1、2008 ASHRAE 379ABSTRACT This study presents the airside performance of the fin-and-tube heat exchangers having plain fin geometry with a largerdiameter tube (Dc= 16.88 mm). The effect of tube row and finpitch on the airside are examined. It is found that the effect offin pitch on the Colburn j facto
2、r is negligible at N = 2. However,a notable drop of heat transfer performance is seen when thenumber of tube row is increased to eight.The effect of tube rowon the friction factor is almost negligible. However, unlikethose of friction factor, the Colburn j factor reveals a differentcharacteristic su
3、bject to the influence of the number of tuberow. If the fin pitch is above a certain threshold value (Fp= 34mm), the influence of tube row on Colburn j-factor is verysmall. However, if the fin pitch is below this threshold value,the Colburn j-factor exhibits a considerable drop with the riseof the n
4、umber of tube row. This phenomenon is especiallypronounced at low Reynolds number region. It is likely that thisphenomenon is related to the complex flow field interactionsbetween the tube row and fin surfaces. The Wang et al.s corre-lation (2000) gives fairly good predictive ability against thepres
5、ent friction factor but under-predicts the j-factor slightly.Based on the present database, a correlation is developed thatcan describe the j-factor with a 3.99% mean deviation and thef-factor with a 2.83% deviation. INTRODUCTIONPlate fin-and-tube heat exchangers having plain finpattern are the most
6、 commonly employed heat exchangers inHVAC Wang,et al. 1996). However, later investigations had showed adetectable effect of fin spacing on the heat transfer perfor-mance for N = 1 and N = 2 (e.g. Seshimo and Fujii 1991 andSensible Airside Performance ofFin-and-Tube Heat ExchangersData with Larger Di
7、ameter TubeY.C. Liu R. Hu, PhD B.C. Yang, PhDFellow ASHRAE Member ASHRAEI.Y. Chen, PhD C.C. Wang, PhDFellow ASHRAEY.C. Liu is a PhD student and I.Y. Chen is a professor at the Graduate School of Engineering Science and Technology at National YunlinUniversity of Science & Technology, Yunlin, Taiwan.
8、R. Hu and B.C. Yang are senior researchers and C.C. Wang is a senior lead researcherat Energy & Environment Laboratories, Industrial Technology Research Institute, Hsinchu, Taiwan.NY-08-0432008, American Society of Heating, Refrigerating and Air-Conditioning Engineers, Inc. (www.ashrae.org). Publish
9、ed in ASHRAE Transactions, Volume 114, Part 1. For personal use only. Additional reproduction, distribution, or transmission in either print or digital form is not permitted without ASHRAEs prior written permission.380 ASHRAE TransactionsWang and Chi 2000a). Therefore, updated correlation wasdevelop
10、ed by Wang et al. (2000) for taking into account theeffect of fin spacing more accurately. The correlation is basedon test results of consistent reduction process from seven datasources. A total of 74 samples were used to develop the corre-lation. The empirical correlation is given as follows:For N
11、= 1,(1)where(2)(3)For N 2,(4)where(5)(6)(7)(8)(9)The friction factor is given as:(10)where(11)(12)(13)The heat transfer correlation gives a mean deviation of7.53% whereas their friction correlation shows an 8.31%mean deviation. Equations (1), (4) and (10) are the most accu-rate plain fin correlation
12、 in the open literature. However, oneshould notice that the foregoing studies did not contain thedatabase of larger diameter tube. In typical applications likefan-coil or ventilator, use of larger diameter like 15.88 mm isalso very common. Unfortunately, there is limitedperformance data of the fin-a
13、nd-tube heat exchanger havinglarger diameter tube in the open literature. Hence, the objec-tive of the present study is to provide relevant performancedata and to examine the applicability of the existingcorrelation to the database.EXPERIMENTAL APPARATUS ANDDATA REDUCTIONThe sample coils are all pla
14、in fin configuration. Theirdetailed geometric parameters are tabulated in Table 1. Rele-vant definitions of geometric parameters are schematicallydepicted in Figure 2. Detailed construction of the circuitryarrangement is identical to those by Seshimo and Fujii (1991)and Wang and Chi (2000). The pres
15、ent test was conducted inan open wind tunnel as shown in Figure 1. The ambient airflow was forced across the test section by means of a 5.6 kWcentrifugal fan with an inverter. To avoid and minimize theeffect of flow maldistribution in the experiments, an airj 0.108ReDc0.29PtPl-P1FpDc-1.084FpDh-0.786
16、FpPt-P2=P1 1.9 0.23 eReDc()log=P2 0.236 0.126 eReDc()log+=j 0.086ReDcP3NP4FpDc-P5FpDh-P6FpPt-0.93=P3 0.3610.042NeReDc()log-0.158 eNFpDc-0.41log+=P4 1.2240.076PlDh-1.42eReDc()log-=P5 0.0830.058NeReDc()log-+=P6 5.7351.21 eReDcN-log+=Dh4ACLA0-=f 0.0267ReDcF1PtPl-F2FpDc-F3=F1 0.7640.739PtPl- 0.177FPDc-0
