The effect of a boiling additive on R123 condensation on a vertical integral finsurface.docx

1、The effect of a boiling additive on R123 condensation on a vertical integral fin surfaceThe effect of a boiling additive on R123 condensation on a vertical integral fin surfaceAbstractThis paper examines the effect of the addition of 0.5% mass isopentane to R123 on the vapor-space condensation heat

2、transfer of R123. In a previous study, the pool boiling performance of R123 was improved by adding 0.5% mass isopentane. Consequently, the impetus of the present study was a desire to quantify the consequence of the boiling additive on the condensation heat transfer performance of pure R123. In this

3、 way, the net effect of the additive on the cycle performance of pure R123 can be estimated. The data consisted of the heat flux and the wall temperature diference measurement for pure R123 and R123/isopentane (99.5/0.5) on an integral-trapezoidal-fin surface. The tem- perature of the saturated vapo

4、r was held constant at 313.15 K for all of the tests. On average, the R123/isopentane mixture exhibited a 4% smaller heat flux than that of pure R123. Presumably, the degradation was caused by the zeotropic behavior of the mixture, which led to a loss of available driving temperature diference for h

5、eat transfer across the liquid film .Considering that the boiling performance was enhanced on average by 10% with the addition of 0.5% mass isopentane, isopentane may still be a viable means of improving the cycle performance of R123 despite the 4% condensation heat transfer degradation. # 2000 Else

6、vier Science Ltd and IIR. All rights reserved.kyewords: Heat transfer; Mass transfer; Condensation; Refrigerant; R123; Additive; Heat transfer coefficient; Surface; Finned tube.IntroductionFor the refrigeration and air-conditioning industry, a liquid additive would be an economical means to reduce m

7、anufacturing and/or operating costs. For example, a liquid additive for 1,1-dichloro-2,2,2-tri-uoroethane (R123) would enable existing water chillers to operate more effciently or enable new water chillers to meet the same duty with fewer tubes. However, the economic benefit of additives that enhanc

8、e boiling heat transfer can be realized only when the additive does not significantly degrade the condensation heat transfer.Kedzierski 1 measured a significant enhancement of R123 pool boiling with the addition of 1 and 2% hexane by mass to R123. He used the Gibbs adsorption equa- tion and the Youn

9、g and Dupre equation to speculate that the boiling heat transfer enhancement of R123 by the addition of hexane was caused by an accumulation of hydrocarbon at the boiling surface. In essence, the greater concentration of hydrocarbon or excess layer at the heat transfer surface caused a reduction of

10、the surface energy between the solid surface and the liquid. The existence of an excess layer at the liquid solid interface is analogous to the existence of a surfactant induced excess layer at a liquid vapor interface. Conse-quently, the hydrocarbon is not a typical surfactant because it accumulate

11、s at the solid liquid interface rather than the liquid vapor interface. However, the reduction in the liquid solid surface energy results in a similar reduction in bubble departure diameter that occurs with a conventional surfactant. As a consequence of the bubble size reduction, the active site den

12、sity increases. A boiling heat transfer enhancement existed when a favorable balance between an increase in site density and a reduction in bubble size occurred.In another boiling additive study, Kedzierski 2 speculated that fouling caused a more modest improve- ment in the heat fux of R123 with the

13、 addition of iso- pentane and hexane. Overall, the R123/isopentane (99.5/0.5) by mass mixture exhibited a 10% heat enhancement for heat within the range of 10 to 90 kW/m2. Similarly, the R123/hexane (99.5/0.5) mixture showed an overall 4% and a maximum of 13% heat enhancement over that of pure R123.

14、The purpose of the present study is to determine the effect of a boiling additive on the condensation heat transfer performance of R123. A boiling additive is unlikely to be commercially viable if it causes a heat transfer degradation in the condenser that more than of sets the heat transfer enhance

15、ment in the evaporator. Assuming that isopentane is a better additive than hex- ane for the enhancement of R123 boiling on all surfaces, isopentane may potentially produce the greatest net heat transfer improvement between the condenser and the evaporator. Based on that premise, the vapor-space cond

16、ensation heat transfer performance of pure R123 and an R123/isopentane (99.5/0.5) by mass mixture were measured on a vertical, trapezoidal an surface.ApparatusFig. 1 shows a schematic of the apparatus that was used to measure the vapor-space condensation heat transfer data of this study. Specificall

17、y, the apparatus was used to measure the vapor saturation temperature (Tv), the average condensation heat (q00 ), and the wall temperature (Tw) of the test surface at the root of the fin. The three principal components of the apparatus were test chamber, post condenser, and boiler. The internal dime

18、nsions of the test chamber were approximately 254 200 130 mm. The boiler was charged with approximately 10 kg of R123. Hot city water flowed inside the tubes of the boiler to heat the test refrigerant on the shell-side of the boiler. The test section was visi- ble through three, flat quartz windows.

