International Journal of Heat and Mass Transfer 104 (2017) 11351144

Contents lists available at ScienceDirect

International Journal of Heat and Mass Transfer

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

Experimental study on flow condensation of mixture in a hydrophobic

microchannel
Chengbin Zhang a, Chaoqun Shen b, Yongping Chen a,b,
a
b

Key Laboratory of Energy Thermal Conversion and Control of Ministry of Education, School of Energy and Environment, Southeast University, Nanjing, Jiangsu 210096, PR China
School of Hydraulic, Energy and Power Engineering, Yangzhou University, Yangzhou, Jiangsu 225127, PR China

a r t i c l e

i n f o

Article history:
Received 11 June 2016
Received in revised form 12 September
2016
Accepted 12 September 2016

Keywords:
Condensation
Microchannel
Mixture
Flow pattern

a b s t r a c t
The condensation flow patterns of ethanolwater mixture in a wide range of ethanol mass concentration
inside a hydrophobic microchannel is experimentally studied via a high speed imaging system. The
effects of channel surface wettability and ethanol concentration on flow condensation are compared
and discussed. The experimental results indicate that the surface hydrophobic modification and ethanol
concentration play a significant role in the flow condensation of mixture in a microchannel. The droplet
condensation appears almost the whole two-phase flow region when the water steam is the main component, in which the droplet flow, droplet-streak flow, droplet-annular flow, droplet-injection flow and
droplet-slug/bubble flow occur sequentially in a hydrophobic microchannel. With increasing ethanol
concentration, the droplet-slug/bubble flow after droplet-injection flow disappears and is replaced by
the pure slug/bubble flow. When the ethanol vapor is the main component, the droplet condensation
almost disappears, and the annular-streak flow, annular flow, injection flow and slug/bubble flow appear
sequentially along the flow direction. Both an increase in vapor Reynolds number and a decrease in ethanol concentration cause the injection location move toward the channel outlet. In addition, the surface
hydrophobic modification introduces the droplet condensation, which is beneficial for the flow condensation heat transfer.
2016 Elsevier Ltd. All rights reserved.

1. Introduction
Miniaturization is a notable feature of the modern devices due
to the rapid development of microfluidic science and technology.
As a typical physical process, the microscale flow condensation
widely exists in miniaturized systems, including micro heat pipes,
micro fuel cells, and micro coolers. The flow condensation in a
microchannel involves the gasliquid two-phase flow, phase
change heat transfer, flow instability as well as the coupled
heat and mass transfer. The flow regimes are the result of
comprehensive effects of channel scale, surface wettability and
fluid properties. Either the hydrophobically-modified surface or
the introduction of binary miscible mixture is beneficial to the
enhancement of flow condensation heat transfer. However, it is
not properly known how the mixture concentration affects the
flow condensation regime in a hydrophobic microchannel. In this
particular situation, it is of considerable significance to understand
Corresponding author at: Key Laboratory of Energy Thermal Conversion and
Control of Ministry of Education, School of Energy and Environment, Southeast
University, Nanjing, Jiangsu 210096, PR China.
E-mail address: ypchen@seu.edu.cn (Y. Chen).
http://dx.doi.org/10.1016/j.ijheatmasstransfer.2016.09.029
0017-9310/ 2016 Elsevier Ltd. All rights reserved.

the two-phase flow condensation of mixture in a hydrophobic

microchannel.
In a microchannel, the small dimension highlights the importance of surface forces, causing the flow condensation regime to
be significantly different from the conventional channels [1]. The
surface tension and shear force, rather than the gravity and buoyancy, are dominant the two-phase flow condensation regimes
under microscales [26]. The flow patterns along the flow stream
in microchannels are usually the droplet flow, annular flow, injection flow and slug/bubble flow, which are related to the mass flux
[711], channel size [7], steam quality [8,9] and cooling rates
[10,11]. Owing to the microscale effect, the injection flow is the
characteristic condensation flow pattern in microchannels [713].
Several special flow patterns are observed in microchannels,
including the droplet-annular flow in a wide rectangular silicon
microchannel [14], and droplet-injection flow in a hydrophobic
microchannel [15]. In addition, the hydrophobically-modified
surface could make the droplet condensation occupy the whole
gasliquid two-phase flow region in a microchannel [15,16], which
is beneficial to the enhancement of condensation heat transfer.
It is generally accepted that the condensation heat transfer can
be improved by adding a miscible additive to the vapor. By this

