Refrigerant distribution in minichannel evaporator manifolds.

Received November 15, 2006; accepted January 18, 2007

The effects of geometry and operating conditions on the distribution of refrigerant R-410A in heat exchangers with horizontal manifolds and vertically oriented minichannels were experimentally investigated to provide the essential design information for the minichannel evaporators. Flow visualization in the horizontal manifold showed that the flow patterns of the two-phase refrigerant before and after the liquid-vapor transition are in stratified flow for the end-inlet location and in bubbly and stratified flows for the side-inlet location. Test results showed that the normalized standard deviation of the liquid mass flow rate, which indicates the degree of maldistribution, changes from 0.088 to 0.263 with manifold inlet location changes, from 0.034 to 0.141 with manifold inlet mass flow rate changes, from 0.037 to 0.082 with tube number changes, and from 0.027 to 0.055 with tube pitch changes. The normalized standard deviation of the liquid mass flow rates showed that the liquid refrigerant flow distribution is strongly affected by the manifold inlet location and the manifold inlet mass flow rate but is primarily independent of tube pitch. The side-inlet location showed a better liquid refrigerant flow distribution than the end-inlet location by more effectively mixing of the liquid and the vapor refrigerant from the inlet. Therefore, the side-inlet location is preferred to the end-inlet location.

INTRODUCTION

The two basic demands for air-conditioning and refrigeration systems are improving the efficiency and reducing the size of the equipment. Compact heat exchangers, especially aluminumbrazed heat exchangers, meet these demands very well. As manufacturing technology in aluminum-brazed compact heat exchangers advances, the hydraulic diameter of the tube channel has been reduced to a size of less than 1 mm. The reduced cross-sectional area of each tube requires many parallel tubes to keep the pressure drop across the heat exchangers within a reasonable range. Having many parallel tubes leads to refrigerant maldistribution issues when the minichannel heat exchanger is used as an evaporator. Refrigerant flow maldistribution is defined as a nonuniform distribution of the liquid flow rates. Since refrigerant maldistribution causes an overall deterioration of the heat exchanger performance, it is essential to understand the optimal design parameters and operating conditions.

Literature Review

Many researchers experimentally and numerically investigated this issue of refrigerant maldistribution in the evaporator for conventional and minichannel heat exchangers, as summarized in Table 1. Several researchers investigated two-phase distribution in manifolds with branch tubes (Horiki and Osakabe 1999; Kariyasaki et al. 1995; Rong et al. 1995; Zietlow et al. 2002; Tompkins et al. 2002; Lee and Lee 2005; Webb and Chung 2005; Watanabe et al. 1995a, 1995b; Asoh et al. 1991; Vist and Pettersen 2004; Fei et al. 2002; Sa et al. 2003; Cho et al. 2002; Vist 2004), most of which were conducted with air/water or low vapor pressure fluorocarbons. Only a few works were conducted with high-pressure working fluids such as R-22 and [CO.sub.2] (Sa et al. 2003; Cho et al. 2002; Vist 2004). It has been reported that gravity is an important force affecting the two-phase distribution in horizontal manifolds with vertical branch tubes. In a manifold with upward branch tubes, vapor flows predominantly through the branch tubes near the inlet, while liquid flows through the branch tubes near the far end of the manifold. Although many authors reported experimental measurements for the two-phase refrigerant distribution in the manifold, there are few works published regarding experimental measurements under realistic operating conditions, with realistic geometry and with R-410A as a working fluid. The objective of the current study is to provide essential design information through an experimental study of the effects of the geometry of the heat exchangers and operating conditions on the distribution of refrigerant R-410A in heat exchanger manifolds.

Table 1. Summary of Previous Works on Two-Phase Flow Distribution

Authors         Heat          Flow        Working      Inlet
                Exchanger     Direction   Fluids       Conditions
                              in
                              Manifold
                              and Tubes

Horiki and      Square        HM-VUT     Air-water    Adiabatic
Osakabe(1999)   manifold (40
                x 40 mm)
                with4 tubes
                ([D sub.i]
                10 mm)

Kariyasaki et   Manifold      HM-VUT     Air-water    Adiabatic
al. (1995)      ([D.sub.i]
                13.9 mm)
                with 3 tubes
                ([D sub.i]
                4.3 mm)

Rong et al.     Plate HX      HM-VDT     Air-water    Quality:
(1995)          with 7        HM-VUT                  0-0.22
                tubes                                 Adiabatic

