Full text of DOI:10.1080/10789669.2002.10391294

HVAC&R Research
ISSN: 1078-9669 (Print) 1938-5587 (Online) Journal homepage: http://www.tandfonline.com/loi/uhvc20
Electrohydrodynamically Enhanced Convective
Boiling of Alternative Refrigerants: Fundamental
Understanding and Applicability
James E. Bryan & Jamal Seyed-Yagoobi
To cite this article: James E. Bryan & Jamal Seyed-Yagoobi (2002) Electrohydrodynamically
Enhanced Convective Boiling of Alternative Refrigerants: Fundamental Understanding and
Applicability, HVAC&R Research, 8:4, 337-355
To link to this article: http://dx.doi.org/10.1080/10789669.2002.10391294
Published online: 28 Feb 2011.
Submit your article to this journal
Article views: 18
View related articles
Citing articles: 1 View citing articles
Full Terms & Conditions of access and use can be found at
http://www.tandfonline.com/action/journalInformation?journalCode=uhvc20
Download by: [University of Auckland Library]
Date: 05 June 2016, At: 09:34

VOL. 8, NO. 4
HVAC&R RESEARCH
OCTOBER 2002
Electrohydrodynamically Enhanced Convective
Boiling of Alternative Refrigerants:
Fundamental Understanding and Applicability
James E. Bryan, Ph.D.
Jamal Seyed-Yagoobi, Ph.D., P.E.
Associate Member ASHRAE
Associate Member ASHRAE
Electrohydrodynamically (EHD) enhanced convective boiling, based on the extraction phenom-
enon, is experimentally investigated. The working fluids are R-134a and R-404A refrigerants.
Both smooth and enhanced tubes are investigated in a horizontal configuration. The heat trans-
fer enhancement and suppression with the corresponding pressure-drop data are presented. The
ratio of EHD power consumption relative to the heat transferred was less than 1.5% for the data
presented. A theoretical analysis is presented to determine the contributing terms of the EHD
power consumption, and it was determined that the electric conduction mechanism was the pri-
mary contributing factor to current flow. The transient application of EHD, in the case that
resulted in the suppression of the heat transfer under steady-state conditions, provided transient
enhancements for convective boiling of R-404A in a horizontal smooth tube. Additionally, appli-
cability of the EHD extraction phenomenon to convective boiling is discussed.
INTRODUCTION
Active enhancement of convective boiling using electrohydrodynamic (EHD) phenomena has
been studied for about a decade (Seyed-Yagoobi and Bryan 1999). The largest EHD convective
boiling enhancements achieved from seven available references are shown in Table 1; informa-
tion on the corresponding experimental parameters is given in Table 2. Bryan and Seyed-
Yagoobi (1997, 2000) showed that enhancements were highly dependent on the quality, flow
regime, heat flux, mass flux, and strength of the EHD forces relative to the flow momentum.
The occurrence of the maximum enhancements shown in Table 1 for the different researchers
was primarily due to the electrode designs, the geometry of the heat transfer surface, and the
operating conditions. However, both Bryan and Seyed-Yagoobi (1997, 2000) and Norris et al.
(1997) showed that in many cases the EHD forces can drastically reduce the rate of heat transfer.
In convective boiling with the electrode designs considered in these studies, the EHD force is
primarily perpendicular to the bulk fluid motion, and its influence is strongly dependent on the
variables mentioned previously and on the fluid electrical properties. The EHD force, which acts
primarily perpendicular to the fluid flow, can also result in increased flow resistance, thereby
increasing the pressure drop. Bryan and Seyed-Yagoobi (2000) developed a simple theory to
determine the mean radial EHD pressure, which included the characteristics of two-phase flow.
Their analysis showed that the amount of heat transfer enhancement and the pressure-drop pen-
alty depended on the size of the mean radial EHD pressure relative to the flow axial momentum
flux rate.
The ranges of EHD-enhanced heat transfer and pressure-drop penalty in convective boiling
obtained by these researchers as well as by Cotton et al. (2000) are shown in the results. EHD
James E. Bryan is a professor in the Department of Mechanical and Aerospace Engineering, University of Missouri,
Columbia. Jamal Seyed-Yagoobi is chair and professor in the Department of Mechanical, Materials, and Aerospace
Engineering, Illinois Institute of Technology, Chicago.
337