1136

C. Zhang et al. / International Journal of Heat and Mass Transfer 104 (2017) 11351144

Nomenclature
Acr
D
L
l
ml
n
Rev
t
W
X

cross-sectional area of channel, m2

hydraulic diameter, m
effective length of microchannel, m
distance between the channel inlet and view point, m
flow rate, kg s1
channel number
vapor Reynolds number
temperature, K
ethanol weight concentration, %
dimensionless location

mean, even on a wettable surface, the condensate does not always

occur as a liquid film. In the binary vapor condensation, owing to
the presence of surface tension gradient, any disturbance on the
vapor-condensate interface causes the so-called pseudo-droplet
condensation [17,18]. The previous experimental investigation
indicates that the condensation performance of a vapor mixture
is dependent on the mixture composition [1922], vaporto-surface temperature difference and vapor velocity [23]; the
condensation modes on a solid surface mainly includes the film
condensation, transition state and dropwise condensation [24].
Considering that the mixture and microscale are two approaches
to enhance condensation heat transfer performance, it is a great
potential for practical applications to combine the advantages of
microscale condensation and vapor mixture condensation.
Recently, Jiang et al. [25] have taken a visualization experiment
to study the condensation flow regimes of ethanolwater mixtures
in a trapezoidal microchannel. They have observed several special
flow patterns under different ethanol concentrations which are not
observed for the pure vapor condensation in microchannels.
Despite several experimental efforts have been conducted to
explore the vaporliquid two-phase flow behaviors of flow condensation in a microchannel [716], conventional condensation
of a mixture on a solid surface [1724] and even vapor mixture
condensation in a microchannel [25], few studies have focused
on flow condensation of mixture in a hydrophobic microchannel.
Especially, it is unclear what the relation is between the condensation flow patterns and the mixture concentration in a hydrophobic
microchannel. To gain a further insight into the microscale flow
condensation regimes, an experimental investigation is conducted
to explore the flow condensation in a hydrophobic microchannel
by using ethanolwater mixtures in a wide range of ethanol concentration. The observed flow patterns of vapor mixture condensation in a hydrophobic microchannel are presented and analyzed. In
addition, the effects of channel surface wettability and ethanol
mass concentration on condensation flow patterns are compared
and discussed.

2. Descriptions of the experiment

2.1. Experimental setup
In order to elucidate the flow condensation regimes of a
mixture in a hydrophobic microchannel, the ethanolwater mixtures in a wide range of ethanol mass concentration are applied.
The experimental setup for the observation of condensation flow
patterns in a microchannel is designed and assembled. Fig. 1 shows
the experimental setup of flow condensation in a microchannel,
which is mainly composed of the test section, pipeline system,
thermostatic water bath and high speed imaging system.

Greek symbols
vapor mixture viscosity, m2 s1
vapor mixture density, kg m3

tv
qv

Subscripts
p
injection flow
in
channel inlet

Voltage regulator
Voltage
stabilizer
Power meter

Microscope
and CCD

PC

Filter

Syringe pump

Chip

Heater
Cooler
Data acquisition

IR camera
Flow meter

T, P

PC

Container
Constant temperature
coolin water bath

Vapor and condensate

Cooling water

Thermocouple

Fig. 1. Schematic of experimental setup.

The test section, which is fixed on a Teflon base, contains the

experimental chip and aluminum cooler. The experimental chip
is made of 10 parallel microchannels with a rectangular crosssection (width: 300 lm, height: 100 lm, effective length
L = 56 mm), as shown in Fig. 2. The microchannel is fabricated by
the use of deep reactive ion etching (DRIE) on h1 0 0i silicon wafers.
Pyrex glass is anodically bonded to the silicon wafer to form the
parallel microchannels. The microchannel wall is treated to be
hydrophobic by using thin Au film and the equilibrium contact
angle of water is 96 (see W = 0% in Fig. 3). The contact angles of
ethanolwater
mixtures
with
different
ethanol
mass
concentrations on the channel wall is measured and shown in
Fig. 3. The contact angle of ethanolwater mixtures is much smaller than 96. The circulation fluid supplied by a thermostatic
waterbath is used to control the temperature of aluminum cooler
for the regulation of flow condensation in a microchannel.
Ethanolwater mixtures with different ethanol concentrations
(W = 2%, 10%, 20%, 40%, 60% and 80%) in the syringe are injected
by syringe pump (LSP01-1BH with the flow accuracy of 0.5%.) into
a heater where the saturated vapor is generated. In the experiment,
a long vertical upward minitube with electric heating is utilized as
a heater to generate the vapor. The heat power as measured by
power meter is changed by adjusting the voltage according to
the flow rate of ethanolwater mixtures. For all the experimental
cases, no temperature jump is observed for the wall temperature
at the downstream of vertical upward minitube. This implies that
the generated vapor is saturated at the microchannel inlet. Fig. 4
presents the relationship between the saturated temperature and
pressure under different mass concentrations of ethanol. After flow