Zietlow et al.  Manifold      HM-VDT     Air-water    MFR: 20 g/s
(2002)          with 19                               Quality: 0.18
                tubes                                 Adiabatic

Tompkins et     Manifold ([D  HM-VDT     Air-water    MF: 50-400
al. (2002)      sub.i] 14                             kg/[m.sup.2]
                mm) with 15                           s Quality:
                minichannel                           0-1
                tubes (6                              Adiabatic
                ports, [D
                sub.i] 1.6
                and 19.1 mm
                width)

Lee and Lee     Manifold ([D  VUM-HT     Air-water    MF: 70-165
(2005)          sub.i] 14                             kg/[m.sup.2] s
                mm) with 15                           Quality:
                flat tubes                            0.3-0.7
                (11.6 mm                              Adiabatic
                width, 1.4
                mm height,
                10 mm
                intervals)

Webb and Chung  D-shape       HM-VDT     Air-water    MFR: 10-60
(2005)          manifold      HM-VUT                  g/s Quality:
                with 20                               0.3-0.8
                minichannel                           Adiabatic
                tubes (11
                ports
                [D.sub.hi]
                1.3 and 25.4
                mm width)

Watanabe et     Manifold ([D  HM-VUT     R-11         MFR:
al. (1995a)     sub.i] 20                             12.6-37.7 g/s
                mm) with 4                            Quality:
                tube ([D                              0-0.3HF: 0,
                sub.i] 6 and                          6.25, 12.5
                40 mm                                 kW/[m.sup.2]
                intervals)

Watanabe et     Manifold ([D  HM-VUT     R-11         MF: 40-120
al. (1995b)     sub.i] 20     HM-HT                   (VT), 440-620
                mm) with n                            (HT)
                tubes ([D                             kg/[m.sup.2]
                sub.i] 6 and                          s Quality:
                40 mm                                 subcool-0.4
                intervals)                            Adiabatic
                (n: 2,3,4
                (VT), 3,4,5
                (HT))

Asoh et al.     Manifold ([   HM-VDT     R-113        MFR: 28-53
(1991)          sup.i] 13.6                           g/s Quality:
                mm) with 3                            0.1-0.25
                tubes ([                              Heat: 80-230
                sup.i] 4.5                            W
                mm)

Vist and        Manifold ([D  HM-VUT     R-134a       MFR: 2.5-4.2
Pettersen       sub.i] 8 and                          g/s Quality:
(2004)          16 mm) with                           0.11-0.50
                10 tubes ([D                          Pressure:
                sub.i] 4 mm,                          690-710 kPa
                21 mm tube
                pitch)

Fei et al.      Plate HX      HM-VDT     R-134a       MFR: 20-60
(2002)          with 5                                g/s Quality:
                tubes                                 0-0.30
                                                      Pressure:
                                                      690-760 kPa

Sa et al.       Minichannel   HM-VUT     R-22         MFR:
(2003)          evaporator:                           11.1-19.4 g/s
                manifold ([D                          Quality: 0.2
                sub.i] 19.4                           Pressure: 0.7
                mm) with 15                           MPa
                minichannel
                tubes (8
                ports,
                [D.sub.hi]
                1.2 mm)

Cho et al.      Manifold ([D  VM-HT      R-22         MF: 60
(2002)          sub.i] 19.4                           kg/[m.sup.2]
                mm) with 32                           s Quality:
                minichannel                           0.1-0.3
                tubes (10                             Pressure:
                ports, HD                             0.62 MPa
                1.5 mm)                               Adiabatic

Vist and        Manifold ([D  HM-VUT     [CO.sub.2]   MFR: 2.3-4.2
Pettersen       sub.i] 8 and                          g/s Quality:
(2003)          16 mm) with                           0.11-0.50
                10 tubes ([D                          Pressure:
                sub.i] 4                              5.5-5.6 MPa
                mm)

