International Communications in Heat and Mass Transfer 88 (2017) 236–244

Contents lists available at ScienceDirect

International Communications in Heat and Mass Transfer

journal homepage: www.elsevier.com/locate/ichmt

Experimental investigation of condensation heat transfer of R600a/POE/ MARK

CuO nano-refrigerant in flattened tubes
Babak Ghorbania,⁎, M.A. Akhavan-Behabadib, Sasan Ebrahimia, Krishna Vijayaraghavana
a
Simon Fraser University, Mechatronic Systems Engineering Department, Surrey, BC, Canada
b
College of Engineering, University of Tehran, Tehran, Iran

A R T I C L E I N F O A B S T R A C T

Keywords: In this study, the effect of addition of copper oxide nanoparticles on condensing heat transfer coefficient of
Condensation heat transfer R600a refrigerant flowing in a flat tube condenser has been investigated experimentally. The test setup consists
CuO nanoparticles of a pump, condenser test, second condenser, evaporator, heaters, and flow meter. The validation of the study
Flat tube was done by comparing the obtained condensation heat transfer coefficients with different empirical correlations
R-600a
in the literature. Different fluids including pure R600a, R600a-oil with Polyester oil (POE) mass percentage of
Mass flux
1%, and three R600a-oil-nanoparticle mixtures with mass percentages of 0.5%, 1%, and 1.5% were studied
Vapor quality
experimentally. It was shown that adding nanoparticles will result in 4.1%, 8.11%, and 13.7% average increase
in condensing heat transfer coefficient with respect to the R600a-oil mixture. The greatest amount of increase
was reported for the weight fraction of 1.5%, where it was observed that the condensing heat transfer coefficient
for the mixture passing through the flattened tube is averagely 109.3% higher than its corresponding value for
the pure refrigerant flowing in the round tube with the same mass flux. It was also found that an increase in mass
flux resulted in an increase the heat transfer coefficient at all vapor qualities.

1. Introduction
A nano-fluid is defined as a heat transfer fluid produced by sus-
pending nano-scale (10− 9 m) materials in a host fluid [1–4]. The na-
noparticles typically have a higher thermal conductivity, resulting in
higher overall thermal conductivity of the fluid. If refrigerant is used as
the host fluid, the obtained solution is called nano-refrigerant. Owing to
the higher thermal conductivity, nano-refrigerants have higher heat
transfer coefficient at a given Nusselt number. Considering thermal
conductivity as an ability of the material for conducting or transmitting
heat, higher thermal conductivity of nanofluids leads to the idea of
employing them in refrigeration applications.
Recently, natural refrigerants such as R-717, R-744, and R-600a
have attracted attention as an alternative to CFCs, HCFCs, and HFCs
because of their safe operating condition, durability, and environmental
concerns like ozone depleting and global warming. R-600a has been
shown to be better for the environment [5,6] possessing superior en-
ergy performance [7,8] compared to other natural refrigerants. Hence
R-600a was used as the refrigerant in the current study.
Several studies in the literature have used CuO nanoparticles owing
to improvement in thermo-physical properties and their capability to
increase the energy efficiency and effectiveness of refrigeration and air
conditioning units [9]. CuO nanoparticles find wide application as they
can be easily stabilized into the polyester oil and are also commercially
affordable. Kedzierski [10] quantified the effect of CuO nanoparticles
on boiling heat transfer coefficient of R134a/Polyolester flowing on a
rough flat surface and demonstrated that the nanoparticle concentra-
tion plays a crucial role in heat transfer enhancement of the nano-re-
frigerants. Liu et al. [11] studied the thermal conductivity of nano-re-
frigerants containing CuO nanoparticles. Their study demonstrated that
CuO nano-refrigerants possess considerably higher thermal con-
ductivities in comparison with base liquids without solid particles.
Coumaressin and Palaniradja built an experimental apparatus to in-
vestigate the evaporative heat transfer coefficient of R134a/CuO mix-
ture. It was found that using CuO nanoparticles with mass concentra-
tions up to 1% increase the heat transfer to a great extent [12].
Akhavan-Behabadi et al. [13] carried out an experimental research to
investigate the effects of CuO nanoparticles on condensation heat
transfer coefficient of R600a refrigerant in a horizontal round tube.
Different nanoparticle mass fractions of 0.5%, 1.0%, and 1.5% were
used with the vapor qualities ranging from 10% to 80%. The study
reports a maximum heat transfer enhancement of 83% compared to the
pure refrigerant at a mass percentage of 1.5%.
Compactness of heat exchangers has recently become another

⁎
Corresponding author.
E-mail addresses: bghorban@sfu.ca (B. Ghorbani), akhavan@ut.ac.ir (M.A. Akhavan-Behabadi), sasane@sfu.ca (S. Ebrahimi), krishna@sfu.ca (K. Vijayaraghavan).

http://dx.doi.org/10.1016/j.icheatmasstransfer.2017.09.011

0735-1933/ © 2017 Published by Elsevier Ltd.

