1、4743 Condensation Enhancement of R-22 by Twisted-Tape Inserts Inside a Horizontal Tube Ravi Kumar, PhD Associate Member ASHRAE Sachida Nand La Kailash Nath Agarwal, PhD , PhD ABSTRACT In this paper the salient features of an experimental inves- tigation on heat transfer enhancement by twisted-tape i
2、nserts inside a horizontal R-22 condenser are reported. The test condenser has four test sections connected in series, and each test section has an inside diameter of 12.7 mm and length of 950 mm. Three twisted-tape inserts with twist ratios, y, of 14.95, 9.05, and 5.90 wereput, one by one, in the e
3、ntire length of the test condenser. The refrigerant mass velocity G was variedfrom 21 0 to 372 kg/s-m2 in four steps. Enhancement in the average heat transfer coeflcient h was attained up to 25% for the twisted-tape insert with the twist ratio y of 5.90, as compared to the plain tube jlow. It was al
4、so found that the twisted-tape inserts are more efective at low vapor mass veloc- ities inside the tubes. An empirical correlation equation was also developed to calculate the heat transfer coejcients. This equation predicts the experimental heat transfer coeficient within an error band of f30% for
5、a tube with the twisted-tape inserts. INTRODUCTION Since the Montreal Protocol was signed in 1987 forbid- ding the use of ozone-depleting refrigerants, a spate of research has taken place in subsequent years. As a result of these efforts, several new ecofiiendly substitutes for conven- tional refrig
6、erants have been developed. However, in the developing countries, R-22 will continue to be used in the refrigeration and air-conditioning industry until the year 2030. The principal reason for the continued use of R-22 is its very low ozone-depleting potential (5% that of R-1 1) and, secondly, to da
7、te no reliable drop-in substitute for R-22 has emerged. Therefore, experimental investigations on R-22 are still being conducted (Jung et al. 2003, 2004). Hari (rishna Varma, PhD The research investigations for the condensation inside a plain horizontal tube have been carried out for a number of yea
8、rs and, at present, the investigations are still being carried out for the condensation of environment friendly substitutes for conventional refrigerants (Boissieux et al. 2000; Jung et al. 2003). In the earlier years, some investigators (Kirov 1949; Kreith and Margolis 1959; Bergles 1973) discovere
9、d a few techniques to enhance the convective heat transfer coefficient with the help of insertion devices, namely, wires, tapes, baffles, etc. These devices are helpful in creating turbulence in the fluid flow field and, thus, are also known as “turbulence promoters.” These methods are fully effecti
10、ve in enhancing the heat transfer rate during single-phase flow. Considering the usefulness of turbulence promoters in single-phase flow, their performance has also been tested in two-phase flow systems as well (Kaushik and Azer 1988; Luu and Bergles 1980; Schlager et al. 1990; Behabadi et al. 2000;
11、 Sami and Maltais 2000), such as in refrigerant condensers. It has been found that the turbu- lence promoters are good heat transfer augmentation devices for refrigerant condensation. The published research is also available, reporting successful use of turbulence promoters to enhance the condensing
12、 heat transfer rate (Royal and Bergles 1978; Luu and Bergles 1980; Cavallini et al. 2003). In fact, among all enhancement techniques employed, swirl flow generation by twisted-tape insert in the vapor-carrying tube of a condenser is very promising due to its simplicity, ease of fabrication, low weig
13、ht, and low cost. However, the improve- ment in heat transfer coefficient depends upon the range of experimental parameters. The research in the area of swirl- flow condensing heat transfer has an application in the area of refrigeration, but work in the area of swirl-flow condensation of refrigeran
14、ts is quite meager. Ravi Kumar is an assistant professor and Kailash Nath Agarwal is a professor in the Department of Mechanical hence, the two concen- ASHRAE Transactions: Research 19 Ail dimensions in mm. Not to scale X 1, test-condenser, 2. outer pipe of annulus, 3. thermal insulation, 4. wooden
15、box Figure 2 Details of test section. tric tubes formed a counterflow annulus. Four such test sections were put in series to form the complete test condenser. In the test condenser, pure vapors of refrigerant were carried inside the inner tube; however, the cooling water flowed in the annular space
16、outside the inner tube. The outer wall tempera- ture measurement of the inner tube was done at four axial loca- tions in each test section. The T-type thermocouples were fixed on the top, side, and bottom positions at all four loca- tions. Hence, on each test section tube were 12 thermocouples, with
17、 a total of 48 thermocouples fixed on the whole test condenser. The entire test condenser was completely insulated with glass wool to prevent any heat loss to the surroundings. All four test sections of the test condenser were instru- mented to measure the inlet and outlet temperatures of the coolin
