1、4727 Natural Convection Heat Transfer for Fin-Tube Heat Exchangers Y.K. Chuah, Ph.D. Fellow ASHRAE C.T. Chen AB ST RACT In thisstudy naturalconvection heat transferofl5firn-tube heat exchangers was measured. There WereJiveJinpitches and three diferent tube numbers. The results show that lower chille
2、d water temperature will give more heat transfel: but water velocity in tubes is a weak factor of heut transfer It was also found that the single-tube heat exchangers have thehigh- est heat transfer coeflcient, with the three-tube heat exchang- ers the lowest. The heat transfer coeflcient and the un
3、it area heat transfer rate of the heat exchangers can reach 1 O W/m2 “C and 1 O0 W/m2, respectively. A hydraulic diameter forfin-tube heat exchangers was dejned and used as the length scale of natural convection heat transfer The experimental results were analyzed and the correlation Nu = 0.43 was o
4、btained. INTRODUCTION Air moving and distribution often constitute more than 30% of the power consumption for an air-conditioning system. Therefore, air-moving power is an important research area for energy conservation. There have been development projects (Wilkins and Kosonen 1992) on natural conv
5、ection type cool- ers for air-conditioning applications. These coolers are often called cold beams, as they are placed under the ceiling. The natural convection effect of the higher density cooler air is used to cause the air distribution for an air-conditioned room. For better heat transfer, fin-tu
6、be heat exchangers were used in this study as the natural convection coolers. Compact heat exchangers of the fin-tube type were used in this study to maximize the heat exchange area on a per unit volume basis. Heat transfer data for many compact heat exchangers can be found in the open literature, n
7、otably those presented by Kays and London (1984). However, most heat exchangers were studied and tested for forced convection performance. There have been studies of natural convection heat transfer for a single bare tube and tube banks, and the results are widely published (ASHRAE 2001). However, t
8、here are very few studies on natural convection heat transfer for compact heat exchangers. Only some data and information are found in some published literature (Lebrun et al. 2000; Bravo et al. 1995), usually as part of the study in developing natural convection coolers. In this study, aluminum fla
9、t fins were used for the heat exchangers, with five different fin pitches, namely, 2, 3,4, 5, and 6 mm. Different sizes of the heat exchangers were made up with one, two, and three tubes. Smooth wall copper tubes with outside diameter of 9.525 mm and wail thickness of 0.35 mm were used. The finned c
10、opper tube length was 40 mm. The dimensions of the heat exchangers studied are shown in Figure 1. The combination of different numbers of copper tubes and fin pitches makes up 15 different heat exchangers in the study, as shown in Table 1. EXPERIMENTS The performance of these heat exchangers was stu
11、died experimentally. Natural heat transfer of these heat exchangers was measured in an environmental control chamber after rather stable temperature conditions were achieved at about 24C. A schematic diagram ofthe experimental setup is shown in Figure 2. Room air was cooled by passing 10-18C chilled
12、 water through the heat exchanger tubes. The lower tempera- ture was chosen to avoid condensation. Temperature measure- ments were taken for the water inlet and outlet, the exterior surface, and the fin surface of the heat exchangers. Tempera- Y.K. Chuah is a professor in the Department of Air-condi
13、tioning and Refrigeration Engineering, National Taipei University of Technology, Taipei, Taiwan. C.T. Chen graduated with an ME degree from the National Taipei University of Technology. 354 02004 ASHRAE. Table I. Experimental Cases Experimental Cases o1 02 Tube Number Fin Pitch Fin Length Fin Height
14、 Fin Thickness one 6mm 25 mm 22 mm 0.25 mm one 5mm 25 mm 22 mm 0.25 mm 03 04 05 one 4 mm 25 mm 22 mm 0.25 mm 0.12 mm one 3mm 25 mm 22 mm one 2 mm 25 mm 22 mm 0.12 mm 06 07 08 two 6mm 50 nun 22 mm 0.25 mm two 5mm 50 mm 22 mm 0.25 mm tW0 4mm 50 mm 22 mm 0.25 mm I 15 I three I 2mm 1 76mm 1 22mm 1 0.12m
15、m 09 10 11 II two 3mm 50 mm 22 mm 0.12 mm two 2mm 50 mm 22 mm 0.12 mm three 6mm 76 mm 22 mm 0.25 mm (a) Tube and fin dimensions. 12 13 14 fins tube (b) Tube length and fin pitches three 5mm 76 mm 22 mm 0.12 mm three 4 mm 76 mm 22 mm 0.25 mm three 3mm 76 mm 22 mm 0.12 mm Figure I The dimensions of th
16、e fin tube heat exhangers. ASHRAE Transactions: Research 355 ,heat exchanger Figure 2 Schematic diagram of the experimental setup. constant temperoture water both ,I ri h 1,2n temperoture measurement screen 0,075m 0,15m experimental chonber _ Air side Tube side 1. Average temperature above and below
17、 heat exchanger (“C) 2. Air temperature distribution near the heat exchanger (“C) Average temperature of the heat exchanger surface, including fin surface area (“C) ture distribution of the room air in the vicinity of the heat exchanger was also measured. Variations of inlet water temperature and wa
18、ter flow velocity were tested. With differ- ent numbers of copper tubes and fin pitches, there are four parameters for experimental comparison of heat transfer performance. The experimental chamber used had dimensions of 1.2 m x 1.2 m x 2.1 m. The heat exchanger was placed at the top of the chamber,
19、 at a height of approximately 1.9 m. As seen in Figure 2, the inlet of the fin-tube heat exchanger was connected to the supply pipe of a constant temperature water bath. For measurements of the air temperature, thermocouples were fixed on wire screen as shown in Figure 3. This would allow simultaneo
20、us measurements at several points. The symmetry of temperature distribution was utilized so that thermocouples were fixed on one side and where more variation of temperature was expected. Shifting of the wire screen enabled measurements of air temperatures at different heights. Thermocouples were al
21、so fixed at positions to measure the average air temperature at the top and at the bottom of the heat exchanger. T-type thermocouples were used in the experiments and were calibrated in the range of 8 to 30C. The experimental measurements are summarized in Table 2. EXPERIMENTAL ANALYSIS The applicat
22、ion of cold beam natural convection cooling requires little or no condensation on the heat exchanger surface. Water temperature used in this study was high enough that no significant condensation occurred during the experi- ments. Therefore, only sensible heat exchange performance is analyzed here.