B. Ghorbani et al. International Communications in Heat and Mass Transfer 88 (2017) 236–244

Nomenclature μ Dynamic viscosity

λ Heat transfer coefficient
A Area ρ Density
c Heat capacity σ Surface tension
D Diameter
e Enthalpy Subscripts
Fr Froude number
g gravity B bottom
G Mass flux cw Cooling water
Ga Galileo number e Evaporator
h Heat transfer coefficient h Heater
H Tube height i Inlet
k Thermal conductivity l Liquid
L Tube length m Mean
ṁ Mass flow rate o Outlet
Nu Nusselt number ow Outer wall
q Heat flux p at constant pressure
Q Heat r Refrigerator
Re Reynolds number s Static
T Temperature sat Saturation
w Tube width t Total
W Electrical power T Top
We Weber number tc Test condenser
x Vapor quality v Vapor
X Martinelli parameter w Wall

Greek symbols

η Isolation efficiency

important issue, and different passive methods have been proposed to
increase the heat transfer coefficient of condensers; one of the methods
proposed by the literature is flattening the tube. An experimental study
on condensation heat transfer coefficient and pressure drop of R134a
and R410a inside the flattened tubes with different internal heights of
5.74, 4.15, 2.57, and 0.97 mm obtained by flattening the round tube
with initial diameter of 8.9 mm has been done by Wilson et al.
Flattening the tube resulted in both increase in pressure drop and en-
hancement in heat transfer coefficient [14]. Quiben et al. also in-
vestigated the effects of flattening the tube on boiling heat transfer

Fig. 1. Schematic of the experimental apparatus.

237
B. Ghorbani et al.
coefficient of R22 and R410a. Two different round tubes with the initial
diameters of 8 and 13 mm were flattened to an oblong shape tube with
internal heights of 2 and 3 mm, respectively. The heat flux was found to
have a negligible effect on pressure drop, with channel height sig-
nificantly influencing both heat transfer coefficient and pressure drop
[15,16]. Nasr et al. preformed an identical study with R-134 flowing in
flattened tubes with internal heights of 6.6, 5.5, 3.8, and 2.8 mm made
by flattening the round copper tube with initial diameter of 8.7 mm. It
was found that both pressure drop and heat transfer coefficient increase
with flattening the tube, introducing the tube with internal height of
5.5 mm as the most energy efficient case [17]. Kim et al. also studied
the condensation heat transfer coefficient and pressure drop of R410a
in three flattened stainless tubes. They showed that flow regime plays
an important role when studying the effects of flattening the tube, in-
dicating that decreasing the internal height of the tube results in heat
transfer enhancement at annular flow regime while the opposite is true
for the stratified flow [18]. Darzi et al. investigated the condensation
heat transfer coefficient and pressure drop of R600a in three flattened
copper tubes with inner heights of 6.7, 5.2, and 3.1 mm made by flat-
tening the round copper tube with initial diameter of 8.7 mm. They
concluded that the flattened tube with internal height of 5.2 mm has
the best overall performance [19].
This study uses an experimental setup identical to the one by
Akhavan-Behabadi et al. [13,17,20–23], and the condensation heat
transfer coefficient of CuO/oil/R600a mixture flowing inside a hor-
izontal flattened tube with internal height of 5.2 mm, is experimentally
investigated for the first time. For this purpose, the heat transfer coef-
ficient was computed for a wide range of mass fluxes and vapor qua-
lities, and then it was compared with heat transfer coefficient of the
pure refrigerant, and the baseline mixture (R600a/oil). Mass flux was in
the range of 110–372 kg/m2s, with the vapor quality between 0.06 and
0.78. Five fluids including pure R600a, R600a-oil with Polyester oil
mass percentage of 1%, and three R600a-oil-nanoparticle mixtures with
mass percentages of 0.5%, 1%, and 1.5% were studied experimentally.
2. The experimental setup
An experimental setup has been designed to inspect the condensa-
tion heat transfer coefficient and pressure drop within both round and
flattened tubes, as depicted in Fig. 1.
2.1. Description of the experimental units
The test condenser (#5) which is used for condensing the R-600a
refrigerant, is a counter-flow double-pipe heat exchanger with the re-
frigerant flowing in the inner tube and the cooling water with the
temperature between 15 °C and 18 °C on the outside; the flattened tube
with an internal height of 5.2 mm, a length of 105 cm, and the initial
diameter of 8.7 mm was used in the test condenser section. Studying the
impact of refrigerant vapor quality on the condensation heat transfer
coefficient was one of the purposes of the current work. As the ex-
perimented flattened tube length is only 105 cm, it would be impossible
to cover a wide range for refrigerant vapor quality by using only two
condensers. Hence, two heaters and an evaporator were utilized for
achieving desired range of vapor quality. A copper tube with the length
of 1 m was employed for each heater as well as the evaporator, with
heating elements uniformly twisted around the outside. Each element
can produce up to 3000 W of electrical power, and the watt meter
precision is 1%. The internal height of 5.2 mm has been found to be the
optimum value from the stand point of heat transfer and pressure drop
for a flattened tube with standard initial diameter of 8.7 mm [17]. The
temperature of the flattened tube outer surface is measured by a K-type
thermocouple at six different equally spaced points at the top and
bottom side of the tube. The thermocouple device can measure tem-
peratures as high as 1000 K with the accuracy of 0.1 K. The measured
temperatures can be read by the data logger which is connected to the
238
International Communications in Heat and Mass Transfer 88 (2017) 236–244
computer. The pressure is also measured at the inlet and outlet of the
test condenser, using a pressure gauge with the accuracy of 10 kPa. The
outer surfaces of all heat exchangers were insulated using glass wool.
While the glass wool significantly reduces heat transfer, the heat loss
cannot be completely eliminated and has been included in our calcu-
lations. A post condenser with 12 m of spiral tube was used in order to
fully condense the refrigerant, preventing multiphase flow at the pump
inlet. A one-way valve is utilized before the pumping section in order to
prevent back flow into the post condenser.
Pressure drops within the test section were measured by the Endress
Hauser PMD-75, calibrated by the company. The device can operate in
the range of 0 to 150 kPa with the precision of 0.075SPN percent,
where SPN is the pressure at the condenser. The fluid mass flow rate is
computed by the Danfoss MASS/2100/600 flowmeter; this device can
measure the mass flow rates less than 250 kg/h with the precision of
1%.
2.2. Experimental setup operation
As illustrated in Fig. 1, the refrigerant is pumped through the system
using the gear-pump (#1). The refrigerant then flows through the flow
meter (#2), followed by the heaters (#3). The heater is used to achieve
the varying vapor qualities as the vapor quality at the inlet of the test
section is affected by the amount of heat added by the heating elements
as well as the heat losses in the Sections 3 and 4. The multiphase flow
then passes the test condenser (#4), where data for computing the
condensing heat transfer coefficient is measured. Finally, the refrigerant
is totally condensed in the post condenser (#7) to be pumped again
through the gear-pump (#1).
In the current study, 192 experiments with different mass flux and
vapor quality values were conducted for six different cases at four
different flow rates and eight different vapor qualities for each. For all
experiments, the following parameters were measured:
1. Refrigerant mass flow rate
2. Cooling water mass flow rate at the test condenser
3. Refrigerant temperature at inlet and outlet of test condenser and
post condenser
4. Outer wall of the inner tube temperature at 12 points within the test
condenser section
5. Cooling water inlet and outlet temperatures at the test condenser
6. Pressure at inlet and outlet of the test condenser and the post con-
denser
7. Amount of heat added at the heaters and the evaporator section
8. Pressure drop through the test condenser
Using all of the mentioned parameters, the condensing heat transfer
coefficient within the flattened tube located at the test condenser was
computed. First, the round tube and pure refrigerant were utilized; in
the next step, the flattened tube and pure refrigerant were studied.
Afterwards, the parameters were measured for refrigerant-oil flow in
the flattened tube; finally, the heat transfer coefficient for the R600a-
oil- CuO nanofluids with mass concentrations of 0.5, 1.0, and 1.5%
were investigated experimentally. Note that adding nanoparticles di-
rectly to the refrigerant does not make sense as the R600a-CuO mixture
is not stable [24]. Hence, the nano particles were first dissolved in the
polyester oil with mass concentrations of 0.5, 1.0, and 1.5% using
homogenizer, and then the oil-CuO mixture was added to the re-
frigerant with weight fraction of 1%, as done by [24]. The range of
different operating conditions is mentioned in Table 1.
3. Heat transfer coefficient calculations
Saturation temperature, saturation enthalpy, and all thermo-phy-
sical properties of R600a were computed using EES software. The vapor
quality at the test condenser inlet can be computed by writing the
B. Ghorbani et al. International Communications in Heat and Mass Transfer 88 (2017) 236–244