C. Zhang et al. / International Journal of Heat and Mass Transfer 104 (2017) 11351144

Reservoir

Microchannels
9.3 mm 16 mm

Hole
56 mm
80 mm
(a) Photo of experimental chip

Pyrex glass
500 m

300 m
100 m
Microchannel

450 m

Silicon
(b) Cross-section of a microchannel

Pyrex glass

Microchannel

Silicon

through the filter, the vapor condenses in microchannel chip and

then is gathered in the container. The latent heat of vapor is
released to the cooling water supplied by the thermostatic water
bath. The aluminum cooler contacts with the microchannel chip
by the thermally conductive silicone. During the experiment, the
cooling water temperature is kept at 16 C.
The temperature at the inlet and outlet of both the aluminum
cooler and experimental chip are measured by the T-type thermocouples, and the temperature signal is acquired by the data acquisition (Agilent 34970A). The microchannel surface temperature is
measured via a noncontact measurement method and is recorded
by an infrared radiation imaging device (NEC, TH9260 with the
accuracy of 2%) which measure the surface temperature of
microchannel chip through five small holes manufactured on the
aluminum cooler. The infrared radiation imaging device should
focus on every hole independently in the temperature measurement of chip wall for an experimental case. The measurement of
wall temperature is verified via single phase liquid flow in
microchannel chip without the coolant flowing through aluminum
cooler. A flowmeter with accuracy of 0.1 mL/s is applied to measure the volumetric flow rate of cooling water. The high speed
imaging system is composed of a microscope and a digital camera
incorporating a charge-coupled device (CCD) detector (Photron
SA4), which is employed to record the condensation flow patterns
in the experiment. The photographing frequency of the camera is
set to be 30005000 frame per second in the experiment.
2.2. Data reduction

Inlet tube
Plenum

Teflon base

Cooler

(c) Plenum connecting the inlet tube and the microchannel chip

75

Silicon wafer

60
45
30

Rev

15
0%

20%

40%
W

60%

80%

Fig. 3. Contact angle of water and ethanolwater mixtures on the channel wall.

W = 2%
W = 10%
W = 20%

390
380

W = 40%
W = 60%
W = 80%

360
350
100

120

ml D
ntv qv Acr

where D, ml, n and Acr are the hydraulic diameter, flow rate, channel
number and cross-sectional area, respectively. tv is the vapor
mixture viscosity and can be obtained by the method introduced
by Wilke [26]. qv is the vapor mixture density which is described
by the ideal gas law and varies with the mass fractions of
component. [28]
3. Flow pattern visualization

370

80

where l is the distance between the channel inlet and view point,
and L is the effective length of the microchannel. Additionally,
Xp = lp/L is the dimensionless location to define the distance for
the appearance of injection flow in the process of flow condensation, in which lp is the distance between the microchannel inlet
and location of injection flow.
The vapor Reynolds number Rev (dimensionless form of the
vapor mixture mass flow rate) is derived by

Droplet

90

During the visual observation, the location of the shooting position is calculated in dimensionless form

X l=L

Fig. 2. Schematic of the experimental chip.

T /K

1137

140 160
p /kPa

180

200

Fig. 4. Relationship between the saturated temperature and pressure under

different mass concentrations of ethanol.

Using the above experimental setup, a wide range of variations

in ethanol mass concentration are examined to investigate the
two-phase flow condensation of mixture in a hydrophobic
microchannel. The visual results indicate that the flow pattern evolutions during flow condensation of ethanolwater mixtures in a
hydrophobic microchannel are highly correlated with the ethanol
concentration, as illustrated in Fig. 5. If the water steam is the main
component, the droplet condensation appears almost across the
whole channel surface in the two-phase flow region, in which
the droplet flow, droplet-streak flow, droplet-annular flow,
droplet-injection flow and droplet-slug/bubble flow occur sequentially along the flow stream in a microchannel (see Fig. 5(a)). With

1138

C. Zhang et al. / International Journal of Heat and Mass Transfer 104 (2017) 11351144

ethanol concentration

flow pattern distribution

Flow direction

(a)
I

II

Droplet- streak

Droplet

W = 2%, 10%

III

Droplet- annular Droplet- injection Droplet- slug/bubble

W = 10%

W = 2%

W = 2%

(b)

W = 20%, 40%, 60%

IV

Flow direction
I

II

Droplet

Droplet- streak

W = 40%

W = 20%

III

IV

Droplet- annular

Droplet- injection

W = 40%

(c)

Slug/bubble flow

W = 60%

W = 40%

Flow direction
I

II

III

IV

W = 80%
Streak flow Annular- streak

Annular flow

Injection flow Slug/bubble flow

Fig. 5. Schematics and corresponding images of flow pattern distributions of ethanolwater mixtures in a hydrophobic microchannel.

