7th International Conference on Multiphase Flow
ICMF 2010, Tampa, FL USA, May 30June 4, 2010
Experimental and CFD Analysis of Two Phase Flow of Refrigerants
inside a Horizontal Tube for the Evaluation of Pressure drop
Bhramara Panitapu, Kishen K. Reddy T and Sharma K V
Mechanical Engineering, JNTUH College of Engineering, JNT University, Hyderabad
Hyderabad, 500085, INDIA
bhramara74@ yahoo.com
Keywords: Two Phase Flow, refrigerants, VOF model, pressure drop
Abstract
Intube two phase flow finds its applications in steam power plants, refrigeration and air conditioning, distillation and
desalination units and transport of oil and slurry etc. The two phase flow is described by the parameters like vapor quality, void
fraction and Martinelli Parameter etc while the performance parameters of two phase flow are given by heat transfer
coefficient and pressure drop. The salient feature of intube two phase flow is the formation of flow regimes, viz., stratified,
slug, annular etc due to counter acting forces of gravity and vapor shear. The present study focuses primarily on the
development of test facility to evaluate the two phase pressure drop of refrigerants, R22, R134a and R407C at high pressures
with a maximum pressure limit of 20 bar. Using the three refrigerants, the pressure drop of two phase flow is studied in the
pressure range of 10 16 bar. Secondly, the present study aims to develop a predictive procedure using CFD analysis for the
prediction of flow regimes and pressure drop. The so obtained pressure drop is compared with the experimental results and
some of the widely used correlations of pressure drop and the flow regimes obtained from CFD analysis are compared with the
Thome et al. [2003] flow regime map.
Introduction
Two phase flow in a horizontal tube has
widespread applications, particularly in the condensers and
evaporators of refrigeration and air conditioning systems.
Tendency for flow stratification is one of the main features
of internal horizontal flow compared to vertical and
inclined flows. The vapor tends to migrate towards the
top portion of the tube or channel during condensation
while the lower portion of the channel carries more of the
liquid owing to density difference, regardless of the flow
regime. The flow regimes for airwater mixture given by
Ewing et al [1999] are shown in Figure 1.
a) Stratified
c) Wavy annular
b) Wavy
d) Annular
Figure 1. Flow Regimes at high void fractions
of airwater mixture [1999]
1Atnn
1000
soo
800
f 600
400
200
0 0.2 0.4 0.6 0.8
Vapor Quality
Figure 2. Thome et al. [2003] Flow regime map for R134a
The early attempts made in classifying different
regimes of two phase flow and presentation in the form of
map can be attributed to Baker [1954]. Since then
numerous investigators have contributed to the
understanding of two phase flow phenomena which is
presented chronologically for horizontal cocurrent flow.
Recently developed two phase flow pattern map for
condensation in horizontal tubes is by Thome et al. [2003]
by adapting KattanThomeFavrat flow pattern map [1998]
for evaporation. They employed a new logarithmic mean
void fraction method to handle void fractions from low to
high reduced pressures. The flow regimes were classified by
them as fully stratified flow (S), stratifiedwavy flow (SW),
 r  e  1/n2
i G500 Kgj m2s
 .,.. ..  10 NO 2.........
I  G=100Krl ms
sA
S  
a ,s v
intermittent flow (I) referring to both slug and plug flows,
annular flow (A), mist flow (MF) and bubbly flow (B).
Figure 2 shows the flow regime transitions for R134a at
different mass flux condensing inside an 8 mm diameter
tube at 400C.
The two phase multipliers of Lockhart and
Martinelli [1947] for adiabatic airwater mixtures were the
pioneering work in the modeling of two phase flows. Their
correlations were later modified for diabetic flows by
Martnelli and Nelson [1948]. These multipliers are
functions of Martinelli parameter, X which is
dimensionless combination of the physical properties.
Subsequently, X is being used in several convective
condensation and boiling correlations as one of the
governing parameters. The generality of Lockhart and
Martinelli multipliers is thus well acclaimed in two phase
studies. Later many correlations were developed using two
phase multiplier approach.
Gr6nnerud correlation [1979] is developed for
refrigerants. Chisholm method [1973] is recommended for
fluids with property index, (PG/, )O /( /P ) > 0.01.