EXPERIMENT

Experimental Setup

An experimental facility with a flow visualization section mimicking a real heat exchanger manifold geometry was developed. Figures 1 and 2 show a schematic diagram of the experimental setup and an exploded diagram of the flow visualization section, respectively. The setup was composed of four major parts: the main refrigerant circuit, the test heat exchanger section with distributing manifold visualization, the post-heater section to measure mass flow rates and vapor qualities of the individual tube groups, and the condensing unit circuit. The pre-heater at the inlet of the test section was used to heat the subcooled liquid refrigerant to the desired inlet vapor quality. The branch tubes were heated by tape heaters. The refrigerant flow through each branch tube group in the heat exchanger was directed by three-way ball valves to either the collecting tube or the post-heater section. The distributing manifold was clear plastic and circular in shape with an inner diameter of 19.0 mm. Minichannel tubes were inserted into the centerline of the manifold. The manifolds were installed in a horizontal position with minichannel tubes of overall 1 m in length oriented in the vertically upward flow configuration. To measure the distribution characteristics along the distributing manifold, three branch tubes were grouped together at the collecting manifold using baffles. Tube groups were numbered sequentially from left to right. The minichannel tubes, which are shown in Figure 3, have six rectangular ports of 1.70 mm hydraulic diameter. The inlet pipe to the manifold was straight and ten times longer than the pipe inner diameter to ensure a fully developed flow that eliminates the impact of the upstream obstructions on the flow patterns. Measurements for the refrigerant distribution were conducted by measuring the mass flow rates and the vapor qualities of the individual tube groups. Uncertainties of the instruments are summarized in Table 2. Uncertainties of the manifold inlet vapor quality, the branch tube inlet vapor quality, and the branch tube mass flow rate (MFR) was [+ or -]0.01%, [+ or -]0.06%, and 0.35%, respectively.

[FIGURE 1 OMITTED]

[FIGURE 2 OMITTED]

[FIGURE 3 OMITTED]

Table 2. List of Instruments and Their Uncertainties

Physical     Instrument         Range          Uncertainty
Parameter

Temperature  T-type          -250[degrees]C    [+ or -]
             thermocouple         -            0.5[degrees]C
                              350[degrees]C

Pressure     Absolute         0-1,724 kPa      [+ or -]0.11%
             pressure
             transducer

Mass flow    Coriolis mass    0-75 g/s         [+ or -]0.1%
rate         flowmeter

Heater       Power meter      0-4 kW           [+ or -]0.5%
input

Experimental Parameters

The flow in the distributing manifold was monitored, and the distribution of the refrigerant mass flow rate and the vapor quality was measured for each tube group while the operating conditions and the heat exchanger geometry were varied, as summarized in Table 3. The experimental parameters listed in Table 3 are applicable to a typical residential air-conditioning system.

Table 3. Range of Experimental Parameters

Experimental Parameters                       Range

Inlet temperature                             7.2[degrees]C
Inlet quality                                 0.3
Refrigerant                                   R-410A
Manifold inlet mass flow rate, g/s            30/45/60
Heat load, kW                                 0/5/10
Heat exchanger tube pitch, mm/tube            8/10/12
Location of inlet                             End/Side
Number of heat exchanger tubes in parallel    18/24/30

With regard to the location of the inlet, the "end inlet" refers to the design with the inlet pipe located to the left side of the distributing manifold and aligned straight with the manifold centerline, while the "side inlet" refers to the design with the inlet pipe located in the middle of the manifold and aligned at a 90degree angle with the manifold centerline. The results of the refrigerant distribution in the distributing manifold are presented as the mass flow rate ratio (MFRR) of the vapor and the liquid phases in the ith branch tube group to total MFR as shown in Equation 1.

[R.sub.[bt,ph]](i)(%) = [[m.sub.[bt,ph]](i)/[N.summation over (i = 1)] [m.sub.[bt,ph]](i)] x 100 (1)

where ph = vap (vapor) and liq (liquid).

EXPERIMENTAL RESULTS

Flow Visualization Results

Figure 4 shows the flow pattern under adiabatic conditions for two inlet locations and two mass flow rates. For the end-inlet location, "stratified flow" or "stratified-wavy flow" pattern was observed so that vapor and liquid flowed at the upper and lower part of the distributing manifold, respectively, due to the gravitational effect. This flow pattern matches one identified from the flow pattern maps provided by Kattan et al. (1998) and Thome et al. (2002). Based on visual observation, the profile of the liquid-vapor interface was of a "stepwise" shape. The liquid level was almost constant and was located below the ends of the branch tubes near the inlet, while it was almost constant and was located above the ends of the branch tubes on the far end of the manifold. Between these two regions, there was a short transition area, where the liquid level increased abruptly in a slope. As the refrigerant MFR increased, the liquid-vapor interface (shown in dashed line) traveled farther from the inlet due to a higher momentum. Based on the measurement for the ten branch tube groups under adiabatic conditions, the liquid-vapor interface for the MFRs, 30, 45, and 60 g/s, was located right after the branch tube group numbers, 1, 2, and 3, respectively. It should be noted that another transition seems to occur just under branch tube group number 4 in Figure 4, but this is due to the different reflection between the rear sight glass and the main body, which can be further explained by the exploded diagram of the flow visualization section shown in Figure 2.

