1、OR-05-1-2 Air-Side Thermal-Hydraulic Performance of Louvered-Fin, Flat-Tube Heat Exchangers with Sequential Frost-Growth Cycles Yanping Xia Student Member ASHRAE Predrag S. Hrnjak, DSc Fellow ASHRAE Anthony M. Jacobi, PhD Member ASHRAE ABSTRACT The thermal-hydraulic performance offolded-jn, flat-tub
2、e heat exchangers is experimentally studied for conditions of an initial frost growth on the air-side surface and for subsequent “refrosting after a defrost period. Typically, the performance under refrosting conditions becomes periodic and repeatable after the third orfourth refrosting cycle. This
3、behavior is caused by the need to reach aperiodic distribution ofwater liquidat the end of a defrost, and the roles offrostgrowth and water retention on the surface are explored in this context. The allowablefiost growth period (before a defi.ost is required), the defrost require- ment, and the ther
4、mal-hydraulic performance are seen to depend on heat exchanger geometry for the two specimens used in this study. The performances for two heat exchangers are compared in the paper to explore geometry effects. INTRODUCTION Louvered, folded-fin, flat-tube heat exchangers are finding broader applicati
5、on in thermal management systems because they offer thermal-hydraulic performance and compactness higher than provided by expanded, round-tube heat exchangers. Often flat-tube heat exchangers are constructed by brazing folded aluminum fins to extruded aluminum microchannels. In air-conditioning and
6、refrigeration systems, these heat exchang- ers have found increasing application as condensers, where they operate under dry-surface conditions. Recently, however, there has been interest in using microchannel heat exchangers as evaporators where they will be subject to wet and frosted surface opera
7、ting conditions. Under frosting conditions, the normal operation of the system will cause the heat exchanger to frost, defrost, and refrost in a cyclic manner. The thermal-hydraulic performance under such conditions and in particular the toler- ance of the design to defrosting and refrosting must be
8、 consid- ered if flat-tube designs are to serve as evaporators. Defrost methods for vapor-compression systems include hot-gas or reverse-cycle defrost (Cole 1989; Al-Mutawa and Sherif 1998), electric defrost (Sherif and Hertz 1998), and others, such as the sublimation approach of haba and Imai (1996
9、). In most systems, the defrost method involves melting the frost while it is on the heat exchanger surface, and that will be the focus of this study in which we will use a simulated hot- gas or reversed-cycle defrost. Thus, the frost will be melted to liquid water, some of which will drain from the
10、 air-side surface during the defrost cycle, and then a new frosting cycle will begin. This approach is intended to model the prevalent condi- tions of heat-pump or refrigeration applications. Over the past several decades, several studies of frost growth and defrosting on finned tube heat exchangers
11、 have been reported. Kondepudi and ONeal(l987) provided a review of frost growth on extended-surface heat exchanger performance, and Kondepudi and ONeal(l989) also conducted research on frost growth on louvered finned heat exchangers. Verma et al. (2002) developed an experimental validated quasi-ste
12、ady finite- volume model for frosting of plain-fin-round-tube heat exchangers. Tassou and Marquand (1987) studied the effects of frosting and defrosting on the performance of air-to-water heat pumps, and Machielsen and Kerschbaumer (1989) did similar research on air coolers. Kim and Groll (2002) stu
13、died the perfor- mance of microchannel heat exchangers as an outdoor coil during frosting and defrosting conditions, and they discussed the effects of using different orientation and fin density of the microchannel heat exchanger on heating capacities, power consumption, and system efficiencies. How
14、ever, the air-side thermal-hydraulic data and quantities of frost and water reten- - Yanping Xia is a graduate student, Pega S. Hrnjak is a research professor, and Anthony M. Jacobi is a professor in the Department of Mechanical and Industrial Engineering, University of Illinois at Urbana-Champaign.