Friedel [1979] developed a correlation for two phase
multiplier for vertical upward and horizontal flow in round
tubes and is recommended for fluids with (p/, l) < 1000.
MiullerSteinhagen and Heck [1986] proposed an empirical
interpolation between all liquid and all vapor flow.
All these correlations though developed for two
phase flows at atmospheric pressure and for evaporating
flows, were used extensively for pressure drop predictions
and analytical modeling of condensing flows also.
The quasi local experimental work reported in the
literature is primarily to study the performance of
alternative refrigerants. Shao et al. [1998], Boissieux et al.
[2000], Smit et al. [2002a, 2002b], Infante Ferreira et al.
[2003] experimentally studied the performance of
refrigerant mixtures to test their applicability as alternative
refrigerants. Lee et al. [2006] reported the performance of
hydro carbons for condensing flows.
Recent experimental work reported focuses on
the development of better predictive procedures for thermal
design of condensers. Cavallini et al. [2001] had conducted
experimental investigations of HFCs, inside a horizontal
tube of diameter, 8mm and length, im for a mass flux
ranging from 100 to 750 kg/m2s. They also measured
pressure drop data for HFCs and observed that low
pressure fluids show higher pressure drop. They suggested
the use of Friedel model [1979] for predicting frictional
pressure drop.
A more recent experimental investigation of
pressure drop and heat transfer for intube condensation of
ammonia with and without miscible oil inside smooth
aluminum tube of diameter, 8.1 mm for a mass flux range
of 20270 kg/m2s is performed by Park and Hrnjak [2008].
They also measured pressure drop of ammonia and
observed that MUiller Steinhagen and Heck and Friedel
correlations based on separated flow model predict the
pressure drop relatively well at pressure drop higher than 1
kPa/m, while a homogeneous model [1992] (McAdams
Model) yielded acceptable values at pressure drop less than
1 kPa/m.
7th International Conference on Multiphase Flow
ICMF 2010, Tampa, FL USA, May 30June 4, 2010
Limited work is reported in the literature on the
modeling of two phase flows using CFD analysis. The
recent work of Schepper et al. [2008] presented the flow
regime predictions for airwater and gasoil mixtures at
atmospheric pressure using CFD analysis and compared
with Baker [1954] flow regime map. Except slug flow
regime, they could simulate all other flow regimes using
VOF model from the commercial CFD software, FLUENT.
In the present study, their work is extended to the
vaporliquid flow of refrigerants at high pressures.
The present study focuses primarily on the
development of test facility to evaluate the two phase heat
transfer coefficient and pressure drop of refrigerants, R22,
R134a and R407C at high pressures with a maximum
pressure limit of 20 bar. Secondly, the present study aims to
develop a predictive procedure using CFD analysis for the
prediction of flow regimes and pressure drop
Nomenclature
x Vapor Quality ( )
X Martinelli Parameter ( )
F External Force (N)
G Mass Flux ( kg/m2s)
N Number of phases ()
S Source Term ( )
U Velocity vector (m/s)
t Time (s)
Greek letters
Y Dynamic viscosity ( Pas)
P Density ( kg/m3)
a Volume fraction ()
Subsripts
I Liquid
v Vapor
k Phase number
Experimental Facility
The test facility is designed based on quasi local
experimental technique using which the refrigerant is partly
condensed inside the test section with a small decrement of
vapor quality, (A ), typically in the range of 0.13 to 0.33.
The experimental heat transfer coefficient and pressure
drop are reported for the average of inlet and outlet quality
for the test section and hence are referred as quasi local
values.
The experiments were conducted for different
mass fluxes of 200, 400 and 600 kg/m2s at a saturation
pressure range of 10 16 bar. The quasi local vapor quality
of refrigerants ranges from 0.36 to 0.87. In total, 32 sensors
for measuring parameters viz., temperature, flow, pressure
and differential pressure are logged to the computer and the
remaining are measured using analog devices.
The main features of test facility are shown in
Figure 3. The test facility is skid mount type with all the
components housed on MS frame as shown in Figure 3a).
All the components are mounted on the skid using
vibration pads.