Table 1 Where, TT and TB are the measured temperature at the upper and at the
The range of different operating conditions for R600a, R600a-oil, and R600a-oil-CuO. lower surfaces of the tube, respectively.
Fluids Values Units
2. The mean temperature of the outer wall of the tube is calculated by
Fluid mass flux 110 − 372 kg/m2s averaging the values of Towthrough the whole section
Average condensation temperature 36.2 − 45.6 °
C
6 k
Vapor quality at the inlet of test condenser 0.10 − 0.85 − ∑k = 1 Tow
Vapor quality at the outlet of test condenser 0.01 − 0.66 − Tow, m =
6 (6)
Average vapor quality at the test condenser 0.06 − 0.78 −
Cooling water mass flow rate 0.065 − 0.21 −
Cooling thermal flux 22.31 − 23.19 kW/m2 3. The rate of heat transfer within the test condenser is computed by
Cooling water temperature change 0.7 − 2.8 °
C
measuring the inlet and outlet cooling water temperatures, as well
as its mass flow rate
energy equation for a control volume including the heaters and the Qtċ = ṁ cw Cp cw (Tcw, o − Tcw, i ) (7)
evaporator as follows:
Qḣ ,1 + Qḣ ,2 + Qė = Qṫ = ṁ r [Cpr (Tsat , r − Tr , in ) + er |e, out − ef , r ] (1) 4. The radial heat flux is then calculated using the following equation
̇ and Qh,2
where, Qh,1 ̇ represent the effective heat input from the two
Qtċ
heaters and Qė is the heat input from the evaporator. The first term in qtċ =
πDL (8)
the right hand side of the Eq. 1 represents the change in the enthalpy of
the liquid refrigerant from the inlet of the first heater to the saturation
point; the second term is the enthalpy of the two-phase refrigerant at
the outlet of the evaporator or at the inlet of the test condenser, where Where D is the diameter for the round tube, or the hydraulic dia-
the last term corresponds to the enthalpy of the liquid refrigerant in meter for the flattened tube. The Cross sectional area of the flattened
saturation temperature. It may be noted that tube is needed for computing the hydraulic diameter which is calcu-
lated as:
er |e, out − ef , r = x e, out efg (2)
πH 2
To account for the heat loss, a factor, η, the fraction of the electrical A= + wH
4 (9)
power that is transferred to the refrigerant at the heater sections is
Where, w is the flattened tube width, which was 5.2 mm; the value
included. Hence,
of 7.29 mm was found for the hydraulic diameter.
Qṫ = η Wṫ (3)
5. The temperature drop through the flattened tube wall ΔTw is com-
Where, Wṫ is the total electric power shown by the watt meter. The
puted by the following equation:
value of η was experimentally determined to be 0.7. Combining the Eqs.
1 to 3, the vapor quality at the test condenser inlet is given by:
ΔTw =
qtċ Di Ln ( )Do
Di
1 ⎡ ηWṫ 2k w (10)
x tc, in = x e, out = − Cpr (Tsat , r − Tr , in ) ⎤
efg ⎢
⎣ ṁ r ⎥
⎦ (4)
Condensation heat transfer for each experiment is computed by
using the amount of heat absorbed by the cooling water and the tem- In which kw is the thermal conduction of the copper wall, and Do and
perature difference between the inner wall of the tube and refrigerant Di are outer and inner diameters of the tube.
according to the following procedure:
6. The mean temperature of the inner wall Tiw, m is assumed equal to
1. The local temperature of the tube outer wall at six axial points is the mean outer wall temperature, Tow, m.
computed by the following equation as depicted in Fig. 2 7. The saturation temperature of the vapor Ts corresponds to the mean
static pressure at the condenser which was assumed to be the
k TTk + TBk
Tow = average of inlet and outlet static pressure values.
2 (5)
8. The average condensation heat transfer coefficient is finally calcu-
lated using the following equation.