Rev

flow pattern distribution

Flow direction

(a)
Droplet

Droplet- annular

Rev = 327
Droplet- streak

Droplet- injection

(b)

Rev = 546

Flow direction

Droplet Droplet-streak Droplet-annular Droplet- injection Slug/bubble flow

(c)

Rev = 828

Slug/bubble flow

Flow direction

Droplet

Droplet- streak

Droplet- annular Droplet- injection Slug/bubble flow

Fig. 6. Flow patterns of ethanolwater mixtures under different Rev (W = 40%).

1139

C. Zhang et al. / International Journal of Heat and Mass Transfer 104 (2017) 11351144

Flow direction

0 ms

Flow direction

0 ms

0.67 ms

2 ms

16 ms
8 ms
(a) W = 2%, Rev = 241, X = 0.19, tin = 381.1K

1 ms
2.67 ms
(a) W = 2%, Rev = 644, X = 0.02, tin = 393.2K

steaks
0 ms

1 ms

0 ms

3 ms
4 ms
(b) W = 10%, Rev = 407, X = 0.02, tin = 383.2K

0 ms

2 ms

8 ms
7 ms
(b) W = 10%, Rev = 407, X = 0.13, tin = 383.8K

1 ms

2 ms
4 ms
(c) W = 40%, Rev = 546, X = 0.02, tin = 393.8K

0 ms

1 ms

2 ms

4 ms

(c) W = 40%, Rev = 546, X = 0.19, tin = 393.8K

Fig. 9. Effect of ethanol concentration on droplet-streak flow.

0 ms

1 ms
Flow direction
Film

2 ms

3 ms

Condensate streak

(d) W = 60%, Rev = 608, X = 0.03, tin = 371.9K

Film
Fig. 7. Effect of ethanol concentration on droplet flow at the entrance of
microchannel.

Flow direction

Streak

Fig. 10. Image of annular-streak flow (W = 80%, Rev = 879, X = 0.45, tin = 388.1 K).

vapor Reynolds number may also affect the two-phase flow patterns. Fig. 6 shows the flow patterns of condensation in a
hydrophobic microchannel as a function of vapor Reynolds number
for a given ethanol concentration (W = 40%). As shown, with
increasing Rev, the position of injection flow is postponed to the
channel outlet, and the length of two-phase flow region extends
accordingly.

Fig. 8. Image of pure streak flow (W = 80%, Rev = 879, X = 0.09, tin = 388.1 K).

3.1. Flow patterns in region I

the rising of ethanol concentration, the droplet-slug/bubble flow
disappears at the downstream of the microchannel, which is
replaced by the pure slug/bubble flow (see Fig. 5(b)). When the
ethanol vapor is the main component, the droplet condensation
almost disappears from the whole flow pattern evolution, and
the streak flow, annular-streak flow, annular flow, injection flow
as well as slug/bubble flow appear sequentially along the flow
stream in a microchannel (see Fig. 5(c)). On the basis of the
analysis of the observed flow condensation, the condensation flow
pattern evolutions in a hydrophobic microchannel along the flow
direction can be divided into five regions (marked as I, II, III, IV,
V in Fig. 5). In addition to the ethanol mass concentration, the

In the region I, i.e. the entrance of microchannel, the droplet

flow and streak flow are observed under different ethanol mass
concentrations. When the ethanol mass concentration is no more
than 60%, the droplet flow occurs at the region near the channel
inlet, as described in Fig. 5(a) and (b). Being similar to the condensation process of pure steam in a hydrophobic microchannel, the
condensation droplets slip along the channel wall and sweep the
other droplets downstream if the droplets become large enough
(see the red ellipses and pink rectangles in Fig. 7(a) and (b)).
Fig. 7 compares the droplet flow in a hydrophobic microchannel
under different ethanol mass concentration. It is seen that, as the

1140

C. Zhang et al. / International Journal of Heat and Mass Transfer 104 (2017) 11351144

Flow direction

0 ms

Discontinuous film

1 ms

2 ms
4 ms
(a) W = 2%, Rev = 380, X = 0.44, tin = 384.6K

Flow direction

0 ms

0.4 ms

Continuous film

0.8 ms
1.6 ms
(a) W = 80%, Rev = 879, X = 0.57, tin = 388.1K

Continuous film
0 ms

1 ms

4 ms
3 ms
(b) W = 10%, Rev = 408, X = 0.36, tin = 382.4K

0 ms

Continuous film

0.4 ms

0.8 ms

1.4 ms
(b) W = 80%, Rev = 879, X = 0.69, tin = 388.1K
Fig. 12. Images of annular flow.