The setup is based on vapor compression cycle
with condensation performed in four heat exchangers, viz.,
pre condenser, test section, after condenser and bypass
condenser. Figure 3b) shows the insulated condensers and
evaporator. All heat exchangers are plate type, except the
test section which is concentric tube type. The test section
is made of hard drawn copper of inner diameter of the inner
tube, 8mm and length, 1200 mm. The main feature of the
test facility is that a separate set of scroll compressor along
with suction accumulators as shown in Figure 3c),
expansion valve and receiver are used for each of the
refrigerants, R22, R134a and R407C to avoid the complete
evacuation of compressor while charging and testing each
refrigerant. Figures 3d) and 3e) shows the data acquisition
software and hardware.
7th International Conference on Multiphase Flow
ICMF 2010, Tampa, FL USA, May 30June 4, 2010
The set up basically consists of two main
subsystems, one refrigerant loop and three water loops.
The refrigerant cycle is shown in Figure 4. Three water
loops, each for test section, for evaporator and for three
condensers supply water at the required operating
temperature.
The insulated water tanks are provided with
chillers and heaters to maintain required source/ sink
temperature. Only those readings are taken for which the
ratio of difference between the refrigerant side heat
loss/gain and the coolant side heat gain/loss to the coolant
side heat gain/loss is less than 1%. The pressure drop
across the test section is measured by Rosemount
differential pressure transducer. The experimental
uncertainty in the measurement of pressure drop is 0.081
kPa.
Figure 3 Main Features of Test Facility a) Skid Mount Test Facility b) Heat Exchangers
c) Scroll Compressors and Suction Accumulators
d) Data Acquisition Window and e) Data Acquisition Hard ware
CFD Analysis
The CFD analysis of two phase flow using VOF
model with second order Geo reconstruct scheme for
interface interpolation, PISO algorithm [2005] for
pressurevelocity coupling, PRESTO algorithm for pressure
interpolation [2005], a second order upwind calculation
scheme [2005] for the determination of momentum and
volume fraction; and with appropriate boundary conditions
is performed using commercial CFD software, FLUENT.
In the volume of fluid (VOF) model, a single set of
conservation equations is shared by the phases and the
volume fraction of each of the phases is tracked in each
computational cell throughout the domain. These
conservation equations can be solved using the appropriate
boundary conditions at the interface.
Governing Equations
In the VOF approach, the participating fluids share
a single set of conservation equations. The governing
equations can therefore be written as,
Conservation of Mass:
(p) + V.(pU) = S (1)
at
Conservation of Momentum :
(pU)+V.(pUU) =V. + pg + F (2)
at
The first term of the lefthand side of the mass
conservation equation represents accumulation; the second
term represents the contribution of convection. The term
on the righthand side represents the sum of volumetric
sources of all the fluids which is zero in the present case as
only flow is considered here. Similarly, the first term on
the lefthand side of momentum equation is the rate of
increase in momentum per unit volume; the second term
represents the change in momentum per unit volume,
caused by convection. The first term on the righthand side
represents molecular contributions, which include pressure
and viscous force per unit volume. The last two terms on
the righthand side represent the gravitational force per
unit volume and any other external force. The numerical
solution of the set of Eqs. (1) and (2) for multiphase flows
is extremely difficult and computationally intensive. The
main difficulty arises from the interaction between the
moving interface and the fixed Eulerian grid that is
employed to solve the flow field.
The motion of the interface is deduced indirectly
from the motion of different phases separated by an
interface. Motion of the different phases is tracked by
solving a continuity equation for a marker function or for
the volume fraction of each phase. Thus, when a control
volume is not entirely occupied by one phase, mixture
properties are used while solving Eqs. (1) and (2). This
approach avoids abrupt changes in properties across a very
thin interface. The properties appearing in Eqs. (1) and (2)
are related to the volume fraction of all phases as follows:
P=a p ; pp (3)
The volume fraction of each fluid a, is
calculated by tracking the interface between different
7th International Conference on Multiphase Flow
ICMF 2010, Tampa, FL USA, May 30June 4, 2010
phases throughout the solution domain. Tracking of the
interfaces between N different phases present in the system
is accomplished by solving continuity equations of the
phase volume fraction for Nl phases. For the k th phase,
this equation has the following form:
S(u,.V) a,= S,
The first term of the lefthand side represents
accumulation; the second term represents the contribution
of convection. The term on the righthand side represents
the sum of volumetric sources of the volume fraction. As
only flow is considered here, this contribution equals zero.