Fig. 2. Schematic view of thermocouples locations used for finding temperature in test condenser and cross section view of the flattened tube.

239
B. Ghorbani et al. International Communications in Heat and Mass Transfer 88 (2017) 236–244

qtċ annular flow pattern exists at the experiment conditions. However, ef-
h=
Ts − Tiw, m (11) fects of the gravity force should also be investigated [28]. As the flow
pattern is annular, which occurs at high velocities and high vapor
qualities, the shear forces are dominated within the fluid zone. Hence,
4. Results and discussion
the gravity force does not affect the flow significantly and it can be
assumed that a layer of saturated liquid phase is produced on the inner
The value of the condensation heat transfer coefficient for all 192
side of the tube with vapor phase flowing in the middle. Although this
cases were calculated using the procedure explained in the previous
phenomenon increases the pressure drop, it is highly desired from the
section. Effects of different parameters such as refrigerant mass flux,
heat transfer point of view [20]. With decreasing the vapor quality and
vapor quality, and flattening the tube on the condensation heat transfer
velocity to much lower values, which did not happen in the current
coefficient for six different cases mentioned in Table 2 were in-
study, the gravity forces play a major role and liquid phase flows on the
vestigated. Note that for the cases of nanoparticle-oil-refrigerant, the
lower side of the tube, while the vapor phase passes through the upper
CuO nanoparticles are solved within the oil first, then the solution is
side; this will reduce the heat transfer coefficient because the con-
Added to the R600a refrigerator.
densed liquid acts as a thermal resistance against the heat flux [29].

4.1. The flow regime

4.2. Study validation
The flow regime for test cases were also investigated in the current
Each case presented in Table 2 was experimentally studied at 4
study using two different methods. First, according to the Breber flow
different mass fluxes and the value of condensing heat transfer coeffi-
pattern Map shown in Fig. 3, almost all of the experimented cases lie in
cient was measured using flows with 8 various vapor qualities. In order
the annular and mist annular region except for the very low values of
to validate the experimental setup, obtained heat transfer coefficients
vapor quality which results in transition regime from wavy and strati-
for pure R-600a were compared with those predicted by previous stu-
fied to the annular and mist annular.
dies. There are so many analytical models for computing the con-
In the Breber Map, Xtt is the Martinelli parameter calculated as:
densation heat transfer coefficient within the round tubes in the lit-
0.1 0.5
1 − x 0.9 ⎡ μl ⎤ ⎡ ρl ⎤ erature. In the current study, the experimental results have been
Xtt = ⎡ ⎤
compared with five well-known correlations to see if the experiment
⎣ x ⎦ ⎢ ⎥ ⎢ ⎥
⎣ μ v ⎦ ⎣ ρv ⎦ (12)
errors lie within an acceptable range. Correlations used for the vali-
JG∗ presents the vapor dimensionless velocity which is computed by dation purpose are all mentioned in the Table 3, and the comparison of
the following equation: the empirical correlations against the current data is shown in Figs. 4–8.
Gx According to the presented figures, Shah correlation shows the
JG∗ = minimum deviation from the experimental results with the maximum
Dgρg (ρf − ρg ) (13) deviation of 13%; the same trend has been reported by other studies in
Secondly, the flow pattern was studied using the method presented the literature, introducing the Shah correlation as the most reliable tool
by Nitheanandan et al. [25] The proposed method uses the Froude and for computing the condensation heat transfer coefficient of mixtures
the Weber dimensionless numbers for predicting the flow regime. The [34–39]. However, other correlations predicted greater heat transfer
Froude number was computed using the equation presented by Dobson coefficients in comparison with the current data. Many factors can
[26]: cause deviation of the experimental results from the empirical corre-
lations. For example, impurities affect the rate of heat transfer to a high
1.5
1 + 1.09Xtt0.039 ⎞ degree. Besides, the transport properties of the refrigerants presented
Fr = 0.025Rel1.59 ⎜⎛ ⎟ Ga−0.5 for Re ≤ 1250
⎝ Xtt ⎠ (14.1) by the various references vary from 20 to 40%. Furthermore, one must
consider the error due to inadequate instrumentation, faulty instru-
1.5
1 + 1.09Xtt0.039 ⎞ ments and human errors [40]. In view of the many sources of error
Fr = 1.26Rel1.04 ⎛⎜ ⎟ Ga−0.5 for Re > 1250
⎝ Xtt ⎠ (14.2) which was discussed, a good agreement between measurements and the
empirical correlations exists, and the maximum mean deviation was
Where, found to be about 25%.
ρl (ρl − ρg ) gD3
Ga = 4.3. Heat transfer enhancement
μl2 (15)