338
HVAC&R RESEARCH
Table 1. EHD Enhancements and Corresponding Operating Parameters for Researchers
6
6
G,
x,
q″,
h ,
h
,
E (10 ) E (10 ) Q
,
o
EHD
E
G
EHD
W
2
2
2
2
Reference
Fluid kg/(m ·s) kg/kg kW/m W/(m ·K) W/(m ·K) V/m
V/m
b,f
c
Yabe et al. (1992) R-123/
R-134a
33
0.2 to
0.6
4
500 to
1000
1050 to
2000
4.04
2.02
N/A
a
b,f
b,f
g
Singh et al. (1994) R-123
Singh et al. (1995) R-134a
75
200
50
0
5
5
800
3800
4600
5.83
5.83
2.22
4.55
1.86 0.34
1.86 0.23
a
0
2100
2000
5000
3000
3500
a
c
c
Singh (1995)
R-134a
0.1
5
13 900
50 000
7000
1.84 0.36
e
a
g
g
c
d
d
Ohadi et al. (1995) R-134a
Salehi et al. (1996) R-404A
50
50
0.6
25
4.04 0.01
a
c
c
0.3
10
1.66
8.62
1.38 0.01
b,f
Bryan and
R-134a
99.9
0.1
11
22 870
0.97
N/A
Seyed-Yagoobi
(1997, 2000)
aMeasured at inlet to test section and different from
local change of quality in the test section
bAverage heat flux; not constant along test section
cElectric field based on radius; local electric field is
different because of geometrical shape of surface
dPower estimated from power ratio presented by researchers
eEstimated from Reynolds number presented by researchers;
based on cross-sectional area of microchannel
fTest section heated by hot water
gTest section heated by resistive electrical heater
Table 2. Experimental Parameters Investigated by Various Researchers
G
x
q″
D ,
D ,
mm
T ,
sat
range, range, range,
o
E
2
2
Reference
Tube
mm L, m Electrode
°C kg/(m ·s) kg/kg kW/m
Yabe et al.
(1992)
Perforated
tube
c
Smooth
10
9.4
9.4
3.75
5
30 33 and 66 0 to 0.9
4
Singh et al.
(1994)
b
c
c
Smooth
Smooth
1.22 Cylindrical
1.22 Cylindrical
3
25 50 to 400 0 to 0.5 5 to 20
Singh et al.
(1995)
b
3
25 50 to 400 0 to 0.5 5 to 20
b
Singh (1995)
Microfin 12.7
0.3
Helical
9.8
25 50 to 150 0 to 0.8 5 to 30
Ohadi et al.
(1995)
50 and
100
a
a
b
Microfin 12.7
0.11 Cylindrical
9.5
25
0 to 0.6
25
Salehi et al.
(1996)
b
Microfin 12.7
0.3
Helical
9.8
1.6
25 50 to 200 0 to 0.8
5, 10
Bryan and
Seyed-Yagoobi Smooth
(1997)
0.1,
100 and
c
15.9 0.2, 0.3, Cylindrical
0.5
5, 25
0 to 0.6 4 to 153
300
aAnnulus resulting from tube and cylindrical elec-
trode is separated into 12 microchannels, each with
a hydraulic diameter on order of 1 mm.
bMeasured at inlet to test section and different from test
section quality.
cHeat flux is averaged and is not constant along test section.
enhancements are affected by the mass flux, quality, flow regime, heat flux, operating tempera-
ture, geometrical characteristics of the heat transfer surface and electrode, applied electric field,
and fluid properties. In this paper, the effects of the EHD force on convective boiling are
described based on the knowledge to date. Experimental results are presented to further explain
the role of the heat transfer surface geometry and the refrigerant when the EHD force is applied.
Additional results are provided to illustrate the transient effects that are much different from the
steady-state effects. Finally, the applicability of the EHD phenomena to convective boiling is
briefly discussed.