The simulations are carried out for the inner tube
of test section which is a horizontal tube with a diameter of
8 mm and a length of 1200 mm under adiabatic conditions.
To test the grid independency, the wall shear stress for
different grids is computed for R22 at different mass fluxes
using five different grids as shown in Figure 4. The wall
shear stress is considered to study grid independency as it
quantifies the boundary layer phenomenon particularly at
medium to high vapor qualities where a very thin liquid film
forms around the circumference of the tube due to the onset
of annular flow regime. Figure 5 shows the variation of wall
shear stress with vapor quality for a mass flux of 400 kg/m2s.
If less number of cells is available at the wall, lesser number
of cells will be patched with liquid phase and the resulting
wall stress will exhibit an oscillating trend with the vapor
quality as shown in Figure 4.
0.2 0.4 0.6
Vapor Quality
0.8 1.0
Figure 4 Variation of Wall Shear Stress with Different
Grids for R22 at G= 400 kg/m2s
Based on the grid independence study, the mesh
model with 278,712 hexahedral cells is selected for
simulations. The wall shear stress is considered to study grid
independency as it quantifies the boundary layer
phenomenon particularly at medium to high vapor qualities
where a very thin liquid film forms around the
circumference of the tube due to the onset of annular flow
regime.
The influence of the gravitational force on the flow
has been taken into account as the flow stratification is the
main feature of two phase horizontal flow.
The properties of refrigerants are taken from
REFPROP Version. 6.01. First, the analysis is performed at
steady state for a given vapor quality to determine the flow
field of one of the phases so that the starting point for the
transient simulation is fully converged flow field of one
phase.
.
R22, G= 400 kg/m2s
i 332.196 cells
* 278.712 cells
s P 232.864 cells
y W 193.890 cells
 181,000 cells
Transient Simulations Prediction of Flow Regimes
The flow regimes are obtained by plotting the
contours of mixture density. As mixture density is
proportional to its phase composition, the distribution of
vapor and liquid phases of refrigerant is clearly seen in these
contours. The flow conditions selected based on the flow
regime predictions of Thome et al. [2003a] flow regime
maps for R22, R134a and R407C that spread in different
flow regimes are simulated, including the flow regime
transitions. The flow conditions simulated and the
corresponding flow regimes predicted by Thome et al. map
as shown in Figure 5 are presented in the Table 1. The
contours of mixture density for R22 based on transient
simulations at different flow conditions as mentioned in
Table 1 are presented in Figure 6. The flow regimes
obtained by Ewing et al. [1999] are also shown to
corroborate the respective flow regimes.
SG 100 kgs/ms
S10 2 00akg/mn st
SG= 400 kg/ms
v G= 600 Kg/ms b)
0A b)
Intermediate Annular
V V V
10 A A A
00 02 04 06 08 1
Vapor Quality
800 
00 02 04 06
Vapor Quality
Figure 5 Flow Conditions and Flow Regimes
Predicted by Thome et al Flow Regime Maps
for a) R22 b) R134a and c) R407C
The red color represents the pure liquid
refrigerant and blue color represents pure vapor. The scale in
the left hand side of Figure 6.1 represents the density
variation of mixture from liquid density of 1129 kg/m3 to
7th International Conference on Multiphase Flow
ICMF 2010, Tampa, FL USA, May 30June 4, 2010
vapor density of 66.2 kg/m3 of R22. The contours of
mixture density represent a stratified wavy flow regime for
the flow conditions mentioned for low mass flux and at low
qualities which is similar to wavy regime given by Ewing et
al. [1999] for airwater mixture shown in Figure 6.1e). The
Ewing's flow regimes shows very unstable interface with
vigorous interfacial waves compared the respective contours
shown in Figures 6.1 to 6.4. This is due to larger value of
liquid to vapor density ratio of airwater mixture at
atmosphere compared to liquid to vapor density ratio of R22
at a high pressure of 15.56 bar.
