1、OR-05-1 -1 Design of Air-Cooled R=41 OA Microchannel Condensers Vishwanath Subramaniam ABSTRACT Air-cooled condensers are routinely designed for a variety of applications, including residential air-condition- ing systems. Recent attempts at improving the performance of these heat exchangers have inc
2、luded the consideration of microchannel tube, multilouver Jin heat exchangers instead of the more conventional round tube-plate jn designs. Such microchannel tube, multilouver Jin heat exchangers have a large number of geometric parameters that affect the performance of the condenser: The present wo
3、rk provides a systematic procedure to optimize the performance of such air-cooled microchannel condensers with refrigerant blend R-41 OA for minimum material and energy consumption. Flow regime-based models from the literature and the authors prior work are adapted to model condensation of refrigera
4、nt blend R-41 OA in the microchannel tubes under consideration. A variety of refrigerant-side pass arrange- ments are also considered to help tailor the pass-wise performance to maximize heat transfer within the allow- able pressure drop constraints. Similarly, correlations for airflow over multilou
5、verjns are used to predict the air-side behavior: Features of the air-side geometry such as fin pitch, height, and louver details are varied to obtain the optimal conjiguration. Through the analysis of several cases, it is shown that material savings can result through judicious choice of surface ar
6、ea and refrigerant-side flow area allocations. The results from this study can be used to guide the design of air-cooled condensers for a wide range of design conditions. INTRODUCTION Microchannel-tube, multilouver-fin heat exchangers are fast replacing conventional round-tube, plate-fin condensers,
7、 Srinivas Garimella, PhD Member ASHRAE particularly in automotive air-conditioning applications. The larger air-side heat transfer coefficients due to the louvers and the larger surface area per unit volume of these heat exchangers are believed to lead to more compact geometries. The basic geometry
8、of a microchannel condenser is shown in Figure 1. Garimella and Coleman (1 998) studied the design and optimization of air-cooled, round-tube condensers for ammonia-water absorption heat pumps. They studied the performance of these systems with flat, wavy, louvered, and annular fins. The goal of the
9、ir optimization was to obtain the design heat duty of 18 kW with the minimum heat exchanger mass possible. It was found that wavy fins resulted in the Webrhicknesr Tuberhiciners Air Flaw F/n GEOMETRY OF A SEGMENT Figure I Schematic of a microchannel condenser and multilouverjn. Vishwanath Subramania
10、m is a graduate research assistant and Srinivas Garimella is an associate professor and director, Sustainable Ther- mal Systems Laboratory, Woodruff School of Mechanical Engineering, Georgia Institute of Technology, Atlanta. 02005 ASHRAE. 471 smallest heat exchanger mass among the fin geometries con
11、sidered within the constraints of the allowable pressure drop. According to the authors, this is because of secondary flows induced by the wavy fins, which increase the air-side heat transfer coefficient and result in a lower mass of the heat exchanger. Garimella et al. (1997) also considered the su
12、bstitution of conventional air-coupled single-phase exchangers for residential absorption heat pumps with such microchannel-tube, multilouver-fin heat exchangers. They demonstrated that the use of microchannel heat exchangers led to a 59% reduction in the heat exchanger mass compared to a round-tube
13、, wavy-fin heat exchanger. Condensation of ammonia in flat-tube, multilouver fin heat exchangers was investigated by Garimella and Wicht (1 995), who optimized the various geometric parameters to obtain a design heat duty of 21 kW with lowest heat exchanger mass. They found that tube and fin depth a
14、nd fin spacing had a significant effect on heat exchanger performance. Also, the available airflow was shown to have a significant effect on performance. A 20% decrease in airflow resulted in a 60% increase in the heat exchanger mass. Jiang and Garimella (2001) investigated air-coupled and hydronica
15、lly coupled heat pumps using microchannel tube, multilouver fin heat exchangers as evaporators and condensers with refrigerant R-22 and compared them with conventional round-tube systems. In the hydronically coupled system, the refrigerant in the evaporator transfers heat to an intermediate ethylene
16、-glycol loop, which is connected to the indoor/ outdoor air through air-to-hydronic fluid heat exchangers. They found that the indoor and outdoor units ofthe air-coupled microchannel system could be packaged in one-half and one- third the volume required for a conventional system, respectively. The
17、evaporator and condenser of the hydronically coupled system were found to require 35% and 65% less material than the air-coupled system, respectively, due to higher heat transfer coefficients in these counterflow heat exchangers with microchannels on both sides. The refrigerant charge required was f