Continuous film
0 ms

2 ms

6 ms
19 ms
(c) W = 20%, Rev = 391, X = 0.61, tin = 390.7K
Fig. 11. Effect of ethanol concentration on droplet-annular compound flow.

ethanol concentration increases, the condensation droplets

increasingly turn to be indistinct, and the droplet condensation
region gradually transforms from the whole channel wall to the
middle region. When the ethanol mass concentration is up to
40%, condensate lump appears near the solid wall, as shown by
the blue rectangle in Fig. 7(c). As the ethanol mass concentration
further increases to 60%, the condensate streak comes out at the
region of the condensate pump, as illustrated by the blue rectangle
in Fig. 7(d).
When the ethanol mass concentration reaches 80%, the pure
streak flow is observed in region I, as shown in Fig. 8. During the
condensation of ethanolwater mixture, the condensate liquid film
is potentially unstable [27]. The surface tension gradient induced
by the binary vapor enhances the irregularity of the liquid film,
and the condensate streak forms on the channel surface. In addition, the condensate liquid is drawn to the streak by this gradient,
and it flows downstream driven by the vapor mixture.
3.2. Flow patterns in region II
Following the droplet flow, the condensate streak flow appears
on the solid surface in the region II, i.e. the droplet-streak flow is
observed after the channel inlet, as shown in Fig. 9. The behaviors
of streak flow are varied with ethanol concentration. When the
ethanol concentration is small (W = 2%), the streak flow is formed
due to the coalescence of slipping droplets when the condensed
droplets are on the way downstream. As a result, the intermittent
streak flow is detected at the upstream of the microchannel, of

which the volume increases rapidly and then gradually decreases,

as illustrated in Fig. 9 (a).
With the increasing of ethanol concentration, the continuous
streak flow starts to come out at the region of condensate lump
due to the droplet coalescence, as shown in Fig. 9(b) and (c). There
are two streaks on both sides of the channel near the side surface.
The detached droplets, if reaching its critical size, flows along the
streak to the downstream, e.g. the droplet marked by the blue
ellipse in Fig. 9(b).
When the ethanol vapor is the main component (W = 80%,),
after the steak flow, with the gathering of the condensate from
upstream, the condensate accumulates at the channel corner
owing to the capillary force. As a result, the condensate covers
the side wall. In addition, the condensate streak flow still exists
on the channel surface. Consequently, the annular-streak flow is
observed in the region II, as presented in Fig. 10.

3.3. Flow patterns in region III

The condensate streak flow in region II is close to the side wall,
and the streak liquid gradually approaches to the channel corner
under the capillary force when the streak flow appears at the
downstream. Consequently, in region III, the droplet-annular flow
is observed when the ethanol mass concentration lies between
2% and 60% (see Fig. 11), but the annular flow is observed when
ethanol mass concentration is larger than 80% (see Fig. 12).
When the ethanol concentration is very small (e.g. W = 2%), the
liquid film on the side wall is discontinuous and intermittent with
the appearance of condensate from upstream, as indicated by the
Fig. 11(a). If the ethanol concentration is up to 10% or above, the
side wall is totally covered by the condensate liquid. Owing to
the instability of flow condensation for mixture vapor, the thickness of the continuous liquid film changes periodically, as shown
in Fig. 11(b) and (c). Additionally, the condensation droplet at
the region of droplet-annular flow gets larger with the increase
of ethanol concentration. For instance, the droplet size under the
condition of W = 20% is significantly larger than that under conditions of W = 2% and W = 10%. Once the ethanol concentration is

C. Zhang et al. / International Journal of Heat and Mass Transfer 104 (2017) 11351144

1141

Flow direction

0 ms

Discontinuous film

Slug

11.33 ms

20 ms

21.33 ms
(a) W = 2%, Rev = 309, Xp = 0.57, tin = 386.2K

Continuous film

Slug

0 ms

6 ms

6.67 ms
(b) W = 10%, Rev = 456, Xp = 0.48, tin = 385.5K

Continuous film

0 ms

Bubble

2.67 ms

18.67 ms

19.33 ms
(c) W = 40%, Rev = 614, Xp = 0.55, tin = 393.1K
Fig. 13. Images of droplet-injection compound flow.

very high (e.g. W = 80%), the droplet disappears and the channel
wall is covered by the condensate liquid film in the region III.
3.4. Flow patterns in region IV
Following the droplet-annular or annular flow, the dropletinjection flow (see Fig. 13) and injection flow (see Fig. 14) are
observed in the region IV of a hydrophobic microchannel during
the flow condensation. Since the liquid film on the side wall of
the channel is intermittent when the ethanol mass concentration
is 2%, the liquid film at the region of injection flow is also discontinuous. Furthermore, the slug shot from the vapor column is much
bigger than that under W > 10%, which fills the channel and cuts
the condensate liquid downstream. With increasing ethanol concentration, the interface of the condensate film becomes smoother.
This may be attributed to the recession of Marangoni effect owing
to decreasing surface tension gradient arising from the binary
vapor condensation.