17、.00758N-+=F2 15.68964.021eReDc()log-+=F3 1.69615.695eReDc()log-=Table 1. Detailed Geometric Parameters of the Text SamplesNo. Fp(mm) f(mm) Dc(mm) Pt(mm) Pl(mm) N, Row1 2.12 0.12 16.68 38.1 33 82 2.54 0.12 16.68 38.1 33 83 3.17 0.12 16.68 38.1 33 84 2.06 0.12 16.68 38.1 33 45 2.54 0.12 16.68 38.1 33
18、46 3.13 0.12 16.68 38.1 33 47 2.12 0.12 16.68 38.1 33 28 2.54 0.12 16.68 38.1 33 29 3.17 0.12 16.68 38.1 33 2ASHRAE Transactions 381straightener-equalizer and a mixer were provided. The inletand the exit temperatures across the sample coil weremeasured by two T-type thermocouple meshes. The inletmea
19、suring mesh consists of twelve thermocouples while theoutlet mesh contains 36 thermocouples. The sensor locationsinside the rectangular duct were established followingASHRAE (1993) recommendation. These data signals wereindividually recorded and then averaged. During the isother-mal test, the varian
20、ce of these thermocouples was within0.2C. In addition, all the thermocouples were pre-calibratedby a quartz thermometer with 0.01C precision.The pressure drop of the test coil was detected by a preci-sion differential pressure transducer, reading to 0.1 Pa. The airflow measuring station was a multip
21、le nozzle code tester basedon the ASHRAE 41.2 standard (1987). The working mediumin the tube side was hot water. The inlet water temperature wascontrolled by a thermostat reservoir having an adjustablecapacity up to 60 kW. Both the inlet and outlet temperatureswere measured by two pre-calibrated RTD
22、s (Resistancetemperature device, Pt-100W). Their accuracy was within0.05C. The water volumetric flow rate is detected by amagnetic flow meter with 0.002 l/s resolution.All the data signals are collected and converted by a dataacquisition system (a hybrid recorder). The data acquisitionsystem then tr
23、ansmitted the converted signals through GPIBinterface to the host computer for further operation. During theexperiments, the water inlet temperature was held constant at65.00.2C, and the tube side Reynolds number was approx-imately 38,000. Frontal velocities of inlet air ranged from 1 to4 m/s. The e
24、nergy balance between air side and tube side waswithin 2%. The water side resistance (evaluated as 1/hiAi) wasless than 10% of the overall resistance in all cases. The test fin-and-tube heat exchangers are tension wrapped having an “L”type fin collar. Thermal contact conductance provided by themanuf
25、acturers ranged from 11000 to 16000 Wm-2K. To obtain the heat transfer and pressure loss characteris-tics of the test coil from the experimental data, the e-NTUFigure 1 Schematic of the test facilities.Figure 2 Schematic of geometric parameters.382 ASHRAE Transactionsmethod is applied to determine t
26、he UA product in the analysis.Detailed reduction can be seen from previous work (e.g. Wangand Chi, 2000). The airside heat transfer characteristics arepresented in terms of the Colburn j factor:(14)Where Vfr is the face velocity denoting the average veloc-ity in front of the test coils and Vmax = Vf
27、r/s, representing themaximum velocity within the heat exchangers. The term, s, isthe ratio of the minimum flow area to frontal area. All the fluidproperties are evaluated at the average values of the inlet andoutlet temperatures under the steady state condition. The fric-tion factors are calculated
28、from the pressure drop equationproposed by Kays and London (1984). The relation for thedimensionless friction factor, f, in terms of pressure drop isshown below:(15)Where Ao and Ac stand for the total surface area and theflow cross-sectional area, respectively. Uncertainties in thereported experimen
29、tal values of the Colburn j factor and fric-tion factor f were estimated by the method suggested byMoffat (1988). The uncertainties ranged from 3.2% to 15.9%for the j factor, and 3.3% to 22.3% for f. The highest uncer-tainties were associated with lowest Reynolds number.Test results are in terms of
30、Colburn j factor and fanningfriction factor. Complete listing of the data is shown in Table2. The effect of fin pitch on the airside performance for differ-ent number of tube row is seen in Figure 3. The correspondingnumber of tube row is 2, 4, and 8, respectively. In the testrange, the effect of fi
31、n pitch on the Colburn j factor is negli-gible at N = 2. However, a notable drop of heat transfer perfor-mance is seen when the number of tube row is increased toeight. The performance drop is especially pronounced at thelow Reynolds number region. In fact, a 30% drop of heat trans-fer performance i
32、s seen for Fp = 2.12 mm relative to Fp = 3.17mm at N = 8 and ReDc= 2000. The relative independence ofheat transfer performance for N = 2 may be associated withdevelopment of boundary layer along the fin surface. Aspointed out by Kim and Kim (2005) who examined the influ-ence of wider pitch on the fi