The contours of mixture density at a mass flux
of 200 kg/m2s and a vapor quality of 0.5 are shown in
Figure 6.2 at different time steps of 100, 200 and 500. The
interface becoming more unstable at a vapor quality of 0.5
compared to that at 0.3 can be clearly seen from Figures.
6.1c) and 6.2a) for the same mass flux, 200 kg/m2s. Waves
trapping liquid slugs can also be seen in Figure 6.2a). The
progressive rising of waves from the bottom pool of liquid
and touching the upper portion of tube to form a wavy
annular flow regime with a discontinuous liquid film
around the circumference is seen is Figures 6.2 a) to 6.2 c).
A very unstable interface with liquid slugs can
be clearly seen in the cross section view represented in
Figure 6.3 c). Figures. 6.3 a) and 6.3 b) show the increased
slug formation with the increase of mass flux.
The contours of mixture density obtained at a
vapor quality of 0.6 for mass fluxes, 400 and 600 kg/m2s
are shown in Figure 6.4. A thin annular film with vapor
flowing in the core can be seen in cross sectional view
given in Figure 6.4c). The thickness of the liquid film at
the bottom of the tube is more compared to the top owing
to high density of liquid as shown in Figure 6.4c). Thus for
all refrigerants considered in the present study, the
contours of mixture density are in an excellent agreement
with the predicted flow regimes mentioned in Table 1 and
match the flow regimes given by Ewing et al. [1999]. Thus
the work of Schepper et al. [2008] is successfully extended
to the vapor liquid flow of refrigerants at high pressures.
This shows that the phase volume fractions and mixture
density are unaffected by the slight variations in
temperature that occur during condensation. Hence the
wall shear stress and pressure drop across the tube is
evaluated under adiabatic conditions.
Steady State Simulation Results Prediction of
Pressure drop
The pressure gradient obtained from CFD
analysis using VOF model is compared with the
experimental data and with pressure drop correlations
available in the literature. The graphs of comparison are
presented for R22.
At low mass flux, only CFD model and MUiller 
Steinhagen and Heck correlation predicted the experimental
data within 20 40% deviation as shown in Figures. 7a) and
7d). All other correlations exhibited further larger
deviations. At a medium mass flux of 400 kg/m2s, CFD data
closely follows the trend of experimental data with a
tendency of over prediction compared to the correlations
and its predictions are the best with a deviation of 5% from
the experimental data as shown in Figures 7b) and 7e). At a
high mass flux of 600 kg/m2s, the CFD predictions show an
* G= 100 kglm2s
0 G= 200 kg/m2s
 A G= 40Okgms T a)
SG= 600 kglm2s R
A
Intermediate N Annular
S
V V
S StrafedWavy
00 02 04 06
Vapor Quality x
G= 100 kg/m's
G= 200 kgfm's
G= 400 kgfm s T
G= 600 kg/m s R
A
Intermediate N Annular
A rN A
ratified Wa
Stratled Wav 0
08 1 0
1000
08 10
excellent agreement with the experimental data with a
deviation of 12%, as represented in Figure 7f).
Thus the pressure drop data obtained from CFD
simulations using VOF model is in good agreement with
the experimental pressure drop data, compared to the
correlations of pressure drop for the refrigerants considered
in the present study irrespective of the refrigerant and mass
flux considered.
Conclusions
The objectives of the present study are achieved
primarily by developing a test facility to evaluate the two
phase heat transfer and pressure drop at high pressures
using refrigerants, R22, R134a and R407C. Secondly, by
developing a predictive procedure for pressure drop using
a combination of CFD analysis.
The VOF model in the commercial CFD
software, FLUENT perfectly tracked the
geometry of vapor liquid interface of
refrigerants as the resulting mixture density
contours are in excellent agreement with the
flow regimes predicted using Thome et al.
[2003a] flow regime map.
The work of Schepper et al. [2008] for
airwater and gasoil mixtures at atmospheric
pressures is successfully extended to
vaporliquid flow of refrigerants at high
pressures as all the flow regimes are
simulated in the CFD analysis.
Pressure drop data estimated using VOF
model is in good agreement with the
experimental data compared to the pressure
drop correlations with a minimum deviation
of 3% and a maximum deviation of 16%.