Under some conditions at W = 80%, the bubble can not completely separate from the vapor ligament, which is squeezed by
the high speed vapor and broken into two or more separated bubbles, as illustrated in Fig. 14(b). Owe to the small size, the separated bubbles can not coalesce with each other until they flow
out of the channel.
Fig. 15 presents the injection flow location as a function of the
inlet vapor Reynolds number under different ethanol mass concentration. It can be seen that at large Rev (i.e. high mass flow rate), a
long flow path is needed to provide enough cooling time, which
ensures the sufficient condensate to form injection. As a result,
the dimensionless injection location increases with the increasing
Rev, as shown in Fig. 15. In addition, the condensation latent heat
of the ethanol vapor is much smaller than that of water steam, leading to less latent heat release during flow condensation when the
ethanol mass concentration is larger. For this reason, the ethanol
water mixture is easy to be condensed under larger ethanol mass
concentration, which decreases the injection flow location at the

1142

C. Zhang et al. / International Journal of Heat and Mass Transfer 104 (2017) 11351144

lesces with the following one to be a bigger one (slug 2 in Fig. 16

(b)). However, differing from slug/bubble flow with W P 20%
(see Fig. 17), the droplet condensation still exists in the dropletslug/bubble compound flow when W 6 10% (see Fig. 16) owing to
the effective hydrophobic effect of channel surface under small
ethanol mass concentration.

Flow direction

0 ms

0.33 ms

4. Influence of channel surface wettability on condensation

flow patterns

2 ms

4 ms

4.67 ms
(a) W = 80%, Rev = 827, Xp = 0.56, tin = 378.9K

0 ms

0.33 ms

In order to gain a further insight into the channel surface wettability effect on condensation flow patterns, we compare the current work with the previous experimental investigation by Jiang
et al. [25] on flow condensation of ethanolwater mixtures in the
non-treatment silicon microchannels, as shown in Fig. 18. Via the
comparison, the contact angle of water on the non-treatment silicon surface is measured to be 75 (see Fig. 19), which is smaller
than that on the treated channel wall in this paper (96), suggesting the smaller surface energy on the current hydrophobic channel
wall. It is experimentally observed by Jiang et al. [25] that droplet
condensation totally disappears for the flow condensation of ethanolwater mixtures in microchannels without treatment when the
ethanol mass concentration is 60%. However, for the same ethanol
mass concentration, the droplet condensation is still observed in
hydrophobic microchannels in the current study. This implies that
the hydrophobic surface is beneficial to the flow condensation heat
transfer.

0.67 ms
(b) W = 80%, Rev = 879, Xp = 0.75, tin = 388.1K

5. Wall temperature distributions along microchannels

Fig. 14. Images of Injection flow.

1.0
0.8

Xp

0.6
2%
10%
40%
60%
80%

0.4
0.2
0.0
200

400

600
800
Rev

1000

1200

Fig. 15. Injection location versus Rev (Symbols and lines are the experimental data
and fitted values, respectively).

same Rev. In summary, both an increase in vapor Reynolds number

and a decrease in ethanol concentration cause the injection location
move toward the channel outlet. The experiment results also indicate that the ethanol mass concentration has little influence on the
injection flow location when the concentration is larger than 60%.
3.5. Flow patterns in region V
In the region V, after the injection flow, the droplet-slug/bubble
compound flow occurs when W 6 10% (see Fig. 16), and slug/bubble flow occurs when W P 20% (see Fig. 17). For both of these two
flow patterns, the condensate stream mainly occupies the channel,
which entrains with vapor slug/bubble to flow towards the channel
outlet. Synchronously, the vapor slug/bubble shrinks due to the
continuous condensation (slug 1 in Fig. 16(a)) and usually coa-

Fig. 20 depicts the wall temperature distribution along the

channel under conditions of almost the same inlet vapor Reynolds
number with different inlet ethanol mass concentration. The wall
temperature decreases along the channel, and its average temperature is higher under the lower inlet ethanol mass concentration.
The explanation is that the ethanolwater mixtures with lower
inlet ethanol mass concentration (i.e. higher inlet water mass concentration) results in a more latent heat. Under a low inlet ethanol
mass concentration, e.g. W = 2%, the channel wall is mainly occupied by the droplet condensation before the injection flow. Hence,
the vapor with high temperature is easier to directly contact with
the cool channel wall, and thus the good performance of flow condensation heat transfer induces mild decrease in wall temperature
before the injection flow. As mentioned above, with the rising of
ethanol concentration, the droplet condensation becomes indistinct and the channel wall tends to be covered by condensate film,
which weakens the flow condensation heat transfer performance.
Therefore, the wall temperature drop before the injection flow
becomes more obvious under higher inlet ethanol mass concentration (e.g. W = 40%), as illustrated in Fig. 20.