23、The heat balance on the water is calculated using Equa- tion l, Qcooi Mwater . Cp, water water (1) where Qcoo, (W) is the heat exchange capacity. The heat transfer of the exchanger is shown in Equation 2, where U is the overall heat transfer coefficient and A is the heat exchanger area. Figure 3 The
24、rmocouples fixed on wire screen. Table 2. Experimental Measurements I Chilled Water Side IInlet and Outlet Water Temperature (OC) Lebrun et al. (2000) defined the log mean temperature difference AtLOG as in Equation 3. in which twsu and t, are the inlet and exit water temperatures, respectively, and
25、 is the average temperature of the fins. In the equation, tmom is taken as the average air temperature at the top of the heat exchanger. The leaving air temperature is taken to be Therefore, the heat transfer can also be calculated using the air heat balance as in Equation 4. Bravo et al. (1995) sug
26、gested thatQCooI and At, have relationship as in Equation 5, where K and n depend on the type of heat exchanger and the mode of heat transfer. 356 ASHRAE Transactions: Research 35, 150 120 e.; 90 e- 4 6 60 30 8 - 0.168 m/s - 4- - 0.336 m/r -C.- 0.603 m/r - 4- - 0.335 mis -.+- 0.503 mls -o- 0.671 mis
27、 - - - - I I l I I 10 12 14 16 18 I I I I I 10 12 14 16 18 W aterTernp. (OC) 12 10 y 8- - “E 2 6- 8 4- 2- Figure 4 Heat transfer per unit length for case 01. - - Giblin (1974) suggested that the heat transfer coefficient of natural convection for a horizontal fin-tube heat exchanger could be express
28、ed as in Equation 6. For pure natural convection, n2 lies between 0.25 and OS. Temoveanu and Wei (1999), however, suggested that for hori- zontal tubes, n2 lies between O. 15 and 0.3. It is noted here that air-side resistance dominates in natural convection flows of this type; therefore, h in Equati
29、on 6 can be taken as approxi- mately the overall heat transfer coefficient and therefore, can be derived experimentally using Equations 1,2, and 3. To obtain the heat transfer correlation for the heat exchangers used, this study uses hydraulic diameter defined as in Equation 7: ACL D, = 4- A (7) in
30、which A, is the minimum flow area, L is the length of the air flow path, and A is the total heat transfer area. Also the average Nusselt number is defined as in Equation 8: - h.0, Nu = - k in which h is the average heat transfer coefficient and k is the conductivity of air. Grashof number is thus de
31、fined as in Equa- tion 9: (9) Rayleigh number is calculated as in Equation 10: Ra = GrxPr (10) Last, the average Nusselt number can be- expressed in terms of Rayleigh number as in Equation 1 1 : 14 - 4- - 0.335 mis -+- 0.603 mis -1- 0.611 mis I I l I I J 10 12 14 16 18 20 Water Temp. (OC) Figure 6 H
32、eat transfer coeficient for case 01. Nu = C(Ra)m In Equation I 1, rn is usually equal to 114 for laminar flow. The constant C was suggested by Ostrach (1 952) as 0.5 16, but Touloukian et al. (1948) took C to be 0.726 for laminar flow and C to be 0.056 for turbulent flow. RESULTS AND DISCUSSION Effe
33、cts of Water Velocity Figures 4,5, and 6 show the experimental results for the single-tube heat exchanger with fin pitch of 6 mm. It seems that water velocity in the tube has little effect on the heat transfer. This is incongruent with the above discussion that air-side resistance dominates. Figure
34、4 shows that heat trans- fer per unit length of the heat exchanger can reach approxi- mately 17 W/m when the inlet water temperature is 10C and the water velocity is 0.671 ds. In Figure 5, when water velocity increases from O. 168 to 0.67 1 m/s, the heat transfer per unit area increases only from ab
35、out 101 W/m2 to about 104 W/m2. Also, this increase in water velocity will only bring the heat transfer coefficient from about 8.7 W/m2 “C to 9.3 W/m2 “C as seen in Figure 6. The experimental results ASHRAE Transactions: Research 357 i qn 8“- 120 - -c- Case01 ioc - -A - Case06 10C -.+- Casei1 ic c m
36、 I - : t I I I II 0.2 0.4 0.6 O .8 WaterVelocity (mk) Figure 7 Heat trunsferper unit area for diflerent number of tubes. verify that water velocity does not affect the heat transfer much for this natural convection heat exchanger. Effects of Inlet Water Temperature Water temperatures used in this st