Acknowledgements
Acknowledgements are extended to AICTE for granting
funds to develop the test facility and M/s Hemair Systems
India Ltd, Hyderabad for fabricating the experimental set
up.
References
Baker, O., Simultaneous Flow of Oil and Gas, The Oil and
Gas Journal, Vol.53, pp.185195, 1954.
Boissieux, X., Heikal, M.R., and Johns, R.A., Twophase
Heat Transfer Coefficients of three HFC Refrigerants inside
a Horizontal Smooth Tube, Part II: Condensation, Int. J. of
Refrigeration, Vol.23, pp.345352, 2000.
Carey, V.P, LiquidVapor Phase Change Phenomena, Taylor
& Francis Group, 1992.
Cavallini, A., Censi, G., Del Col, D., Doretti, L., Longo,
G.A., and Rosetto, L., Intube Condensation of Halogenated
Refrigerants, ASHRAE Transactions, Vol.108, No.2, paper
4507, 2002.
7th International Conference on Multiphase Flow
ICMF 2010, Tampa, FL USA, May 30June 4, 2010
Chisholm, D., Pressure gradients due to friction during the
flow of evaporating two phase mixtures in smooth tubes and
channels, Int. J. Heat and Mass Transfer, No. 16,
pp.347358, 1973.
El Hajal, J., Thome, J.R., Cavallini, A., Condensation in
Horizontal Tubes, Part I Two Phase Flow Pattern Map, Int.
J. of Heat and Mass Transfer, Vol. 46, pp.33493363, 2003.
Ewing, M.E., Weinandy, J.J., Christensen, R.N.,
Observations of Two Phase Flow Patterns in a Horizontal
Circular Channel, Heat Transfer Engineering, Vol.20, No.l,
pp.914, 1999.
Friedel, L., Improved friction pressure drop correlations for
horizontal and vertical two phase pipe flow, paper E2,
European Two Phase Flow Group Meeting, Ispra, Italy,
1979.
Fluent Inc., Fluent 6.2 User Guide, Fluent Inc., Lebanon,
USA, 2005.
Gr6nnerud. R., Investigation of liquid holdup,
flowresistance and heat transfer in circulation type of
evaporators, part iv: twophase flow resistance in boiling
refrigerans. In Annexel9721, Bull. de l'Inst. du Froid,
1979.
Infante Ferreira, C.A., Newell, T.A., Chato, J.C., Nan, X.,
R404A Condensing Under Forced Flow Conditions inside
Smooth, Microfin and CrossHatched Horizontal Tubes, Int.
J. of Refrigeration, Vol.26, pp.433441, 2003.
Kattan, N., Thome, J.R., Favrat, D., Flow Boiling in
Horizontal Tubes: Part I Development of Flow Regime
Map, ASME Journal of Heat Transfer, Vol.120, pp.140147,
1998.
Lee, H.S., JungIn Yoon, Kim, J.D., Bansal, PK.,
Condensing Heat Transfer and Pressure drop Characteristics
of Hydro Carbon Refrigerants, Int. J. of Heat and Mass
Transfer, Vol.49, pp.19221927, 2006.
Lockhart, R. W., Martinelli, R. C., Proposed correlation of
data for isothermal two phase, two component flow in pipes,
C im '~. I Engineering Proceedings, Vol. 45, No.l, pp. 3948,
1947.
Martinelli, R.C., Nelson, D. B., Prediction of pressure drop
during forced circulation boiling of water, Trans. ASME,
Vol. 70, pp. 695702, 1948.
MillerSteinhagen. H., Heck. K., A Simple Friction Pressure
Drop Correlation for TwoPhase Flow in Pipes, Chem. Eng.
Process, vol. 20 pp. 297308, 1986.
Park, C.Y., Hrnjak, P, NH3 InTube Condensation Heat
Transfer and Pressure drop in Smooth Tube, Int.J. of
Refrigeration, pp.19, 2008.
Schepper, S.C.K., Heynderickx, G.J., Martin, G.B., CFD
Modeling of All GasLiquid and VaporLiquid Flow
Regimes Predicted by Baker Chart, Chemical Engineering
Journal, Vol.138, pp.349357, 2008.