[FIGURE 4 OMITTED]

For the side-inlet location, the incoming refrigerant impinged on the inner side wall of the manifold and was divided symmetrically near the inlet. The flow patterns of two-phase refrigerant were in a bubbly flow before the liquid-vapor transition and in a stratified flow after the liquidvapor transition. The liquid-vapor interface was in a V-shape near the inlet. As the refrigerant MFR increased, the vapor traveled farther toward both ends of the manifold and the liquid-vapor interface took on a wider V-shape. At the far end of the manifold, the liquid level of the liquid-vapor interface was almost constant and was above the ends of the branch tubes. For both inlet locations, the locations of the liquid-vapor interface were almost independent of tube pitch. Additionally, as the heat load increased, the liquid-vapor interface traveled slightly farther from the inlet.

Flow Distribution Results

Effects of Manifold Inlet Mass Flow Rate. Figure 5 shows the refrigerant distribution results under the condition of 5 kW heat load for two inlet locations and three mass flow rates. For the end-inlet location, the profile of the branch tube inlet vapor quality was in a "stepwise" shape. Two regions of almost constant value existed, with a very short transition region between the two regions--one of about 100% vapor quality near the inlet and another of about 12% vapor quality near the end of the manifold. As the manifold inlet mass flow rate increased, the number of branch tube groups with almost 100% tube inlet vapor quality also increased because the liquid-vapor interface was located closer to the end of the manifold due to the increased momentum. At a manifold inlet mass flow rate of 30 g/s, 10% of the branch tubes had almost 100% tube inlet vapor quality, and at 60 g/s, about 30% of the branch tubes had almost 100% tube inlet vapor quality. However, for the side-inlet location, there was no region with 100% branch tube inlet vapor quality; instead, the profile of the branch tube inlet vapor quality was symmetric. The branch tube inlet vapor quality was about 60~70% near the inlet and about 20% near the end of the manifold. Between the two regions, the branch tube inlet vapor quality decreased steadily along the manifold.

[FIGURE 5 OMITTED]

For both inlet locations, the MFRRs deviated within a 2% range from the average MFRR. As the manifold inlet MFR increased, the MFRRs for the branch tube groups near the manifold inlet decreased slightly, while the MFRRs for the branch tube groups near the end of the manifold increased slightly because the liquid traveled farther due to the higher momentum. Moreover, the liquid MFRRs for the branch tube groups near the inlet were much lower than those for the branch tube groups located at the far end of the manifold. However, for the end-inlet location, approximately 7%~15% of the inlet manifold liquid MFR flowed through the first three branch tube groups near the inlet. For the side inlet, about 25% of the inlet manifold liquid MFR flowed through the four branch tube groups near the inlet.

Effects of Heat Load. Figure 6 shows refrigerant distribution results under the 60 g/s MFR condition for the two inlet locations and three heat loads. For both inlet locations, the vapor quality along the branch tubes also increased as the heat load increased. As a result, the gravitational pressure drop along the branch tubes was reduced for the tubes located far from the inlet. Consequently, more liquid refrigerant flowed to the far end of the manifold, and the liquid-vapor interface was located farther from the inlet as the heat load increased. However, the overall effect of the heat load on the liquid refrigerant flow distribution was within the measurement error range, except at the branch tube groups near the liquid-vapor interface.

[FIGURE 6 OMITTED]

Effects of the Number of Branch Tubes. Figure 7 shows the refrigerant distribution results under the 55 g/s MFR condition for the two inlet locations and three sets of branch tubes, 18, 24, and 30 each. As shown in the figure, the vapor quality profiles of the reduced number of branch tubes were similar to that of the 30-tube configuration for both inlet locations. Additionally, the liquid MFRRs increased as the number of branch tubes decreased for both inlet locations due to fewer tubes at the fixed manifold inlet MFR. In order to analyze quantitatively the degree of maldistribution for the three tube numbers, the normalized standard deviation of the liquid MFR of the branch tube groups was evaluated by using Equation 2.