37、udy were lO“C, 12“C, 14“C, 16“C, and 18C. The experimental results are also shown in Figures 4, 5, and 6. When the water velocity is 0.168 ds, and when the water inlet temperature decreases from 18C to 1 OOC, the heat transfer per unit length increases from about 4.9 W/m to 17 W/m. Also, per unit ar
38、ea heat transfer will increase from about 30 W/m2 to 101 W/m2. It is also seen that the heat transfer coefficient k will increase from 8.77 W/m2 “C to 11.6 W/m2 “C. Similar trends were found for higher water velocity. At water velocity O.671m/s7 heat transfer coefficient h is 8.2 W/m2 “C at 18C and
39、will increase to 11.8 W/m2 “C when the inlet temperature decreases to 10C. This shows that water temperature is an important factor of heat transfer. Effects of Number of Tubes Cases with fin pitch of 6 mm are used in the comparison and the results are shown in Figures 7 and 8 for water temper- atur
40、e of 10C. Figure 7 shows the comparison in terms of per unit transfer area. Single tube (case 01) seems to be superior to others, with three tubes (case 11) the worst. It can also be seen in Figure 8 that case O1 has higher heat transfer coeffi- cient. This shows that, although more tubes and more t
41、ransfer area may bring higher heat transfer rate, the heat transfer in terms of per unit transfer area is better for smaller heat exchangers. This can be due to more free flow path for natural convection flow for a smaller heat exchanger. Effects of Fin Pitch The fin pitch tested was 2 mm, 3 mm, 4 m
42、m, 5 mm, and 6 mm, and the experimental results are shown in Figures 9,l O, and 1 1. It can be seen in Figure 9 that the heat exchangers with l4 lot Ya E 5 6 c 4 2 -c- Case01 10C - 4 - Case06 1OoC -. + -. Case1 1 10C I I I I I L 0.2 0.4 0.6 0.8 Water Velocity (mls) Figure 8 Heut trunsfer coecient fo
43、r diferent number of tubes. 35 I I 25 301 - Case01 iOc - 4 - Case02 IOC Case03 10% - + - Case04 t 0-c -.-O- Case05 10C 5t O0 1 0.2 0.4 O .6 O .8 Water Velocity (mk) Figure 9 Heut transfer per unit length for diferent fin pitches. smaller fin pitches have higher heat exchanger capacities per unit len
44、gth. That seems to be the effects of more heat transfer area per unit length. Fin pitch of 2 mm (case 05) performs better the others (3 mm, 4 mm, 5 mm, and 6 mm) in terms of heat transfer per unit length. However, Figures 10 and 11 show that larger pitch is better in terms of heat transfer per unit
45、area and heat transfer coefficient. This shows that heat transfer does not increase proportionally with the increase in transfer area due to fins. In contrast to forced convection, smaller fin pitch does not cause enhancement in heat transfer coefficient. It shows in Figure 10 that in terms of heat
46、transfer per unit area, fin pitch of 6 mm is about optimal. 358 ASHRAE Transactions: Research - -A - Case02 10C -+-. Case03 IOC I I I I 0.2 0.4 0.6 0.8 Water Velocity (mls) Figure10 Heat transfer per unit area for diferent jn pitches. Last, a correlation ofNusselt number was attempted with the resul
47、ts in Equation 12. The results are also shown in Figure 12. Nu = 0.43(Ra)* (12) The above equation was obtained by assuming laminar flow at the heat exchanger and the exponent of Ra was taken to be 0.25. CONCLUSION This study experimented with natural convection heat transfer of a fin-tube heat exch
48、anger by varying the tube inlet water temperature, water velocity in the tube, number of tubes, and fin pitch on the tube. The following conclusions can be derived: 1. Inlet water temperature is a significant factor of heat trans- fer; lower water temperature results in more heat transfer. Water vel
49、ocity in the tube is a weak factor of heat transfer. It was found that more tubes and smaller fin pitch results in lower heat transfer coefficient, contrary to what is expected from forced convection heat transfer. For the heat exchangers studied, the heat transfer coefficient obtained could reach 10 W/m2 OC, and the per unit area heat transfer rate could reach 100 W/m2. The experimenta1 results were analyzed and the correlation Nu = 0.43(Ra)0.25 was obtained. 2. 3. 4. AKNOWLEDGMENT Support from the National Science Council at Taiwan is ack