GD (1 − x ) Experiments were conducted for pure refrigerant, refrigerant-oil

FRel =
μl (16) mixture with weight fraction of 1% and refrigerant based nanofluids

The weber number was determined using the equation provided by

Table 2
Soliman et al. [27]: Properties of different experimented cases.
2 0.3
μ
We = 2.45Re v0.64 ⎛⎜ v ⎞⎟ (1 + 1.09Xtt0.039)−0.4 for Re ≤ 1250
Case Fluid Tube type Nanoparticle Oil/oil-nanoparticle
concentration in oil concentration in
⎝ Dσ ⎠
ρv (17.1)
refrigerant (%)
0.3 2 0.084
μ2 ⎡ μ ρ ⎤ 1 R600a Round tube 0.0 0.0
We = 0.85ReV0.79 ⎜⎛ V ⎟⎞ (1 + 1.09Xtt0.039)−0.4 ⎢ ⎜⎛ v ⎟⎞ ⎜⎛ L ⎟⎞ ⎥ Xtt 0.157 for 2 R600a Flattened tube 0.0 0.0
⎝ ρV Dσ ⎠ ⎣ ⎝ μL ⎠ ⎝ ρV ⎠
⎦ 3 R600a- Flattened tube 0.0 1.0
Re > 1250 (17.2) oil
4 R600a- Flattened tube 0.5 1.0
It was found that for all cases We < 40 and Fr > 7, indicating that oil-CuO
the flow regime was annular for the experiments, as discussed earlier. 5 R600a- Flattened tube 1.0 1.0
oil-CuO
The same result was achieved using the Breber pattern flow map.
6 R600a- Flattened tube 1.5 1.0
Besides, the We number did not exceed the value of 40 which means oil-CuO
that the flow regime is not even mist annular for any case and pure

240
B. Ghorbani et al. International Communications in Heat and Mass Transfer 88 (2017) 236–244

Fig. 4. Comparison of the Traviss correlation against the present data [30].

Fig. 3. Breber flow pattern map.

Table 3
The empirical correlations used for the study validation.

Number Flow pattern Model/correlation

correlation

1 Traviss et al. Ref0.9 Pr

[30] Nu = . F (Xtt )
F2
F(Xtt) = 0.15(Xtt− 1 + 2.85Xtt− 0.476)
F2 = 0.707 Pr Ree0.5 Ref ≤ 50
F2 = 5 Pr + 5Ln(1 + Pr(0.096Ref0.585 − 1))
50 < Ref ≤ 1125
F2 = 5 Pr + 5Ln(1 + 5Pr) + 2.5 ln(0.00313Ref0.812)
Ref > 1125
2 Cavalini et al. 1
0.8
[31] Nu = 0.05Prf 3Reeq
1
2
⎛ μg ⎞ ⎛ ρf ⎞
Reeq = Reg ⎜ ⎟ ⎜ ⎟ + Ref
μ ρ
⎝ f ⎠⎝ g ⎠ Fig. 5. Comparison of the Cavallini correlation against the present data [31].
3 Dobson [26] Annular flow; Re > 1250, Fr > 20
2.22
Nu = 0.023Pr 0.4Rel0.8 ⎡1 + 0.889 ⎤
⎢ Xtt ⎥
⎣ ⎦
Fr is computed using the equation ….
4 Shah et al. 3.8Xtt0.76 (1 − Xtt )0.04 ⎞
[32] h = hf ⎜⎛ (1 − Xtt )0.8 + ⎟

⎝ Pr 0.38 ⎠
k
hf = 0.023 Ref0.8 Pr 0.4
D
5 Jung et al. 0.81 0.29
Nu = 22.4hf ⎛1 +
2 ⎞ ⎛ q ⎞
[33] ⎜ ⎟
⎜ ⎟
⎝ Xtt ⎠ ⎝ hfg G ⎠
hf = 0.023Ref0.8Pr0.4

with three different nanoparticle mass fractions of 0.5, 1, and 1.5%.

Finally, 192 experiments were done and the obtained values of con-
densing heat transfer coefficients were computed as presented in Fig. 9.
For the case of pure refrigerant in the flattened tube, it was observed
that heat transfer coefficient decreases with decreasing the vapor
quality at all values of mass flux. The same trend is true for the pure
refrigerant flowing in the round tube. This makes sense because the
liquid layer on the inner side of the tube is thinner at higher vapor
qualities which results in lower thermal resistance against the heat flux.
Before studying the effects of adding CuO nanoparticles to the re- Fig. 6. Comparison of the Dobson correlation against the present data [26].
frigerant, the behavior of the R600a-oil solution with mass