18、ound to be 20% and 90% lower in the air-coupled and the hydronically coupled systems, respectively, as compared to a conventional round-tube system. Kim and Bullard (2002b) compared the performance of a microchannel condenser with a finned-tube condenser for a window room air conditioner using refri
19、gerant R-22. They concluded that the heat transfer rates per unit core volume are 14% to 331% higher for microchannel condensers as compared to conventional finned round tube condensers. Also, the refrigerant charge and condenser core volume and mass were found to be 35%, 55%, and 35% lower, respect
20、ively, for the microchannel condenser. Kim and Groll (2003) tested a 3 ton (10.5 kW) residential split heat pump system with refrigerant R-22 using a baseline spine-fin heat exchanger and a microchannel heat exchanger for the indoor and outdoor heat exchangers. The microchannel heat exchangers were
21、tested for vertically placed and 15“ angularly placed configurations with 6 and 8 fins per cm. The microchannel heat exchangers had about 23% less face area and 32% less refrigerant-side volume compared to the baseline heat exchanger. In the cooling mode, the COP was found to increase from 1 % to al
22、most 6% for the microchannel heat exchangers depending on fin density and heat exchanger orientation. The 15“ angular installation was found to perform better than the vertical installation, as the airflow to the heat exchanger was more normal in the angular installation. This closeness to normal ai
23、rflow increased the air-side heat transfer due to increased interruption of the thermal boundary layer on the louvered fins. Also, the 8 fins per cm system was found to perform better than the 6 fins per cm system. In the heating mode, the performance of the microchannel heat exchanger system was fo
24、und to be lower than that of the baseline case. Also, the vertical installation was found to offer a better performance than the 15“ angular installation. The 8 fins per cm system was found to offer a better performance than the 6 fins per cm system in the vertically placed configuration, while the
25、two were found to offer the same performance in the angularly placed case. In the heating mode, the baseline system was also found to have a lower frequency of defrost cycles compared to the case with microchannel heat exchangers. Among the microchannel heat exchangers, the frequency of defrost cycl
26、es in the heating mode was found to increase from the lowest value for the vertically placed 6 fins per cm system to progressively higher number of cycles for the 8 fins per cm vertically placed system, the 6 fins per cm angularly placed system, and the 8 fins per cm angularly placed system. In the
27、present study, a systematic procedure is developed to obtain the optimal geometrical parameters of a microchannel-tube, multilouver fin condenser for refrigerant blend R-4 1 OA, which is one of the leading replacement fluids for the currently used refrigerant R-22. The goal is once again to achieve
28、the desired duty with minimal material consumption. The configuration of these heat exchangers is such that it offers numerous opportunities for variation of refrigerant-side and air-side parameters to address a given design need. Typical parameters available for optimization include refrigerant-sid
29、e dimensions, such as height, depth, number of webs, and thicknesses, and fin parameters, such as height, density, and louver angle and pitch. These refrigerant- side and air-side parameters are depicted schematically in Figure 1. A computer code was developed to simulate the performance of the micr
30、ochannel condenser with refrigerant R-410A. This was used to study the effect of each of the geometric parameters on the condenser performance and arrive at an optimum design. ANALYSIS The basic geometry and the refrigerant-side and air-side details of the microchannel condenser analyzed in this stu
31、dy are shown in Figure 1. As this figure shows, the heat exchanger 472 ASHRAE Transactions: Symposia consists of a series of flat tubes with internal webs that form the microchannels. Multilouver fins are brazed to these tubes, with appropriately sectioned headers at the ends of the tubes that estab
32、lish the refrigerant-side pass arrangement, The entire condenser is constructed from aluminum. The louvers are fabricated in a roll-forming operation by creating slits in the fins and then bending the material at a desired angle. For the sake of simplicity, the fins are assumed to be perfectly recta
33、ngular and at right angles to the tube outer surface. Details of the calculation of the flow area, surface area, and other parameters required for the heat transfer and pressure drop analyses on both sides are provided in Appendix A. In a phase-change heat exchanger such as the one under considerati