19、 The opposing side of the finned condensing test surface was cooled with high velocity (2.5 m/s) water flow. Varying the temperature of the cooling water varied the heat flux of the test section. The vapor produced by the boiler was condensed by the post condenser and the test section and returned b

20、y gravity to the liquid pool. The post condenser was identical to the shell-and-tube boiler; however, chilled waterflowed inside the tubes while the vapor condensed on the outside of the tubes. The duty of the boiler and the post condenser were significantly large so that a wide variation in the dut

21、y of the test surface would not afect the saturation pressure of the test apparatus. The purger and the desiccant filter removed non-condensible gases and water, respectively, from the test refrigerant after charging and before testing.To reduce the errors associated with the saturation temperature

22、measurement, the saturation temperature of the vapor was measured with two 450 mm long 1.6 mm diameter stainless steel sheathed thermocouples. The small diameter provided for a relatively rapid response time. Approximately 180 mm of each thermo- couple length was exposed to the vapor of the test cha

23、mber. The portion of each thermocouple that was in the test chamber was shielded with a 6 mm diameter stainless steel tube and was in contact with the saturated refrigerant vapor. The tips of the two thermocouples were placed near the lower edge of the test plate and approximately 60 and 95 mm, resp

24、ectively, from the front of it.Test surfaceFig. 2 shows the oxygen-free high-conductivity (OFHC) copper integral-trapezoidal-fin test plate used in this study. The integral-trapezoidal-finsurface in this study was machined directly onto the top of the test plate by electric discharge machining (EDM)

25、. Fig. 3 shows a drawing of the fin cross section. The fin pitch was 1.36 mm. The surface had nominally 746 fins per meter oriented along the long axis of the plate. The ratio of the surface area to the projected area of the surface was 2.87. The ratio of the finarea (Af) to the total area (Ao) was

26、0.74. The fin-tip width and the fin-height were 0.24 and 1.53 mm, respectively.Measurements and uncertaintiesThe standard uncertainty (ui) is the positive square root of the estimated variance u2. The individual standard uncertainties are combined to obtain the expanded uncertainty (U). The expanded

27、 uncertainty is commonly referred to as the law of propagation of uncertainty with a coverage factor. All measurement uncertainties are reported for a 95% confidence interval.The copper-constantan thermocouples and the data acquisition system were calibrated against a glass-rod standard platinum res

28、istance thermometer (SPRT) and a reference voltage to a residual standard deviation of 0.013 K. The NIST thermometry group calibrated the fixed SPRT to two fixed points having expanded uncer- tainties of 0.06 mK and 0.38 mK. A quartz thermo- meter, which was calibrated with a distilled ice bath, agr

29、eed with the SPRT temperature to within approxi- mately 0.003 K. No correlation was found to exist between the measured thermocouple electromotive force (EMF) and a measured 1 mV reference. Consequently, there was no measurable drift in the acquisition voltage measurement over a month period. Before

30、 each test run,the measurements of a thermocouple in the bath were compared with the SPRT. The median absolute diference between the thermocouple and the SPRT was 0.02K over the duration of the entire study. Considering theuctuations in the saturation temperature during the test and the standard unc

31、ertainties in the calibration, the expanded uncertainty of the average saturation tem- perature was no greater than 0.04 K. Consequently, it is believed that the expanded uncertainty of the tempera- ture measurements was less than 0.1 K. The saturation temperature was also obtained from a pressure t

32、rans- ducer measurement with an expanded uncertainty of less than 0.03 kPa. The expanded uncertainty of the saturation temperature from a regression (with a resi- dual standard deviation of 0.6 mK) of equilibrium data 3 for R123 was 0.17 K. The saturation temperature obtained from the thermocouple a

33、nd the pressure mea- surement nearly always agreed within 0.17 K for the pure R123 data.Fig. 2 shows the coordinate system for the 20 wells where individual thermocouples were force-fitted into the side of the test plate. The wells were 16 mm deep to reduce conduction errors. Using a method given by Eckert a