VOLUME 8, NUMBER 4, OCTOBER 2002
339
EHD is an interdisciplinary phenomenon dealing with the interaction between electric fields
and flow fields in a dielectric fluid medium. This interaction can result in electrically induced
fluid motion and interfacial instabilities, which are caused by an electric body force. The electric
body force density acting on the molecules of a fluid in the presence of an electric field consists
of three terms (Melcher 1981):
1
2
2
1
2
2
∂ε
∂ρ




--
--
------
f
Z ρ E Ó E ∇ε H ∇ E ρ
(1)
e
e
The three terms in Equation (1) stand for two primary force densities acting on the fluid. The
first term represents the force acting on the free charges in the presence of an electric field and is
known as the Coulomb force. The second and third terms represent the polarization forces
induced in the fluid.
The electric body force density components defined in Equation (1) are responsible for the
induced forces on the dielectric fluid medium. Two examples are
• A charged body in a nonuniform or uniform electric field will move along the electric field
lines and impart momentum to the surrounding fluid. This Coulomb force is expressed by
the first term in Equation (1).
• If an interface exists, such as between a liquid and vapor in a nonuniform electric field, an
attraction force is created that draws the fluid of higher permittivity (liquid) toward the region
of higher electric field strength. This dielectrophoretic force results from the application of a
nonuniform electric field in the presence of a permittivity gradient. This force is described by
the second term in Equation (1).
All three components of the EHD force density can be significant, or one can dominate over
the others. In convective boiling, all three forces will be present; however, the second term of
Equation (1) is expected to dominate as the quality and heat flux increase.
EXPERIMENTAL METHODS
The experimental apparatus, shown in Figure 1, is a closed loop that refrigerant is pumped
through with a positive-displacement gear pump instead of a compressor. R-134a and R-404A
refrigerants were used in all experiments. The experimental apparatus is described in more detail
by Bryan (1998).
The four test sections of 10, 20, 30, and 50 cm in length, shown in Figure 1, were connected
in series and heat was provided to each by a recirculating hot-water loop. The different test sec-
tions were used to allow measurement of the local heat transfer coefficients and heat fluxes.
Each test section, as shown in Figure 2, consisted of a copper boiling tube surrounded by an
acrylic tube [inside diameter (ID) 38.1 mm] that the hot water flowed through. The boiling tubes
were either a smooth tube [ID = 14.1 mm, outside diameter (OD) = 15.9 mm] or a microfin tube
(inside back-wall diameter = 14.6 mm, OD = 15.9 mm) and were oriented in a horizontal config-
uration for all experiments. T-type thermocouples were soldered to the surface of the boiling
tubes at the top, middle, and bottom in each test section. In the 10 and 20 cm sections, the sur-
face thermocouples were soldered to the tube surface at the center of the test section. In the
30 cm section, they were soldered in two stations at 10 cm intervals, and in the 50 cm section in
three stations at 12.5 cm intervals. The surface thermocouples allowed measurement of the local
heat transfer coefficient in each test section. The inlet and outlet water temperatures of each test
section were measured with thermistors so that in all experiments the temperature difference
across each test section could be accurately measured and maintained to 1 K or less. The pres-
sure drop across each test section was measured with a variable-reluctance pressure transducer.

340
HVAC&R RESEARCH

VOLUME 8, NUMBER 4, OCTOBER 2002
341
Figure 2. Drawing of Test Section
The electrode used in the experiments was a brass rod 1.6 mm in diameter (see Figure 2). The
brass electrode was suspended in the center of each test section and held in tension by two elec-
trode supports, one at the entrance and one at the exit of the test section. In all experiments, the
brass electrode was energized with a positive dc high voltage while the copper boiling tube was
grounded.
In all experiments, the heat transfer coefficient and quality were calculated from direct mea-
surements. The average overall heat transfer coefficient for the four test sections combined was
determined from an area average of the local heat transfer coefficients for the 10, 20, 30, and
50 cm sections. The local heat transfer coefficient at each location inside the tube was calculated
from
·
m
-----------------------------------------
c (T Ó T )
out
p
in
h Z
(2)
πD L(T Ó T )
sat
i
wi
The local inside wall temperature T was calculated through a one-dimensional conduction
wi
analysis, knowing the local outside surface temperature measured at the different locations in
each test section, as described earlier. The conduction through the wall was accounted for even
though its influence was very small. The change in quality was determined from
·
mc (T Ó T
)
--------------------------------------
p
in
out
∆x Z
(3)
·
m
h
ref fg
The change in quality ∆x was the difference between the exit quality from the last test section
and the inlet quality from the first test section. In this study, however, liquid at the saturated con-
dition always entered the first test section (10 cm section), corresponding to the inlet quality of
zero. Thus, ∆x was equal to the actual quality at the exit of the last test section.