241
B. Ghorbani et al.
Fig. 7. Comparison of the Shah correlation against the present data [32].
Fig. 8. Comparison of the Jung correlation against the present data [33].
concentration of 1% was investigated. According to Fig. 9, adding oil to
the refrigerant increases the heat transfer coefficient at lower vapor
qualities; however, the opposite trend is true at higher vapor qualities
and there is a decrease in heat transfer coefficient compared to the pure
refrigerant for vapor concentrations more than 0.6. This behavior can
be explained as follows. First, the value of surface tension for the
mixture is higher than the pure refrigerant, increasing the wettability
and changing the flow pattern to the annular. Besides, the thermal
conductivity of the mixture, is higher than the thermal conductivity of
the pure refrigerant which helps the mechanism of heat transfer to
occur [13]. On the other hand, there is an increase in oil concentration
in the liquid phase by increasing the vapor quality, as oil remains in the
liquid phase. Consequently, increasing the viscosity, the value of the
Reynolds number is reduced and this will result in lower heat transfer
coefficients.
From the comparison between the heat transfer coefficient of re-
frigerant-oil-CuO in Fig. 9 with the corresponding values of refrigerant-
oil in Table 4, it is obvious that the rate of heat transfer increases by
adding the nanoparticles to the refrigerant. The amount of rise in heat
transfer coefficient is different at various mass fluxes and mass con-
centrations. The greatest improvement was occurred for nanofluid with
242
International Communications in Heat and Mass Transfer 88 (2017) 236–244
mass percentage of 1.5%, where the heat transfer coefficient was found
to be 13.7% higher than the refrigerant-oil mixture. Note that the
percent increase in heat transfer is computed by averaging through the
values of increase in heat transfer for all vapor qualities. This means
that achieving even higher values of increase in heat transfer coefficient
is possible at some vapor qualities. The average values of heat transfer
enhancement for all cases are mentioned in Table 4. This rise in rate of
heat transfer can be accounted for by different phenomenon happening
by adding the nanoparticles and flattening the tube. First, disturbance
of nanoparticles enhances the heat transfer due to the reduction of
boundary layer thickness as the boundary layer acts as a thermal re-
sistant against the heat flux [41]. Second, the liquid molecules can be
absorbed by the nanoparticles which leads to the formation of the
molecular adsorption layer on their surface [42]. The rate of heat
transfer in the body of the fluid is then increased by the formation of
molecular layer and nanoparticles movements.
Likewise, one of the other dominant parameters affecting the
thermal conductivity of the mixture and the heat transfer coefficient
within the fluid media is Brownian motion due to addition of nano-
particles. Brownian motion can increase the thermal conductivity of the
nanofluid by an insignificant value, primarily by increasing the
“random walk” motion instead of motion only through diffusion. This
improving effect only occurs at lower concentrations and after reaching
a limit in weight fraction of nanoparticles, no more increase in heat
transfer coefficient is achievable [43]. Besides, after more increasing
the nanoparticles mass concentration, while the cluster formation leads
to an increase in conductivity, the maximum rise remains well below
the maximum value allowed by Maxwell's upper limit [44]. Further-
more, the movement velocity of larger particles is much lower than that
of smaller ones, which reduces the probability of collision [45].
Therefore, the condensing heat transfer coefficient within the fluid with
maximum weight fraction of 1.5 for CuO nanoparticles was investigated
at this study.
Other reasons can also be accounted for the improvement of heat
transfer by suspending nanoparticles in heating or cooling fluids. The
suspended nanoparticles increase the surface area, the heat capacity of
the fluid, and the effective thermal conductivity of the mixture; they
also intensify the flow passage surface and mixing fluctuations, as well
as the turbulence of the fluid [46].
5. Conclusion
Experiments were conducted to investigate the effects of CuO na-
noparticles on the condensation heat transfer coefficient of R-600a in-
side a horizontal plain tube. For the feasibility of nanoparticle disper-
sion into the refrigerant, the CuO nanoparticles were fist solved into
polyester oil, and then the mixture was added to the refrigerant. Many
predictive methods are available for two phase flow patterns and heat
transfer coefficient in the horizontal tubes which were compared with
the current experimental data and it was found that Shah correlation
has the minimum deviation from the experimental results. Annular flow
pattern was found to be the dominant flow pattern in this study. The
results showed that the heat transfer coefficient increases with in-
creasing mass velocity and vapor quality. However, for the vapor
qualities higher than approximately 0.6, the heat transfer deteriorated
due to the rise in local oil concentration. Adding nanoparticles can
improve condensation heat transfer coefficient up to 13.70% compared
to oil/refrigerant mixture occurred at the mass velocity of 212 kg/m2s
for the highest mass fraction of nanoparticles of this study, i.e. 1.5%.
Some reasons were presented for the heat transfer enhancement during
nanofluid condensation such as disturbance of nanoparticles, reduction
in boundary layer thickness, formation of molecular adsorption layer,
the Brownian motion effects, increase in the surface area as well as in
the heat capacity of the fluid, and rise in the effective thermal con-
ductivity of the mixture. For the 0.5% mass fraction of nanoparticles at
110, 212, 295 and 372 kg/m2s mass flux, the average heat transfer
B. Ghorbani et al. International Communications in Heat and Mass Transfer 88 (2017) 236–244

Fig. 9. Estimated values of condensing heat transfer coefficient for different values of mass flux.