15、 02005 ASHRAE. 487 tion are not available in their paper. Since very little has been published to address frost formation and defrosting on louvered, folded-fin, flat-tube heat exchangers, there is currently little basis available in the open literature for the design of micro- channel heat exchange
16、rs under frosting conditions. The focus of this paper is on the operation of louvered, folded-fin, flat-tube heat exchangers under frosted-surface conditions, with attention to defrost and refrost effects. The performance of two heat exchangers under frosting and defrost- ing conditions will be comp
17、ared in order to analyze the results with attention to the effects of retained liquid water. It is hoped that this experimental study of flat-tube heat exchanger perfor- mance under frosted-surface conditions will provide initial insights into the performance of microchannel heat exchangers under cy
18、clical frosting conditions and that it will stimulate further research on this topic. 2 Heat EXPERIMENTAL METHOD Chamber Experimental Apparatus and Test Procedure The experimental facility used to obtain performance data for microchannel heat exchangers under frosting conditions is shown in Figure 1
19、 and was described in detail in an earlier report (Carlson et al. 200 I). An open wind tunnel was placed inside an environmental chamber. The heat exchanger was positioned at the wind tunnel inlet, suspended on an electronic balance (k3 g) to measure the accumulated frost mass or remaining defrost w
20、ater. Flexible plastic film was used to connect the heat exchanger to the tunnel, allowing the heat exchanger to move freely in the vertical direction for proper weight measurement, while providing a seal that eliminates mass leakage from the Wind dew. o Tunnel Ta o tunnel. A precooler was included
21、in the chamber to set the chamber to the desired temperature prior to the initiation of an experiment. A variable-speed blower was used to provide the air flow, and the flow rate was measured using the pressure drop across standard nozzles. Pressure taps upstream and downstream of the nozzles were c
22、onnected to a Setra model 239 pressure trans- ducer (50.63 Pa), which allowed the air mass flow rate to be determined (kl%) using the methods ofASHRAE Standard 33 (ASHRAE 2000). The air temperature was controlled by regu- lating the power supplied to the heater located in the chamber, using a PID co
23、ntroller and type-T thermocouple (iO.2“C) placed at the inlet of the heat exchanger. Humidification was provided by a steam line and was maintained using a PID controller and a General Eastern model D-2-SR chilled-mirror dew-point sensor (kO.2“C). Two chilled-mirror sensors of the same model were us
24、ed to obtain humidity data upstream and downstream of the specimen heat exchanger. Air temperatures were measured using thermocouple grids at the inlet and exit of the heat exchanger. The upstream thermocouple grid consisted of six ype-T thermocouples in an evenly spaced array, and it provided upstr
25、eam temperature data (*0.2“C). The downstream array consisted of nine type-T thermocouples (kO.2“C). The air- side pressure drop across the heat exchanger was measured with another Setra model 239 pressure transducer connected to static pressure taps as shown in Figure 1 (k0.25 Pa). Ethyl alcohol su
26、pplied by a gear pump was used as the coolant in the experimental loop. The alcohol flow was cooled by a chiller system (not shown in the schematic diagram of Figure I), and coolant temperature was regulated using an elec- trical heater and a PID controller. The coolant temperatures Blower Tc.1 1 1
27、Chilled Mirrors Heater Tc,o pf Pre-cooler Da T 4 4 h4 Solenoid To chiller system Figure 1 Facility schematic. 488 ASHRAE Transactions: Symposia Table 1. Heat Exchanger Geometries / Specimens Total air-side surface, Ao (m2) Heat exchanger width, W(mm) Heat exchanger height, H (mm) Heat exchanger dept
28、h, L,.(mm) Fin height. H,mm) Fin pitch, Pf(mm) #1 #2 3.0 4.0 279 381 171 381 70.0 27.9 1.95 2.12 9.9 8.3 were measured using immersion thermocouple probes at the inlet and exit of the heat exchanger (h0.2“C). The coolant was a single-phase flow of ethyl alcohol, and for the two heat exchangers studi
29、ed, the average temperature differences between the inlet and exit were 1S“C and 3,5“C, respectively. Visually, the frost formation on the front face of the first heat exchanger was evenly distributed, but for the second heat exchanger, more frost accumulated near the inlet header than near the exit
30、 header. A Coriolis-effect mass flow meter was used to measure the coolant flow rate (+O. 1 %). A Campbell Scientific system, composed of a data logger (CR23X) and a multiplexer (AM4 16), was used in the experi- ments. The instruments were sampled at 1 O-second intervals throughout the experiments,
31、and six measurements were aver- aged to provide the results in one-minute intervals for the entire duration of the experiment. The results were written into a text file for subsequent analysis. Still and video images of frost growth were obtained using a CCD camera. At the initiation of an experimen
32、t, coolant was sent to the precooler only until the chamber was cooled to the desired temperature. After achieving the desired chamber temperature, the coolant was diverted to the specimen heat exchanger, and performance data were collected. The frosting experiments were conducted at constant air-in