6. Conclusions
An experimental investigation was performed to investigate the
flow condensation of ethanolwater mixtures in a wide range of
ethanol concentrations inside a hydrophobic rectangular
microchannel. The condensation flow patterns of mixtures in the
hydrophobic microchannel under different ethanol concentration
are presented and analyzed. In addition, the effect of channel surface wettability on condensation flow pattern and the effect of
ethanol concentration on heat transfer performance are discussed.
Finally, the main conclusions are drawn as follows:

1143

C. Zhang et al. / International Journal of Heat and Mass Transfer 104 (2017) 11351144

Flow direction

Slug 2

0 ms

49 ms

47 ms

Slug 1

coalescence

Slug 2

Slug 1

56 ms
Droplets

62 ms

(a) W = 2%, Rev = 197, X = 0.34, tin = 376.2K

Droplets
0 ms

2 ms

6 ms

30 ms
(b) W = 10%, Rev = 456, X = 0.55, tin = 385.5K

Fig. 16. Images of droplet-slug/bubble compound flow.

Flow direction

0 ms

0.67 ms

13.33 ms

19.33 ms

Fig. 17. Images of slug/bubble flow (W = 40%, Rev = 614, X = 0.57, tin = 393.1 K).

W = 2% 4% 6%

W = 20% 31%

W=60%
Fig. 18. Condensation flow patterns of ethanolwater mixtures in non-treated silicon microchannels [25].

(1) Droplet flow, annular-streak flow, droplet-annular flow,

droplet-injection flow and droplet-slug/bubble flow occur
sequentially during the flow condensation of mixtures in a
hydrophobic rectangular microchannel. When water steam
is the main component, the droplet condensation almost
occupies the whole two-phase flow region in the microchan-

nel. With the rising of ethanol concentration, the slug/bubble flow occurs without droplet condensation after the
droplet-injection flow. If the ethanol vapor is the main component, the droplet condensation almost disappears, and
annular-streak flow, annular flow, injection flow and slug/
bubble flow appear sequentially along the flow direction.

1144

C. Zhang et al. / International Journal of Heat and Mass Transfer 104 (2017) 11351144

Water droplet

Non-treated silicon wafer

Fig. 19. Contact angle of water on the non-treatment silicon surface.

Injection flow location

390

(c)

(b)

(a)

375

t /K

360
345
330
315
300
0.0

(a)
(b)
(c)
0.2

0.4

0.6

0.8

1.0

X
Fig. 20. Wall temperature distribution along the channel. (a) W = 2%, Rev = 443,
tin = 391.1 K; (b) W = 10%, Rev = 456, tin = 385.6 K; (c) W = 40%, Rev = 449,
tin = 373.9 K.

(2) The injection location increases with increasing Rev, and it

decreases with the increasing ethanol mass concentration
at the same Rev, which is attributed to small latent heat
release of ethanolwater mixtures.
(3) Compared with non-treated silicon microchannels, the current microchannels with hydrophobic treatment facilitate
the droplet condensation, which is beneficial for the flow
condensation heat transfer.
(4) The wall temperature decreases along the flow direction.
Especially, under a lower inlet ethanol concentration, the
average wall temperature is higher, and the decrease in wall
temperature along the flow direction exhibits milder before
the injection flow.

Acknowledgements
The authors gratefully acknowledge the supports of the
National Natural Science Foundation of China (51306033), Natural
Science Foundation of Jiangsu Province (BK20130621), and 333
project of Jiangsu Province in China.
References
[1] T.T. Hao, X.H. Ma, Z. Lan, R. Jiang, X. Fan, Analysis of the transition from laminar
annular flow to intermittent flow of steam condensation in noncircular
microchannels, Int. J. Heat Mass Transfer 66 (2013) 745756.