342
HVAC&R RESEARCH
Table 3. Experimental Uncertainties
Total Uncertainty
Measurements
Average
0.09°C
0.13°C
0.06°C
Maximum
0.41°C
Surface temperature
Inlet and outlet refrigerant temperature
Inlet and outlet water temperature
Refrigerant mass flux
Water mass flow rate
Pressure drop
0.17°C
0.09°C
2
2
4.2 kg/(m ·s)
4.3 kg/(m ·s)
0.0027 kg/s
0.056 kPa*
Average, %
13.8
0.0027 kg/s
0.074 kPa*
Maximum, %
20.5
Calculated Quantities
Quality
Heat flux
13.1
20.0
Heat transfer coefficient
13.4
39.2
*Uncertainties were smallest on 10 cm section and greatest on 50 cm section, and increased as heat load increased.
A total uncertainty analysis was performed for all the measured data and calculated quantities
based on methods described by Kline and McClintock (1953) and Taylor (1982). The maximum
total uncertainties are shown in Table 3 for the different measured and calculated quantities. Great
care was taken to properly calibrate all instrumentation, which allowed for a complete uncertainty
analysis to be performed with the lowest possible experimental uncertainties. The highest uncer-
tainties resulted when the temperature differences were the smallest, as defined in Equation (2).
This occurred at the lowest heat fluxes and when the EHD enhancement was the greatest.
The experimental data for heat transfer and pressure drop without the application of EHD were
compared to those of several different researchers (Eckels et al. 1994; Kandlikar 1990; Torikoshi
and Ebisu 1993; Wattelet 1994; Wattelet et al. 1994) to verify the quality of the data acquired in
this work. The electrode did influence the heat transfer and pressure drop without the application
of EHD. The heat transfer was enhanced from a few percent up to about 25%, with the high
2
enhancements occurring at mass fluxes above 250 kg/(m ·s). The pressure drop increased about
20 to 25% because of the presence of the electrode and was effectively independent of the mass
2
flux between 100 and 400 kg/(m ·s). For complete details, refer to Bryan (1998).
RESULTS AND DISCUSSION
EHD Enhanced Convective Boiling—Qualitative Discussions
The qualitative discussions given here are supported by the recent theoretical and experimen-
tal work of Bryan and Seyed-Yagoobi (1997, 2000). The EHD force in convective boiling is pri-
marily perpendicular to the bulk fluid motion, especially with the electrode design considered in
this study and other previous studies. The influence of this force is strongly dependent on qual-
ity, flow regime, heat flux, mass flux, fluid properties, electrode and surface geometry, and the
applied voltage.
The saturated liquid begins to change phase as it enters the evaporator tube. The fluid transi-
tions from single-phase liquid through the different two-phase flow regimes. Nucleate boiling
and convection are the primary modes of heat transfer to the fluid. The contribution of these two
heat transfer mechanisms varies depending on the fluid properties, the mass flux, the quality,
and the heat flux. Because the flow is in a horizontal configuration, the mass flux will influence
the distribution of the liquid and vapor in the tube. The distribution of the liquid and vapor will
substantially influence the local heat transfer. At low mass fluxes, the flow will be wavy-
stratified. As the mass flux increases, the amount of stratification will decrease and the liquid