33、let temperature, refrigerant-inlet temperature, air-inlet humidity, refngerant mass flow rate, and blower frequency. The test conditions were: air inlet temperature of -1C to 2C; refngerant inlet temperature of-1 1C; air rela- tive humidity of 70% to 80%; and initial air face velocity of 1 .O ms. It
34、 should be noted that during the course of an experiment the face velocity decreased, owing to the increase in air-side pressure drop associated with frost deposition. I 1 I 1 k-v*. Figure 2 Structure of the heat exchanger. Heat Exchanger Geometry Two different heat exchangers were used in this stud
35、y; both were circuited in a cross-flow configuration. The airflow was horizontal and the tubes were vertical. The first sample only had a manifold at the top of the heat exchanger, whereas the second specimen had manifolds at the top and bottom of the heat exchanger. The air-side geometry for both h
36、eat exchangers is noted in Table 1, with the structure of the heat exchanger as illustrated in Figure 2. It should be noted that specimen #1 had a smaller face area and larger coil length than did specimen #2. Data Reduction Method Basic data analysis includes determining the capacity of the heat ex
37、changer, which is calculated from both the coolant side and the air side, using and with Every refrosting cycle was arbitrarily chosen to be one hour in duration. After each frosting cycle, coolant was diverted to a and bypass loop, where the coolant temperature was raised to 10C qa,i = jsg for defr
38、osting. After a front view of the heat-exchanger fins captured by the CCD camera showed that the accumulated frost was completely melted, the defrosting cycle was terminated. The heat balance, which is defined below, was always within 10%: (3) For the two heat exchangers for which data will be repor
39、ted, all the defrosting cycles lasted four minutes for the first heat (qc - gal (4, + qa)/2 HB = exchanger and five minutes for the second heat exchanger. The blower was turned off during defrosting. The overall heat transfer coefficient was obtained from ASHRAE Transactions: Symposia 489 (4) where
40、ATl, is the log-mean temperature difference computed under the assumption of counterflow conditions, and F is the correction factor to ATl,. RESULTS Amount of Frost Accumulation and Water Retention The weight of the heat exchanger was measured using an electronic balance. Figure 3 shows the total fr
41、ost mass on the heat exchanger at the end of each frosting cycle (tif, i = 1-5) and the liquid water retained at the end of each defrosting cycle (ti,d, i = 1-5). The data are presented in terms of mass per unit of heat transfer area in order to facilitate more direct comparisons between the specime
42、ns. Figure 3 shows that the amount ofwater retained at the end of a defrost reached an asymptotic value after the fourth cycle, and the water retention per unit area of speci- men #1 was about 25% higher than that of #2 after the second cycle. Figure 3 also shows that the frost accumulation per unit
43、 area reached an asymptote after about the fourth cycle. The accumulated frost per unit area of specimen #1 was higher than that of #2 after the third cycle. The accumulated frost is composed of the newly deposited frost and the frozen water remaining on the heat exchanger from the previous defrosti
44、ng cycle. By comparing the values at tifs and ti,is, it is apparent that the differences in the total accumulated frost mass were mainly due to differences in water retention. Images of Heat Exchanger during Defrost and Refrost Images of the heat exchanger face were captured using a CCD camera. Imag
45、es of the heat exchangers between two successive defrosting cycles are shown in Figure 4, with three complete fins and the tube on one side of the fins appearing in the field of view. Figure 4a shows the melting frost, and bright spots in the image are reflections from water droplets. The corner for
46、med by two adjacent fins hindered the downflow of the droplets, as can be seen from Figure 4b, where a droplet was held in a corner at the end of the defrosting cycle. As the next frosting cycle ensued, the droplets froze directly into ice, and their volume expanded (see Figure 4c). Later, frost gre
47、w on that frozen ice. At the end ofthe frosting cycle, the accumulated frost almost totally blocked the airflow passage (as can be seen in Figure 4d). Figure 4(e) shows the melting of the ice (the substance that is jellylike in appearance), which was much slower than the melting of the frost. At the
48、 end (Figure 49, no droplets were held in the fin corners in the field of view, but it is possible that some droplets were held at other places out of view. In addition to providing a site for future frost growth, drop- lets retained during the defrost cycle increase the pressure drop across the hea
49、t exchanger. Figure 5 shows images selected from each frosting cycle, corresponding to the same pressure drop across the heat exchanger. At the same pressure drop, the frost 250 200 150 100 50 O ._ ._. . . .- . ,* i 0 heat exchanger #1, frost accumulation heat exchanger #1, water retention heat exchanger #2. water retention . 0 heat exchanger #2, frost accumulation . . . . . . ? 1 i . i , I Figure3 Added weight per unit area on the heat exchangers. of the first frosting cycle (starting from a dry surface) appears thicker than in subsequent frost cycles. After the t