[2] Y.P. Chen, M.H. Shi, P. Cheng, G.P. Peterson, Condensation in Microchannels,
Nanosc. Microsc. Therm. 12 (2008) 117143.
[3] P. Cheng, H.Y. Wu, Mesoscale and microscale phase change heat transfer, Adv.
Heat Transfer 39 (2006) 461563.
[4] J.W. Coleman, S. Garimella, Characterization of two-phase flow patterns in
small diameter round and rectangular tubes, Int. J. Heat Mass Transfer 42
(1999) 28692881.
[5] B. Mdric, M. Miscevic, V. Platel, P. Lavieille, J. Joly, Experimental study of flow
characteristics during condensation in narrow channels: the influence of the
diameter channel on structure patterns, Superlattices Microst. 35 (2004) 573
586.
[6] A. Agarwal, S. Garimella, Representative results for condensation
measurements at hydraulic diameter 100 l, J. Heat Trans. 132 (2010)
041010.
[7] Y.P. Chen, R. Wu, M.H. Shi, J.F. Wu, G.P. Peterson, Visualization study of steam
condensation in triangular microchannels, Int. J. Heat Mass Transfer 52 (2009)
51225129.
[8] X.H. Ma, X.G. Fan, Z. Lan, T.T. Hao, Flow patterns and transition characteristics
for steam condensation in silicon microchannels, J. Micromech. Microeng. 21
(2011) 075009.
[9] S. Garimella, Condensation flow mechanisms in microchannels: basis for
pressure drop and heat transfer models, Heat Transfer Eng. 25 (2004) 104116.
[10] H.Y. Wu, M.M. Yu, P. Cheng, X. Wu, Injection flow during steam condensation
in silicon microchannels, J. Micromech. Microeng. 17 (2007) 16181627.
[11] X.J. Quan, P. Cheng, H.Y. Wu, Transition from annular flow to plug/slug flow in
condensation of steam in microchannels, Int. J. Heat Mass Transfer 51 (2008)
707716.
[12] Y.P. Chen, P. Cheng, Condensation of steam in silicon microchannels, Int.
Commun. Heat Mass 32 (2005) 175183.
[13] W. Zhang, J.L. Xu, J.R. Thome, Periodic bubble emission and appearance of an
ordered bubble sequence (train) during condensation in a single microchannel,
Int. J. Heat Mass Transfer 51 (2008) 34203433.
[14] J.F. Wu, M.H. Shi, Y.P. Chen, X. Li, Visualization study of steam condensation in
wide rectangular silicon microchannels, Int. J. Therm. Sci. 49 (2010) 922930.
[15] Y.P. Chen, C.Q. Shen, M.H. Shi, G.P. Peterson, Visualization study of flow
condensation in hydrophobic microchannels, AIChE J. 60 (2014) 11821192.
[16] C. Fang, J.E. Steinbrenner, F.M. Wang, K.E. Goodson, Impact of wall
hydrophobicity on condensation flow and heat transfer in silicon
microchannels, J. Micromech. Microeng. 20 (2010) 045018.
[17] V.V. Mirkovich, R.W. Missen, Study of condensation of binary vapors of
miscible liquids, Can. J. Chem. Eng. 39 (1961) 8687.
[18] J.J. Yan, J.S. Wang, Y.S. Yang, S.H. Hu, J.P. Liu, Research on marangoni
condensation modes for waterethanol mixture vapors, Microgravity Sci.
Technol. 21 (2009) 7785.
[19] T. Fujii, N. Osa, S. Koyama, Free convection condensation of binary mixtures on
a smooth tube: condensing mode and heat transfer coefficient of condensate,
Proc.US Eng. Found. Conf. (1993) 171182.
[20] J.S. Wang, J.J. Yan, S.H. Hu, J.P. Liu, Marangoni condensation heat transfer of
waterethanol mixtures on a vertical surface with temperature gradients, Int.
J. Heat Mass Transfer 52 (2009) 23242334.
[21] Y. Utaka, S. Wang, Characteristic curves and the promotion effect of ethanol
addition on steam condensation heat transfer, Int. J. Heat Mass Transfer 47
(2004) 45074516.
[22] D. Del Col, A. Cavallini, J.R. Thome, Condensation of zeotropic mixtures in
horizontal tubes: new simplified heat transfer model based on flow regimes, J.
Heat Transfer 127 (2005) 221230.
[23] T. Murase, H.S. Wang, J.W. Rose, Marangoni condensation of steamethanol
mixtures on a horizontal tube, Int. J. Heat Mass Transfer 50 (2007) 37743779.
[24] X.H. Ma, Z. Lan, W. Xu, M.Z. Wang, S. Wang, Effect of surface free energy
difference on steam-ethanol mixture condensation heat transfer, Int. J. Heat
Mass Transfer 55 (2012) 531537.
[25] R. Jiang, X.H. Ma, Z. Lan, Y.X. Bai, T. Bai, Visualization study of condensation of
ethanolwater mixtures in trapezoidal microchannels, Int. J. Heat Mass
Transfer 90 (2015) 339349.
[26] C.R. Wilke, A viscosity equation for gas mixtures, J. Chem. Phys. 18 (1950) 517
519.
[27] K. Hijikata, Y. Fukasaku, O. Nakabeppu, Theoretical and experimental studies
on the pseudo-dropwise condensation of a binary vapor mixture, J. Heat
Transfer 118 (1996) 140147.
[28] H. Ju, Investigation of the effects of the anisotropy of gas-diffusion layers on
heat and water transport in polymer electrolyte fuel cells, J. Power Sources 191
(2009) 259268.