18、g water and also the inlet and outlet temperatures of refrigerant vapor. A pre-condenser was installed ahead of the test condenser. By regulating a predetermined mass flow rate of cooling water in this condenser, a desired vapor quality could be achieved at the test condenser inlet. The data were ac
19、quired for a vapor quality range of approximately 5 to i0 K superheat at the inlet and about 0.10 at the outlet of the test condenser, The swirl in the refrigerant vapor was created with the help of twisted tapes. These tapes were made from 0.5 mm thick stainless steel flats. To accommodate the cont
20、raction in the twisting, the width of the strips was approximately 5% more than the inside diameter of thc test section. The edges of the strips were gently filed to remove any abrasions and make them smooth. Figure 2 shows a tape fitted inside the test section tube. In the present investigation, th
21、ree inserts with the twist ratio y (the ratio of half-pitch of the helix and the inside tube diameter) of 14.95,9.05, and 5.90 have been used. Energy balance calculations for the pre-condenser, test condenser, and the after condenser were carried out. The unac- Section at x-x counted energy loss of
22、the refrigerant was in the range of 4.2% to 8.6% with an average value of 6.8%. The uncertainty in the experimental determination of heat transfer coefficient h has also been determined. The condensing-side heat transfer coefficient h has been calculated according to the procedure described in Agraw
23、al et al. (2004), and the average heat transfer coefficient is the arithmetic average of the heat transfer coefficients at all the locations at a given R-22 mass velocity. The inside and outside diameters of the test section were measured with an accuracy of 0.02 111111. The coolant flow rate was me
24、asured with the highest error of S% of the indicated flow rate. The accuracy in the measurement of tube wall temperature and the cooling water temperature were on the order of 0.1OC and O.O5OC, respectively. The saturation temperature of the condensing vapor was determined with an uncertainty on the
25、 order of 0.2OC. This resulted in an uncertainty in the determination of heat transfer coefficient h of approximately 7% (La1 1993). RESULTS AND DISCUSSION In order to study the relative performance of different twisted-tape inserts, experimental data were acquired for the condensation during plain
26、tube flow and flow with the twisted- tape inserts occupying the entire length of the test condenser. The performance during plain ube flow and for the tube with different twisted-tape inserts is discussed in the following sections. Plain Tube Flow For plain tube flow, the variation of heat transfer
27、coes- cient h with the vapor qualityx is shown in Figure 3. The vapor qualiy is taken as the ordinate and the heat transfer coefficient (h) inside the tube is taken as the abscissa. The vapor quality 20 ASHRAE Transactions: Research 8000 5000 “E 3000 Y 2 =- 2000 f 1000 $ 1500 8 800 s = 500 E 2 300 2
28、00 150 1 O0 .- L u) c - m O , 2500, 0.0 0.1 0.2 0.3 0.4 0.5 0;6 0.7 0.8 0.9 1.0 vapor quality, x Figure 3 Heat transfer coeficient for the flow in a plain tube at diferent refrigerant mass velocities. is unity near the entrance of the test condenser, and at the exit the quality of vapor is close to
29、O. 15. The condensation of vapor takes place while passing through the test condenser. As the vapor moves toward the exit from the test condenser, the vapor quality x is reduced. The same observation is made as well for the heat transfer coeffi- cient h. The heat transfer coefficient h is reduced do
30、wn the stream because the condensate layer in the inner wall of the test condenser offers more resistance to heat transfer. However, at a particular section of test condenser, as the mass velocity G increases, the heat transfer coefficient h is also enhanced. In Figure 3, the average heat transfer c
31、oefficient h for different mass velocities is also shown. The average heat transfer coefficient h has the hi hest value of 21 15 W/m2 K at the mass velocity of 372 kg/s m , and the avera e heat transfer coefficient h has the least value of 1233 W/m K at the mass velocity of 2 10 kg/s m2. Hence, the
32、average heat transfer coef- ficient h is reduced by 58% for a reduction of 56% in mass velocity G. This relation is almost linear; however, from the mass velocity of 237 kgs m2 to 210 kgs m2, the reduction of average heat transfer coefficient h is not significant. The present experimental results ar
33、e also compared with the correlations developed by other investigators. The present experimental data lie within a range of +40% to -30% with the predictions from the Akers correlation (Akers and Rosson 1960). The correlation of Tandon et al. (1985) predicts the present experimental data in an error