[FIGURE 7 OMITTED]

[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] (2)

where

[AMFR.sub.liq] = [[N.sub.bt].summation over(i - 1)] [m.sub.bt,liq](i)/[N.sub.bt]

The smaller the value, the more uniform the distribution. The calculated values are noted in Table 4, which shows that the normalized standard deviations of the side-inlet location are smaller than those of the end-inlet location. For the end-inlet location, the 30-tube configuration has the lowest normalized standard deviation. For the side-inlet location, the 24-tube configuration has the lowest. Overall, the difference in the normalized standard deviations of the three configurations is within 0.1.

Table 4. Normalized Standard Deviation of Liquid MFR

MFR                                 30g/s                 55g/s

Inlet Location             End Inlet   Side Inlet  End Inlet  Side Inlet

Branch tube no.: 30 ea      0.408       0.320       0.549      0.286
Branch tube no.: 24 ea      0.450       0.235       0.595      0.202
Branch tube no.: 18 ea      0.466       0.256       0.631      0.249

Effects of the Tube Pitch. Figure 8 shows the refrigerant distribution at 55 g/s MFR and 5 kW heat load conditions for the two inlet locations and three tube pitches, 8, 10, and 12 mm. The figure shows that the vapor quality and the liquid MFRR distribution are barely affected by the variation of the tube pitch for both inlet locations. In order to assess quantitatively the degree of maldistribution for the three tube pitches, the normalized standard deviation of the liquid MFR at the branch tube groups was calculated as shown in Table 5. It can be concluded from these results that the vapor and liquid distribution in the manifold is barely affected by the variation of the tube pitch.

[FIGURE 8 OMITTED]

Table 5. Normalized Standard Deviation of Liquid MFR

MFR                         30 g/s                 55 g/s

Inlet Location       End Inlet  Side Inlet  End Inlet  Side Inlet

Tube pitch: 10 mm      0.408       0.320      0.549       0.286
Tube pitch:  8 mm      0.438       0.354      0.580       0.294
Tube pitch: 12 mm      0.439       0.368      0.576       0.341

CONCLUSIONS

Since the state of the refrigerant entering the evaporator is in two-phase and the vapor quality of the refrigerant changes upon the operating conditions, the proper refrigerant distribution to a multiple number of tubes is a challenging task. In order to provide essential design information for the minichannel evaporators, an experimental study was conducted to determine the effects of geometry and operating conditions on the distribution of refrigerant in the horizontal evaporator manifold. Based on the observations made from the flow visualization and experimental data of the flow distribution, the following characteristics were observed.

* The gravitational force and the momentum difference between the liquid and the vapor phases affect the phase separation in the horizontal manifold. Low momentum vapor on the top layer is easily taken off in the branch tubes near the inlet, while the high momentum liquid on the bottom layer travels farther toward the far end of the manifold.

* For the end-inlet location, the profile of the branch tube inlet vapor quality is of a "stepwise" shape. As the manifold inlet mass flow rate increases, the number of branch tube groups having almost 100% tube inlet vapor quality also increases because the liquid-vapor interface is located closer to the end of the manifold due to the increased momentum. However, for the side-inlet location, there are no regions with 100% branch tube inlet vapor quality, and the profile of the branch tube inlet vapor quality is symmetric. The branch tube inlet vapor quality is about 60% to 70% near the inlet and about 20% near the end of the manifold. Between these two regions, the branch tube inlet vapor quality decreases steadily along the manifold.

* The normalized standard deviation of the liquid mass flow rate, which indicates the degree of maldistribution, changes from 0.088 to 0.263 with manifold inlet location changes, from 0.034 to 0.141 with manifold inlet mass flow rate changes, from 0.037 to 0.082 with tube number changes, and from 0.027 to 0.055 with tube pitch changes. Therefore, it can be concluded that the liquid refrigerant flow distribution is rather strongly affected by the manifold inlet location and the manifold inlet mass flow rate but is primarily independent of the tube pitch.

* Overall, the side-inlet location is preferred to the end-inlet location for the range of mass flow rates investigated.

ACKNOWLEDGMENTS

This work was sponsored by the American Society of Heating, Refrigerating and Air-Conditioning Engineers, Inc. (ASHRAE), and the Center for Environmental Energy Engineering (CEEE) at the University of Maryland. The feedback and technical guidance of the project monitoring subgroup is gratefully acknowledged.