conductivity data for water-based nanofluids, Int. J. Therm. Sci. 48 (2) (2009)
Table 4
363–371.
Increase in heat transfer coefficient with respect to the refrigerant-oil case for different
[3] W. Duangthongsuk, S. Wongwises, Measurement of temperature-dependent thermal
mass fluxes.
conductivity and viscosity of TiO2-water nanofluids, Exp. Thermal Fluid Sci. 33 (4)
(2009) 706–714.
Case Percent of increase in heat transfer coefficient [4] C.Y. Tso, S.C. Fu, C.Y.H. Chao, A semi-Analytical model for the thermal con-
number ductivity of nanofluids and determination of the nanolayer thickness, Int. J. Heat
2 2 2 2
G = 110 kg/m s G = 212 kg/m s G = 295 kg/m s G = 372 kg/m s Mass Transf. 70 (2014) 202–214.
[5] F.S. Rowland, M.J. Molina, Stratospheric for chlorofluoromethanes: chlorine atom-
4 0.96 4.10 1.61 3.53 catalyzed destruction of ozone, Nature 249 (1974) 810–812.
5 7.1 8.11 6.37 6.90 [6] M.J. Kurylo, The chemistry of stratospheric ozone: its response to natural and an-
6 11.67 13.70 10.77 11.57 thropogenic influences, Int. J. Refrig. 13 (2) (1990) 62–72.
[7] H.S. Lee, J.I. Yoon, J.D. Kim, P. Bansal, Evaporating heat transfer and pressure drop
of hydrocarbon refrigerants in 9.52 and 12.70 mm smooth tube, Int. J. Heat Mass
Transf. 48 (12) (2005) 2351–2359.
enhancement up to 0.96%, 4.10%, 1.61% and 3.53% was achieved. [8] H.-S. Lee, J.-I. Yoon, J.-D. Kim, P.K. Bansal, Characteristics of condensing and
Besides, for 1% mass fraction of nanoparticles the average heat transfer evaporating heat transfer using hydrocarbon refrigerants, Appl. Therm. Eng. 26
enhancement up to 7.1%, 8.11%, 6.37% and 6.90% was obtained. (10) (2006) 1054–1062.
[9] R. Saidur, S.N. Kazi, M.S. Hossain, M.M. Rahman, H.A. Mohammed, A review on the
Finally, the most average heat transfer enhancement occurred at the performance of nanoparticles suspended with refrigerants and lubricating oils in
1.5% mass fraction of nanoparticles, which are up to 11.67%, 13.70%, refrigeration systems, Renew. Sust. Energ. Rev. 15 (1) (Jan. 2011) 310–323.
10.77% and 11.57% for the mass velocities of 110, 212, 295 and [10] M.A. Kedzierski, Effect of CuO nanoparticle concentration on R134a/Lubricant
pool-boiling heat transfer, J. Heat Transf. 131 (4) (2009) 43205.
372 kg/m2s, respectively. Therefore, one can use oil/CuO/r600a and [11] M.-S. Liu, M.C.-C. Lin, I.-T. Huang, C.-C. Wang, Enhancement of Thermal
flattened tube instead of round tube to increase heat transfer which Conductivity with CuO for Nanofluids, Chem. Eng. Technol. 29 (1) (Jan. 2006)
consequently increases the overall efficiency of any condensing system. 72–77.
[12] T. Coumaressin, K. Palaniradja, Performance Analysis of a Refrigeration System
Using Nano Fluid, Int. J. Adv. Mech. Eng. 4 (4) (2014) 459–470.
References [13] M.A. Akhavan-Behabadi, M. Nasr, S. Baqeri, Experimental investigation of flow
boiling heat transfer of R-600a/oil/CuO in a plain horizontal tube, Exp. Thermal
[1] Y.J. Hwang, et al., Investigation on characteristics of thermal conductivity en- Fluid Sci. 58 (2014) 105–111.
hancement of nanofluids, Curr. Appl. Phys. 6 (6) (2006) 1068–1071. [14] M.J. Wilson, T.A. Newell, J.C. Chato, C.A. Infante Ferreira, Refrigerant charge,
[2] H.A. Mintsa, G. Roy, C.T. Nguyen, D. Doucet, New temperature dependent thermal pressure drop, and condensation heat transfer in flattened tubes, Int. J. Refrig. 26