Shao, D.W., Granryd, E.G., Flow Pattern, Heat Transfer and
Pressure drop in Flow Condensation, Part I: Pure and
Azeotropic Mixtures, Int. J. of HVAC & R, Vol.6, No.2,
pp.175195, 2000.
Smit, F.J., Meyer, J.P, Condensation Heat Transfer
Coefficients of Zeotropic Refrigerant Mixture, R22/R142b
7th International Conference on Multiphase Flow
ICMF 2010, Tampa, FL USA, May 30June 4, 2010
in Smooth Horizontal Tubes, Int. J. of Thermal Sciences,
Vol.41, pp.625630, 2002a.
Smit, F.J., Meyer, J.P, Heat Transfer Coefficients during
Condensation of Zeotropic Refrigerant Mixture,
HCFC22/HCFC142b, ASME J. of Heat Transfer, Vol.124,
pp.11371146, 2002b.
Figure 6.1 Contours of Mixture Density for R22 at a) G = 100 kg/m2s and x=0.3 b) G = 100 kg/m2s and x=0.8
c) G = 200 kg/m2s and x=0.3 and d) Cross sectional view at G = 100 kg/m2s and x=0.3
e) Stratified Wavy regime obtained by Ewing et al. [1999] for air water mixture
6.2 At G = 200 kg/m2s and x=0.5 at a) Time step = 100 b) Time step = 200 and c) Time step = 500
d) Wavy Annular regime obtained by Ewing et al. [1999]
6.3 At a) G = 400 kg/m2s, x=0.3 b) G = 600 kg/m2s, x=0.3 and c) Cross section view for slug flow at
G = 400 kg/m2s, x=0.3 d) Slug flow regime obtained by Ewing et al. [1999]
6.4 a) at G = 400 kg/m2s, x=0.6 b) at G = 600 kg/m2s, x=0.6 and c) Cross section view for annular flow at
G = 400 kg/m2s, x=0.6 d) Annular regime obtained by Ewing et al. [1999]
3 d)
7th International Conference on Multiphase Flow
ICMF 2010, Tampa, FL USA, May 30June 4, 2010
Table 1 Flow Conditions and Flow patterns predicted using Thome et al.
[2003] Flow Regime Maps for R134a, R22 and R407C
Mass Quality Flow Regime Flow Regime Flow Regime
Flux R22 R134a R407C
100 0.3 SW SW SW
100 0.8 SW SW SW
200 0.3 SW SW SW
200 0.5 WavyAnnular WavyAnnular WavyAnnular
400 0.3 Intermediate Intermediate Intermediate
400 0.5 Slugannular Annular Slugannular
400 0.6 Annular Annular Annular
600 0.3 Intermediate Intermediate Intermediate
600 0.5 Slugannular Annular Slugannular
600 0.6 Annular Annular Annular
 Grdnnerud  Friedel
 LockhartMartinelli  Chisholm
  MUller Steinhagen and Heck  CFD Analysis
Experiment
02 04 06 08 1.0
Vapor Quality
7000
6000
5000
4000 0
3000 "
w .I /.:"r ... .
2000
1 0 b)
1000 
0.2 0.4 0.6
Vapor Quality
E
C)
t
Ld
0:
0~
a,
'1
2i
1.4
12
08
0.6
0.4 F
I X e)
0.2 0.3 0.4 0.5 0.6 0.7 0.
Vapor Quality
14
1.2
I a
1 *
nar _
,,0
0.4
.0 02
0.3 04 0.5
Vapor Quality
Figure 7 Comparison of CFD Data of Pressure Gradient with that of Experiment and Correlations for R22
at a) G= 200 b) 400 and c) 600 kg/m2s d) Deviation Graph e) Parity Graph
1600
E
@( 1200
800
400
0 400
SI
x ...^ ..I.
I H : l a)
Sa)
, I
o 0.8
o 0.4
ad)
2 i
0 0.4
Ji c )
__A
05 06
Vapor Quality
E
a
0)
ro
cD
(5
0)
0)
Qa
0)
tD
I)
LU
w
0 2 0.4 0.6
Vapor Quality
12000
E
1m
 10000
o 8000
0
6000
4000
2000
.
.... ..   g
..../  .
c)_
c
U
I A
06 0.7
(
[
0.4
0.7 0.8
1
0.8 1.1
06