NOMENCLATURE 

AMFR = average mass flow rate

HM = horizontal flow in manifold

HL = heat load

HT = horizontal flow in tube

m = mass flow rate

MF = mass flux 

MFR = mass flow rate

MFRR = mass flow rate ratio

N = number of tube group

NSTD = normalized standard deviation

R = ratio

TN = tube number

TP = tube pitch

VDT = vertical downward flow in tube

VUT = vertical upward flow in tube

VUM = vertical upward flow in manifold

x = vapor quality

Subscripts

bt = branch tube

dm = distributing manifold

liq = liquid

ph = phase

vap = vapor

REFERENCES

Asoh, M., Y. Hirao, Y. Watanabe, and T. Fukano. 1991. Phase separation of refrigerant two-phase mixture flowing downward into three thin branches from a horizontal manifold pipe. ASME/JSME Thermal Engineering Proceedings 2:159-64.

Cho, H., K. Cho, B. Youn, and Y.S. Kim. 2002. Flow maldistribution in microchannel evaporator. Ninth International Refrigeration and Air Conditioning Conference, July 16, Purdue University, West Lafayette, IN [Paper no. R6-5 on CD].

Fei, P., D. Cantrak, and P. Hrnjak. 2002. Refrigerant distribution in the inlet header of plate evaporators. SAE World Congress, Detroit, MI, Paper 2002-01-0948.

Horiki, S., and M. Osakabe. 1999. Water flow distribution in horizontal protruding-type header contaminated with bubbles. Proceedings of the ASME, HTD 364(2).

Kariyasaki, A., T. Nagashima, T. Fukano, and A. Ousaka. 1995. Flow separation characteristics of horizontal two-phase air-water flow into successive horizontal capillary tubes through T-junction. Proceedings of ASME/JSME Thermal Engineering Conference 2:75-82.

Kattan, N., J.R. Thome, and D. Favrat. 1998. Flow boiling in horizontal tubes, Part 1: Development of a diabatic two-phase flow pattern map. Journal of Heat Transfer 120:140-47.

Lee, S., and J. Lee. 2005. Aspects of two-phase flow distribution at header-channels assembly. Proceedings of 5th International Conference on Enhanced, Compact, and Ultra-Compact Heat Exchangers, pp. 351-63.

Rong, X., M. Kawaji, and J.G. Burgers. 1995. Two-phase flow distribution in a stacked plate heat exchanger. ASME/JSME Fluid Engineering and Laser Anemometry Conference and Exhibition, August 13-18, Hilton Head, SC.

Sa, Y.C., D.Y. Jang, C.S. Ko, S.Y. Oh, and B.Y. Chung. 2003. Flow maldistribution of flat tube evaporator. Proceedings of The 4th International Symposium on HVAC, Beijing, China, pp. 817-22.

Thome, J.R., and J. El Hajal. 2002. Two-phase flow pattern map for evaporation in horizontal tubes: Latest version. Heat Transfer Engineering 24:3-10.

Tompkins, D.M., T. Yoo, P.S. Hrnjak, T.A. Newell, and K. Cho. 2002. Flow distribution and pressure drop in micro-channel manifolds. 2002 International Refrigeration Conference, Purdue University, West Lafayette, IN, July, Paper R6-4.

Vist, S. 2004. Two-phase refrigerant distribution in round tube manifolds. ASHRAE Transactions 110(1):307-17.

Vist, S., and J. Pettersen. 2004. Two-phase flow distribution in compact heat exchanger manifolds. Experimental Thermal and Fluid Science 28:209-15.

Watanabe, M., M. Katsuta, and K. Nagata. 1995a. Two-phase flow distribution in multi-pass tube modeling serpentine type evaporator. Proceedings of the ASME/JSME Thermal Engineering Conference 2:35-42.

Watanabe, M., M. Katsuta, and K. Nagata. 1995a. General characteristics of two-phase flow distribution in a multipass tube. Heat Transfer, Japanese Research 24:32-44.

Webb, R.L., and K. Chung. 2005. Two-phase flow distribution to tubes of parallel flow air-cooled heat exchangers. Heat Transfer Engineering 26(4):3-18.

Zietlow, D.C., M. Campagna, and J.F. Dias. 2002. Innovative experimental apparatus to measure liquid flow distribution in two-phase flow occurring in the manifolds of compact heat exchangers. ASHRAE Transactions 108(2):501-08.

Yunho Hwang, PhD Member ASHRAE

Dae-Hyun Jin, PhD Student Member ASHRAE

Reinhard Radermacher, PhD Fellow ASHRAE

Yunho Hwang is a research associate professor, Dae-Hyun Jin is a graduate research assistant, and Reinhard Radermacher is a professor in the Department of Mechanical Engineering at the University of Maryland, College Park, MD.