33、n-and-tube heat exchangers havingplain fin geometry. Their data also suggests a tiny increase ofj-factor with a rise of fin pitch. However, they also mentionedthat the dependence of j-factor with fin pitch is reduced whenboundary layer interaction could not occur. The results are inline with the pre
34、sent test results. For N = 2, the interactions ofboundary layer is relatively small, giving rise to a negligibleeffect of fin pitch on the j-factor. By contrast, although theeffect of fin pitch on the friction factor is comparatively smallat N = 2, one can see that its influence becomes even smaller
35、as the row number is further increased to 8. The results areagain attributable to the interaction of boundary layer of finsurfaces. For N = 2, the interaction of boundary layer for alarger fin pitch is small, indicating a comparative influence ofentrance. As a result, the friction factor is slightly
36、 larger for awider fin pitch. With the rise of tube row, most fin surface liesin the region of fully developed, thereby the influence of finpitch on the friction factor is reduced.The results can be further made clear in Figure 4 wherethe effect of tube row on the airside performance can be furtherc
37、larified. As seen in the figure, the effect of tube row on thefriction factor is rather small. For the smallest fin pitch beingtested, Fp 2.1 mm, the entrance effect is the smallest amongthe tested samples. Hence the number of tube row did not castany detectable effect on the friction factor. With t
38、he rise of finpitch, one can see that there is a very tiny entrance effect,although the effect is considerably small, one can still see thatthe friction factor for N = 2 is marginally higher than N = 4 andN = 8. Unlike those of friction factor, the Colburn j factorreveals a different characteristic
39、subject to the influence of thenumber of tube row. As seen in Figure 4(a), a marked drop ofColburn j factor with the rise of tube row is seen, especiallywhen the Reynolds number is smaller than 5000. Actually,compared to N = 2 at ReDc= 2000, the Colburn j factor isreduced by half. Nevertheless, the
40、reduction in heat transferperformance is reduced when the fin pitch is increased. Asshown in Fig. 4(c) where Fp 3.15mm, the effect of thenumber of tube row on the heat transfer performance issignificantly reduced. Even for a Reynolds number of 2000,the drop of heat transfer performance is only rough
41、ly 20%.The results implied that there is a countermeasure of theinfluence of fin pitch subject to change of the number of tuberow. In fact, depending on the diversity of the tested fin pitch,different researches reached different conclusions about theeffect of tube row on the heat transfer performan
42、ce. Forexample, Rich (1975) concluded that there is a significantinfluence of tube row on the j-factor, yet the influence isparticularly evident at the low Reynolds number region(similar to Figure 4(a). By contrast, Jang et al. (1996) andKim and Kim (2005) reported a negligible effect of tube rowon
43、j-factor (analogous to Figure 4(c). By examining the testedrange of fin pitches for the foregoing studies (Fp= 1.75 mm forRich, Fp 7.5 mm for Jang et al. and Kim and Kim), it is nosurprise why the previous researches had reached differentconclusions. For the unexpected drop of heat transferperforman
44、ce associated with the number of tube row, Rich(1975) suspected that it is related to the complex interactionsof flow field within the heat exchangers. It is not totallyunderstood about the actual flow pattern being prevailedwithin the heat exchanger when the number of tube row isincreased. However,
45、 the present authors suspect thisphenomenon may be attributed to the interactions of finspacing and tube row which leads to instability of flow (Xi andTorikoshi, 1996). As pointed out by them, the unsteadinesscaused by the tube row is strongly related to the fin spacing. Inanother numerical work car
46、ried out by Torikoshi et al. (1994),who conducted a 3-D numerical investigation of a plain fin-jhoVmaxCpa-Pr23=fAcA0-m1-2P1Gc2-12+()12- 1=ASHRAE Transactions 383and-tube heat exchanger, had clearly showed that the vortexforms behind the tube can be suppressed and the entire flowregion can be kept st
47、eady and laminar when the fin pitch issmall enough (e.g. Fp/Dc= 0.17). Further increase of Fp/Dcto0.3 would result in a noticeable increase of cross-stream widthof vortex region behind the tube. As a consequence, the mixingof flow field becomes more intense when the fin pitch isincreased above certa
48、in threshold value which could give riseto a rather uniform heat transfer characteristic, hence showinga relative independence of heat transfer with the rise of tuberow. In the meantime, when the fin pitch is reduced, the flowfield behind tube row is stabilized and laminar which causesa change of flow field and an inevitable drop of heat transferperformance when the tube row is increased further. Based onthe arguments and the test results of Rich (1975), Kim andKim (2005) and this study, the t