34、 band of +35% to -40%. In two-phase flow studies involving the condensation of vapor, deviations on this order are not unusual (McAdams 1949). Dobson and Chato (1998) developed the following correlation for the condensation of R- 12 and R- 1 34a: F 5 The Dobson and Chato (1998) correlation has also
35、been reported to overpredict for high mass velocities. However, the above correlation predicts the data of the present investigation for the condensation of R-22 in a range of +20% when the constant 2.22 is modified as 1 .O. In Figure 3 a comparison is also made between the experimental data of Jung
36、 et al. (2003) for a test section of 1000 mm length and 9.52 mm diameter and the present investigation. As the tube dimensions are different for both investigations, the condensing Nusselt number (Nu) at different mass velocities of R-22 was calculated. The Nusselt number of the present investigatio
37、n is approximately 18% less at a lower mass velocity of R-22, and at high mass velocity the data of the present investigation are nearly 14% less than those of Jung et al. (2003). Flow with Twisted-Tape Inserts The variation of heat transfer coefficient h with different twist ratios y at a given vap
38、or mass velocity G is shown in Figures 4 to 8. These figures are drawn taking vapor quality x as the abscissa and heat transfer coefficient h as the ordinate with twist ratio y as a parameter. Each of the above five figures is for a particular vapor mass velocity G. A general observa- tion of these
39、graphs suggests a clear scattering in test data, which is a function of the test conditions, test facility, and robustness and accuracy of measurement. The repeatability of 20% of the test runs was checked and it was found to be within In Figure 4 the graph is drawn for a constant vapor mass velocit
40、y G of 2 1 O kgs m2. At this mass velocity, the twisted- tape insets with twist ratio y of 5.90 and 9.05 yielded the aver- age heat transfer coefficient h close to 1527 W/m2 K, which is an improvement of nearly 24% in comparison to that for a plain tube flow. However, the twisted tape with twist rat
41、io y of 14.95 has improved the condensing-side heat transfer coeffi- cient h by 17%. But for the vapor mass velocity of 237 kgs m2 (Figure 5), the same tape with a twist ratio of 14.95 has improved the average condensing-side heat transfer coeffi- cient h by only 14%, whereas the tape with the twist
42、 ratio y of 5.90 has improved the average condensing-side heat transfer coefficient h by 25% and has remained the best performing twisted-tape insert. Figure 6 is drawn taking vapor quality x as the abscissa and the heat transfer coefficient h as the ordinate for a constant vapor mass velocity of 28
43、2 kg/s m2. The best performing twisted tape at this mass velocity is that with the twist ratio y of 5.90. This twisted tape has enhanced the average heat trans- fer coefficient h by nearly 1 1 %. The twisted tapes with twist ratio of 14.95 and 9.05 have enhanced the average heat transfer Coefficient
44、 by 4.5% and 5.5%, respectively. No firm conclu- sion can be drawn from these results as the uncertainty in the present investigation of the heat transfer coefficient is nearly 7%. 4%. ASHRAE Transactions: Research 21 s 2000 # 1000 s +- 1500 c a .- 8 800 $ 600 400 300 200 ln t c J= 150 - W 14.95 9.0
45、5 5.90 twist ratio, y - O C 0 o: mo 0 +o= O. .O o m. a. O O O O Figure 4 Comparison of heat transfer coeficient for twisted-tape inserts, G = 21 O kg/s.m2. 8000 6000 4000 3000 ,E 5 2000 $ 1500 ! 800 8 600 5 400 2 300 200 150 100 .- 1000 C m m o y =m(plain) y = 14.95 G=282 kgls-m . n 0.0 0.1 0.2 0.3
46、0.4 0.5 0.6 0.7 0.8 0.9 1.0 1.1 vapor quality, x Figure 6 Comparison of heat transfer coeficient for twisted-tape inserts, G = 282 kg/s.m2. 8000 6000 4000 NE 3000 Y 9 =- 2000 $ 1500 1000 8 800 600 400 .- c m 2 300 200 I 50 I O0 y = 5.90 G=237 kgla-m2 Y “E 2000 f 1500 i C Q al S r c E C m S -10 200 2
47、20 240 260 280 300 320 340 360 380 400 refrigerant velocity, G, kg/s-m2 Figure 9 Variation of enhancement ratio with refrigerant mass velociy for different twisted-tape inserts. In Figure 9, a graph has been plotted taking the refrigerant mass velocity as abscissa and enhancement in heat transfer co
48、efficient with respect to a plain flow, on a percentage basis, as the ordinate. The vapor quality varies from 1 .O to O. 1. The twist ratio y of the inserted tape is taken as the parameter. The twisted-tape insert with twist ratio of 5.90 has been found to be the best performing insert and it has en
49、hanced the average condensing-side heat transfer coefficient up to 25%, in comparison to that for a plain flow, for all the mass velocity of R-22. It is also discovered that the effect of all twisted-tape inserts is more pronounced for low mass velocity than high mass velocity. However, since for higher mass velocities, on the order of 372 kgls-m2, the enhancement in heat transfer coefficient again rises, but only with a single observation for any twisted-tape insert, a conclusion cannot be drawn. The twisted tape with twist ratio of 14.95 has performed poorly, and its performanc