VOLUME 8, NUMBER 4, OCTOBER 2002
343
will become more evenly distributed around the tube wall as the flow transitions to the annular
regime. The pressure drop will also be significantly influenced by the two-phase flow. The inter-
action between the two phases will drastically increase the frictional losses in the flow. The
momentum of the flow over a given tube length will increase as more energy is added to the
flow. The increases in both the flow friction and flow momentum will significantly increase the
pressure drop.
When an electric field is applied in the convective boiling process, the EHD force density
components, as defined in Equation (1), produce some interesting behavior. In the bubbly and
plug flow regimes, the electric force generates secondary motions in the liquid. The polariza-
tion component of the electric force also acts on the bubble at the liquid-vapor interface to
move it around (because of the nonuniformity in the electric field distribution caused by the
presence of a bubble) and break it apart. At this point, the EHD force density components are
enhancing the heat transfer; because their influence is significant, they are also generating a
substantial increase in the pressure drop. However, the heat flux can play a significant role in
the amount of EHD enhancement, and potentially lead to heat transfer suppression. The bub-
bles are held around the tube surface by the polarization forces. Because the bubbles are of
lower permittivity than the surrounding liquid, they will be driven to the region of lowest elec-
tric field strength, which is the tube wall. If the heat flux is high enough, the significant bubble
nucleation will start to increase the resistance to heat transfer at the wall, leading to heat trans-
fer suppression.
When the flow transitions to the annular regime, the role of the EHD force density compo-
nents change. The Coulomb and polarization forces still generate secondary motions in the liq-
uid, but in this regime there is less liquid to act on because more energy has been added to the
flow. The polarization forces start to play a more significant role at this point in the flow. The
amount of heat transfer enhancement or suppression as well as pressure drop will depend on the
strength of the EHD forces relative to the flow momentum. If the polarization forces only thin
the liquid layer at the tube surface, heat transfer can be significantly enhanced. However, sup-
pression will occur when the polarization forces hold the vapor bubbles on or near the tube wall
as they form and start to remove the thin liquid layer from the tube surface. Again, the heat flux
will also play a role in establishing when the polarization forces will enhance or suppress the
heat transfer at this location in the flow. The pressure drop will increase regardless of enhance-
ment or suppression because the polarization forces increase the liquid-vapor interaction, result-
ing in an increase in the frictional losses. The influence of the polarization forces is expected to
dominate in refrigerant flows, because transition to annular flow can occur at a quality of 0.1.
This type of EHD polarization force is classified as an extraction phenomenon, which occurs
when a liquid layer of higher permittivity is attracted to a region of high electric field strength
(e.g., the electrode shown in Figure 2), which is surrounded by a vapor of lower permittivity.
The heat transfer surface geometry can also play a role in the EHD enhancement potential. An
enhanced heat transfer surface will result in a nonuniform electric field distribution at the sur-
face. This nonuniform field can potentially help further improve the EHD enhancement. In con-
vective boiling, it is most desirable to have the greatest nonuniform electric field and electric
field strength near the tube surface. However, the greatest electric field presented in this work is
near the tube center at the high voltage electrode.
The EHD forces will change the local void fraction distribution in convective boiling. At low
qualities, the local void fraction will be greatest at the tube wall and decrease toward the tube
centerline when the EHD forces are applied. The EHD forces will also change the local void
fraction distribution in annular flow if the EHD forces are large enough. If the EHD forces are
enhancing the heat transfer, the local void fraction distribution will be more uniform than when
the EHD forces are suppressing heat transfer.

344
HVAC&R RESEARCH
Overall Heat Transfer and Pressure-Drop Results
Applying the EHD forces to the convective boiling process resulted in the enhancement and
even suppression of local heat transfer, as well as an increase in pressure drop. The measure-
ments of the overall EHD-enhanced or suppressed heat transfer and the pressure-drop penalty
are also important in assessing potential industrial applications. The overall heat transfer and
pressure-drop results are presented in this section.
The overall heat transfer coefficients and the total pressure drop as a function of the change in
quality are shown in Figures 3 through 6 for different conditions. Each figure contains the results
for one saturation temperature and two mass fluxes. The data illustrated in these figures are for all
four test sections combined and represent the overall values of the heat transfer coefficient and
2
the pressure drop. However, the 10 cm section was not used in the results for G = 101.1 kg/(m ·s)
shown in Figure 5, because of an electrical connection problem. The increase in ∆x was obtained
by increasing the heat flux to the test sections. The increase in the heat transfer coefficient with-
out EHD as ∆x increases derives from the influence of both nucleate boiling and convection. The
pressure drop increases with an increase in ∆x as well, because as more energy is added to the
flow, the change in frictional losses and the flow momentum become larger because of the
2
increasing vapor velocity. The flow, except for the results of R-404A at G = 101.1 kg/(m ·s) and
T
= 5.2°C, entered the first test section as a single-phase liquid and exited the last test section
sat
2
in the annular regime. The results shown in Figure 5 for R-404A at G = 101.1 kg/(m ·s) and T
sat
= 5.2°C represent stratified-annular flow entering the first test section and exiting the last test
section.
The heat transfer enhancement or suppression and the pressure-drop penalty resulting from
application of EHD forces are different for the different conditions as shown in Figures 3
through 6. The heat transfer enhancement and suppression occurring over the range of ∆ x derive
from the factors discussed earlier. As the flow progresses through the test sections, the role and
influence of the EHD force density components, defined in Equation (1), vary depending on the
Figure 3. Overall Heat Transfer and Pressure-Drop Results for R-134a in
Smooth Tube at T = 5°C
sat

