants and obtained by an iteration process to minimize the errors between the heat transfer coefficient calculated from the above correlations and the experimental results. These values are presented in Table 3.

K can be incorporated into the correlation for mixture. Thus, the Gungor-Winterton correlation [5] can be modified as

(14)h_{tp,m}=E{h}_{sp}+KS{h}_{nb}

K equals 1 for pure refrigerant in Eq. (14). Therefore, Eq. (14) can be used for predicting the saturated flow boiling heat transfer coefficients of both pure refrigerants and refrigerant mixtures.

Figure 10 illustrates the comparison between the experimental data and the prediction value of the correlations for R290/R152a as proposed by Jung et al. [12] and Schlünder [13]. The correlation predictions of Schlünder [13] are higher than the experimental values. Table 4 presents a comparison of the MADs between predicted and experimental values of the refrigerant mixtures. Among these, the predicted result of the present work is the best, with a MAD of 10.41%. Moreover, approximately 97.6% of the prediction values fall within±30%.

Conclusion

Saturated flow boiling heat transfer coefficients in a horizontal tube of the binary azeotropic mixture R290/R152a were measured. Experimental results confirmed that the mixture effects on heat transfer would be absent at the azeotropic composition, therefore exhibiting very little heat transfer degradation.

Analysis of the experimental data of R290/R152a mixtures showed that the degradation of flow boiling heat transfer decreases as mass flux increases. The transitions of flow pattern occur at a higher mass flux and quality for mixtures than those for pure substances because the mass transfer resistance for mixtures can prevent more wetted portions and promote annular flow. Therefore, the heat transfer is decreased.

The Gungor-Winterton correlation [5] shows a better prediction result compared to the other four existing correlations for pure substances. Based on this correlation, a modified correlation was proposed, which is corrected by the previous research [11] on pool boiling heat transfer and the experimental data obtained from this study. The new modified correlation gives the best fit and predicts heat transfer coefficients within a 10.41% deviation. Moreover, 97.6% of the predictions fall within±30% of the experiments. Therefore, it can be used for predicting the saturated flow boiling heat transfer coefficients of pure refrigerants and mixtures.

References

[1] Park Y, Kang J, Choi J, Yoo J K, Kim H. Vapor-liquid equilibria for the 1,1-difluoroethane (HFC-152a) + propane (R-290) system. Journal of Chemical and Engineering Data, 2007, 52(4): 1203-1208

[2] Jung D S, McLinden M, Radermacher R, Didion D. A study of flow boiling heat transfer with refrigerant mixtures. International Journal of Heat and Mass Transfer, 1989, 32(9): 1751-1764

[3] Taylor B N, Kuyatt C E. Guidelines for Evaluating and Expressing the Uncertainty of NIST Measurement Results. NIST Technical Note 1297, 1994 Edition, 1994

[4] Zou X, Gong M Q, Chen G F, Sun Z H, Zhang Y, Wu J F. Experimental study on saturated flow boiling heat transfer of R170/R290 mixtures in a horizontal tube. International Journal of Refrigeration, 2009, 32(2): 371-380

[5] Gungor K E, Winterton R H S. A general correlation for flow boiling in tubes and annuli. International Journal of Heat and Mass Transfer, 1986, 29(3): 351-358

[6] Liu Z, Winterton R H S. A general correlation for saturated and subcooled flow boiling in tubes and annuli, based on a nucleate pool boiling equation. International Journal of Heat and Mass Transfer, 1991, 34(11): 2759-2766

[7] Shah M M. Chart correlation for saturated boiling heat transfer: equations and further studies. ASHRAE Transactions, 1982, 88(1): 185-196

[8] Kandlikar S G. An improved correlation for predicting two-phase flow boiling heat transfer coefficient in horizontal and vertical tubes. In: Kitto J B J, ed. 21th ASME/AIChE national heat transfer conference. ASME, Seattle, 1983

[9] Kew P A, Comwell K. Correlations for the prediction of boiling heat transfer in small diameter channels. Applied Thermal Engineering, 1997, 17(8-10): 705-715

[10] Copper M G. Saturation nucleate pool boiling: a simple correlation. In: 1st UK Natl. Heat Transfer Conf (I Chem E Symp series No. 86). 1984, 2: 785-793

[11] Sun Z H, Gong M Q, Li Z J, Wu J F. Nucleate pool boiling heat transfer coefficients of pure HFC134a, HC290, HC600a and their binary and ternary mixtures. International Journal of Heat and Mass Transfer, 2007, 50(1,2): 94-104

[12] Jung D S, McLinden M, Radermacher R, Didion D. Horizontal flow boiling heat transfer experiments with a mixture of R22/R114. International Journal of Heat and Mass Transfer, 1989, 32(1): 131-145

[13] Niederkrüger M, Steiner D, Schlünder E U. Horizontal flow boiling experiments of saturated pure components and mixture of R846/R12 at high pressures. International Journal of Refrigeration, 1992, 15(1): 48-58

RIGHTS & PERMISSIONS

Higher Education Press and Springer-Verlag Berlin Heidelberg