243
B. Ghorbani et al.
(4) (2003) 442–451.
[15] J.M. Quibén, L. Cheng, R.J. da Silva Lima, J.R. Thome, Flow boiling in horizontal
flattened tubes: Part II – Flow boiling heat transfer results and model, Int. J. Heat
Mass Transf. 52 (15) (2009) 3645–3653.
[16] J.M. Quibén, L. Cheng, R.J. da Silva Lima, J.R. Thome, Flow boiling in horizontal
flattened tubes: Part I – Two-phase frictional pressure drop results and model, Int. J.
Heat Mass Transf. 52 (15) (2009) 3634–3644.
[17] M. Nasr, M.A. Akhavan-Behabadi, S.E. Marashi, Performance evaluation of flat-
tened tube in boiling heat transfer enhancement and its effect on pressure drop, Int.
Commun. Heat Mass Transfer 37 (4) (2010) 430–436.
[18] N.H. Kim, E.J. Lee, H.W. Byun, Condensation heat transfer and pressure drop in
flattened smooth tubes having different aspect ratios, Exp. Thermal Fluid Sci. 46
(2013) 245–253.
[19] M. Darzi, M.A. Akhavan-Behabadi, M.K. Sadoughi, P. Razi, Experimental study of
horizontal flattened tubes performance on condensation of R600a vapor, Int.
Commun. Heat Mass Transfer 62 (2015) 18–25.
[20] M. Saeedinia, M.A. Akhavan-Behabadi, M. Nasr, Experimental study on heat
transfer and pressure drop of nanofluid flow in a horizontal coiled wire inserted
tube under constant heat flux, Exp. Thermal Fluid Sci. 36 (2012) 158–168.
[21] M. Nasr, M.A. Akhavan-Behabadi, M.R. Momenifar, P. Hanafizadeh, Heat transfer
characteristic of R-600a during flow boiling inside horizontal plain tube, Int.
Commun. Heat Mass Transfer 66 (2015) 93–99.
[22] M.R. Momenifar, M.A. Akhavan-Behabadi, M. Nasr, P. Hanafizadeh, Effect of lu-
bricating oil on flow boiling characteristics of R-600a/oil inside a horizontal smooth
tube, Appl. Therm. Eng. 91 (2015) 62–72.
[23] M. Nasr, M.A. Akhavan-Behabadi, S.E. Marashi, The Effect of Tube Flattening on
Flow Boiling Heat Transfer Enhancement, Heat Transfer Eng. 32 (6) (May 2011)
467–475.
[24] K. Henderson, Y.-G. Park, L. Liu, A.M. Jacobi, Flow-boiling heat transfer of R-134a-
based nanofluids in a horizontal tube, Int. J. Heat Mass Transf. 53 (5) (2010)
944–951.
[25] T. Nitheanandan, H.M. Soliman, R.E. Chant, 3328 – a proposed approach for cor-
relating heat transfer during condensation inside tubes, Conference Proceeding by
ASHRAE, 1990.
[26] M.K. Dobson, Heat transfer and flow regimes during condensation in horizontal
tubes, PhD Thesis, Mechanical and Industrial Engineering, University of Illinois at
Urbana-Champaign, Urbana-Champaign, IL, 1994.
[27] H.M. Soliman, On the annular-to-wavy flow pattern transition during condensation
inside horizontal tubes, Can. J. Chem. Eng. 60 (4) (Aug. 1982) 475–481.
[28] S.G. Kandlikar, P. Balasubramanian, An Experimental Study on the Effect of
Gravitational Orientation on Flow Boiling of Water in 1054 Ã 197 μm Parallel
Minichannels, J. Heat Transf. 127 (8) (August 2005) 820–829.
[29] R.M. Macgillivray, Gravity and Gas Density Effects on Annular Flow Average Film,
M.Sc. Thesis, University of Saskatchewan, Sep 2004.
244
International Communications in Heat and Mass Transfer 88 (2017) 236–244
[30] D.P. Traviss, A.G. Baron, W.M. Rohsenov, Forced-convection condensation inside
tube, MIT Report DSR-72591, 1971.
[31] A. Cavallini, J.R. Smith, R. Zecchin, A dimensionless correlation for heat transfer in
forced convection condensation, Sixth Int. Heat Transf. Conf, 2 1974, pp. 309–313.
[32] M.M. Shah, An improved and extended general correlation for heat transfer during
condensation in plain tubes, HVAC & R Res. 15 (5) (Sep. 2009) 889–913.
[33] D. Jung, K.H. Song, Y. Cho, S.J. Kim, Flow condensation heat transfer coefficients of
pure refrigerants, Int. J. Refrig. 26 (1) (2003) 4–11.
[34] D. Mishkinis, J.M. Ochterbeck, Analysis of tube side condensation in microgravity
and earth normal gravity, Int. Semin. “Heat Pipes, Heat Pums, Refrig. Minsk.
Belarus”, Sept. 8–12, 2003.
[35] Y. Yang, L. Jia, Experimental investigation on heat transfer coefficient during up-
ward flow condensation of R410A in vertical smooth tubes, J. Therm. Sci. 24 (2)
(2015) 155–163.
[36] R. Sánta, The analysis of two-phase condensation heat transfer models based on the
comparison of the boundary condition, Acta Polytech. Hungarica 9 (6) (2012)
167–180.
[37] S. Thomas, C. Balakrishna Panicker Sobhan, XA review of experimental investiga-
tions on thermal phenomena in nanofluids, Nanoscale Res. Lett. 6 (1) (May 2011)
377.
[38] M. Derby, H.J. Lee, Y. Peles, M.K. Jensen, Condensation heat transfer in square,
triangular, and semi-circular mini-channels, Int. J. Heat Mass Transf. 55 (1–3)
(2012) 187–197.
[39] J.D. Berrichon, H. Louahlia-Gualous, P. Bandelier, N. Bariteau, Experimental and
theoretical investigations on condensation heat transfer at very low pressure to
improve power plant efficiency, Energy Convers. Manag. 87 (2014) 539–551.
[40] M.M. Shah, A general correlation for heat transfer during film condensation inside
pipes, Int. J. Heat Mass Transf. 22 (4) (1979) 547–556.
[41] H. Peng, G. Ding, W. Jiang, H. Hu, Y. Gao, Heat transfer characteristics of re-
frigerant-based nanofluid flow boiling inside a horizontal smooth tube, Int. J.
Refrig. 32 (6) (2009) 1259–1270.
[42] C.-J. Yu, A. Richter, A. Datta, M. Durbin, P. Dutta, Molecular layering in a liquid on
a solid substrate: an X-ray reflectivity study, Phys. B Condens. Matter 283 (1)
(2000) 27–31.
[43] A. Gupta, R. Kumar, Role of Brownian motion on the thermal conductivity en-
hancement of nanofluids, Appl. Phys. Lett. 91 (22) (2007) 2010–2013.
[44] R.K. Shukla, V.K. Dhir, Effect of brownian motion on thermal conductivity of na-
nofluids, J. Heat Transf. 130 (4) (2008) 42406.
[45] T.-P. Teng, Y.-H. Hung, T.-C. Teng, H.-E. Mo, H.-G. Hsu, The effect of alumina/
water nanofluid particle size on thermal conductivity, Appl. Therm. Eng. 30 (14)
(2010) 2213–2218.
[46] Y. Xuan, Q. Li, Heat transfer enhancement of nanofluids, Int. J. Heat Fluid Flow 21
(1) (2000) 58–64.