VOLUME 8, NUMBER 4, OCTOBER 2002
345
Figure 4. Overall Heat Transfer and Pressure-Drop Results for R-134a in
Microfin Tube at T = 5°C
sat
Figure 5. Overall Heat Transfer and Pressure-Drop Results for R-404A in
Smooth Tube at T = 5°C
sat
total change in quality. When the quality is low, the EHD force density components are acting to
enhance the heat transfer and substantially increase the pressure drop. This occurs in all cases
2
except for R-134a in the microfin tube at G = 105.0 kg/(m ·s) and T = 5.0°C (see Figure 4).
sat
This is a case where the EHD forces at 15 kV are producing an unstable behavior in the two-
phase flow, resulting in mostly heat transfer suppression. The unstable behavior was verified by

346
HVAC&R RESEARCH
Figure 6. Overall Heat Transfer and Pressure-Drop Results for R-404A in
Microfin Tube at T = 5°C
sat
repeating the test conditions shown in Figure 4. When ∆x (i.e., heat flux) increases further, the
role of the EHD force components changes. Specifically, the polarization forces [an extraction
phenomenon primarily from the second term on the right-hand side of Equation (1)] start to play
a more significant role as more energy is added to the flow. If the extraction phenomenon is
large enough, for a given set of operating conditions, it will strip the liquid layer from the wall,
causing local dryout, which will result in suppression of the heat transfer. This is evident in Fig-
2
ures 3 and 4 for a mass flux of approximately 100 kg/(m ·s) and change in quality greater than
0.5. At 5 kV there is little or no heat transfer enhancement, but at 15 kV the extraction phenom-
enon is large enough to significantly suppress heat transfer.
Figures 3 through 6 indicate that application of EHD forces provide enhancement at low qual-
ities and suppression at qualities higher than 0.5 under the conditions in this research. In this
work the change in quality was achieved by increasing the heat flux. Thus, heat flux and quality
both influence the net effect of application of electric forces on heat transfer. In this work, when
∆x was low, heat flux was low in all test sections, usually resulting in EHD enhancement. High
∆x corresponded to high heat flux in all test sections. Only for R-134a in the smooth and micro-
2
fin tubes at G = 100 kg/(m ·s) and T = 5°C (see Figures 3 and 4) did total heat transfer sup-
sat
pression occur at the highest heat fluxes.
The interdependence of the convective boiling parameters and the EHD parameters can be
seen by comparing the results shown in Figures 3 through 6. EHD enhancement was greater
for R-134a in the smooth tube, but not in the microfin tube when compared to R-404A for the
same applied voltage, electrode design, and saturation temperature. This behavior is prima-
rily attributed to differences in the fluid properties, as shown in Table 4. The fact that
R-404A is a mixture (near-azeotropic blend of R-125, R-143a, and R-134a) may play a role
as well because of the differences in the electrical and thermophysical properties of the mix-
ture components.

VOLUME 8, NUMBER 4, OCTOBER 2002
347
EHD Power Consumption
Applying the electric field to heat transfer and fluid flow results in minor power consumption.
The amount of EHD power consumption, or more specifically the current density, will depend
on the specific application, the dielectric fluid medium, the applied potential, and the type and
magnitude of the EHD forces. The total current density is defined as (Bryan 1998; Seyed-
Yagoobi and Bryan 1999)
J Z σ E H ρ H ρ µ E
(4)
u
e
e
e e
The current density will be dependent only on the electric conductivity and the electric field
strength if the electric charge density is small. In the convective boiling experiments in this
research, the electric charge density was expected to be small, and primarily attributed to im-
purities, temperature gradient, and interfacial gradient because direct injection of the free
charges was significantly reduced or eliminated. Then, the EHD power was calculated from
the following:
Q
Z VI Z V(2πr Lσ E )
(5)
EHDIσe
e
eIm
where the electric field E and the mixture electric conductivity σ are defined in Table 5.
e,m
Table 4. Properties for R-134a and R-404A at T = 5°C
sat
σaeI1
S/m
,
ρ ,
ρ , µ × 10 , µ × 10 ,
k ,
σ ,
c ,
h
fg
J/kg
,
4
4
1
v
1
v
1
st
pl
a
3
3
κ
kg/m kg/m kg/(m·s) kg/(m·s) W/(m·K) N/m J/(kg·K) Pr
1
b
R-134a 11.8 2.93E-8 1276.6 17.1
2.504
1.689
0.112
0.122
0.093
0.068
0.011 1354.5 3.7 195151
0.007 1368.9 3.4 163983
b
R-404A
9.8 2.95E-8 1140.5 35.2
aMeasured in this research.
bProperties obtained from the 2001 ASHRAE Handbook—Fundamentals.
Table 5. Electric Field Strength Equations for Annular and Homogeneous Approaches
with Supporting Equations to Calculate EHD Power Defined in Equation (5)*
Electric Field Strength at Energized Electrode
Only liquid present around
electrode, using annular
approach
Only vapor present around
electrode, using annular
approach
Homogeneous mixture
of liquid and vapor around
electrode
εv
ε1
----
----
V
V
ε1
εv
V
--------------------------------------------------------
--------------------------------------------------------
-----------------------
E Z
(6) E Z
(7)
E Z
(8)
εv
r
r
ε1
r
r
r