34、on, the refrigerant liquid- and vapor-phase properties change along the flow with the changing refrigerant quality. To account for these variations and their effects on heat transfer coefficients and pressure drops, the heat exchanger tubes are divided into a number of small segments in the study. T
35、he refrigerant properties are assumed to be constant within each segment. The refrigerant inlet condition is specified for the first segment (and is the specified inlet condition for the entire condenser). The outlet condition for the segment is found by computing the corresponding refrigerant-side
36、and air-side heat transfer and pressure drop parameters, which in turn yield the segment condensation duty and pressure drops. Through the heat exchanger, the outlet condition of each upstream segment serves as the inlet condition for the following segment. The average of the inlet and outlet refrig
37、erant conditions is used to compute the refrigerant properties that are required for the computation of heat transfer coefficients and pressure drops for the segment. Since the outlet condition for any given segment is not initially known, an assumption about the outlet condition is first made and t
38、he heat duty calculated. Iterations are then performed to refine the segment heat duty and other parameters until convergence is achieved. The iterations were terminated when the error in the calculated segment heat duty was less than 0.0 1% between sucessive iterations. This process is repeated thr
39、ough the entire length of the heat exchanger. Details of the refrigerant-side and air-side heat transfer and pressure drop calculations are described below. Refrigerant Side Refrigerant-side calculations are described here with reference to the refrigerant-side and segment details shown in Figure 1.
40、 The refrigerant enters the condenser as a superheated vapor and leaves as a subcooled liquid. The refrigerant-side heat transfer and pressure drop calculations are therefore different for the superheated vapor, two-phase, and subcooled liquid regions. The properties of the refrigerant R-4 1 OA are
41、obtained from REFPROP 7.0 (NISS 2002). The refrigerant is assumed to be distributed uniformly among all tubes in a pass at the beginning of each new pass. The subcooled and superheated flow regime calculations treat the refrigerant as a single-phase fluid. For flow through rectangular channels, the
42、applicable flow regime is deter- mined using the criterion for transition from laminar to turbu- lent flow provided by Bhatti and Shah (1987), which in turn depends on the ratio of maximum and mean velocities in the duct given by Purday (1 949) and Natarajan and Lakshmanan (1972). These expressions
43、depend on the duct aspect ratio a given by a = thifweb,. For laminar flow, the single-phase refrigerant-side Nusselt number is calculated using the expres- sion for flow through rectangular ducts given by Bhatti and Shah (1987). In the absence of a more suitable rectangular flow correlation, the cor
44、responding turbulent flow Nusselt number is computed using the expression developed by Churchill (1977a) for circular tubes. This Nusselt number expression depends on the friction factor, which is given by Churchill (1 977b). Segment single-phase pressure drops in laminar flow are calculated using a
45、 correlation for friction factor in rectangular geometries recommended by Shah and London (1978). For turbulent flow, a correction factor in terms of the duct aspect ratio recommended by Bhatti and Shah (1987) is applied to the Churchill friction factor for circular tubes. For the saturated region o
46、f the condenser, the two-phase heat transfer coefficient is calculated using the Shah (I 979) correlation. It should be noted that other two-phase heat transfer correlations were also considered; however, due to the applicable thermal resistances on the air side and the refrigerant side, the heat du
47、ty is not very sensitive to the correlation used for condensation heat transfer coefficient. The pressure drop in the two-phase region was computed using the Friedel (1979) correlation. Since the heat transfer correlation used here was developed primarily for annular flow, it is expected that near t
48、he inlet and outlet of the condenser, where the flow regime could be mist and intermittent, respectively, the predicted heat transfer coefficient will not be very accurate. However, in view of the dominance of the air side in determining the condenser size, this is not likely to lead to any apprecia
49、ble errors. In these two- phase segments, as the fluid condenses, the bulk fluid decelerates, which results in a pressure rise. This increase in pressure due to deceleration is computed as follows: where the Butterworth (1975) model was used to compute the void fraction q,. Minor losses in the bends at the ends of the tubes are computed using the applicable loss coefficient. For the bends with two-phase flow, the multiplier suggested by Rohsenow et al. (1985) was used to obtain the minor loss as follows: “hend,2-phase “hen, iiq + 2x (2) p2 -phase ASHRAE Transactions: Symposia 473 1