i 
----

o 
----

i 
----

o 
----

o 
----
re ln
----
re ln
----
H ln
H ln
reln










ε1
re
ri
εv
re
ri
re
ri Z (1 Ó αs)r2o H αsre2 (9)
ri Z (1 Ó αs)r2e H αsro2 (10)
0.89
0.18 Ó1
Ó1
ρ
µ
ρ

ρ1
1 Ó x
x
1 Ó x
x

  v 

1 
-----

  v 
----------- -----
----------- -----
αs
Z
1 H
(11)
αH
Z
1 H
(12)

 




 
ρ1
µv
σeIm Z (1 Ó αs)σeI1 H αsσeIv (13)
σeIm Z (1 Ó αH)σeI1 H αHσeIv (14)
Note: Equation (11) is Thom (1964) correlation.
*See Bryan and Seyed-Yagoobi (2000) for further details.

348
HVAC&R RESEARCH
Figure 7. EHD Power Versus Quality for R-134a in Smooth Tube at 15 kV,
2
G = 300.4 kg/(m ·s), and T = 24.9°C
sat
The amount of liquid and vapor present will influence the amount of EHD power consumed.
As quality increases, EHD power decreases, because the amount of vapor (less conductive than
the liquid phase) is increasing. This can be supported analytically by using the theoretical
approach presented by Bryan and Seyed-Yagoobi (2000) to calculate EHD power from Equation
(5) based on the first term in Equation (4). Table 5 lists the equations for E and σ based on the
e,m
theoretical approach of Bryan and Seyed-Yagoobi (2000). Experimental data and analytical
results for EHD power as a function of quality are shown in Figure 7 for different test sections
2
with R-134a in the smooth tube at G = 300.4 kg/(m ·s) and T = 24.9°C.
sat
Both the experimental data and the analytical results indicate that a large portion of the cur-
rent flow derives from electric conduction, the first term on the right-hand side of Equation (4).
The electric field calculated based on Equation (6), given in Table 5, assumes that all the liquid
in the tube surrounds the energized electrode. The liquid extraction phenomenon promotes this
by attracting the liquid, which has higher permittivity than the vapor, toward the energized elec-
trode. However, as flow momentum increases, the attraction of the liquid to the energized elec-
trode will decrease for a given applied voltage. Thus, if the EHD power is only from electric
conduction, then the EHD power should lie within the analytical results shown in Figure 7. Most

VOLUME 8, NUMBER 4, OCTOBER 2002
349
Figure 8. Average Heat Transfer and EHD Power Versus Mass Flux for R-134a
in Smooth Tube at T = 25°C
sat
of the data in Figure 7 follow this trend. The presence of the electric charge density (from impu-
rities and/or interfacial gradients) would increase EHD power further. In Figure 7, EHD power
consumption for each test section is theoretically presented using two different equations for the
electric field. Specifically, Equation (6), which is based on the electric field calculated by
assuming that only the liquid phase is present around the energized electrode, corresponds to the
lowest EHD power consumption. On the other hand, using Equation (7) results in the highest
EHD power consumption as long as electric conduction is the only component of the electric
current. At low quality values, Equation (8) is used instead of Equation (7) because the low
quality corresponds to a more homogeneous mixture of liquid and vapor surrounding the elec-
trode than the vapor alone.
The increase in mass flux corresponds to an increase in flow momentum, which would cause
EHD power to increase toward the curve defined by Equation (8) at low qualities and Equation
(7) at qualities greater than 0.05. This is mainly because of the change in the liquid/vapor distri-
bution with the increase in flow momentum (e.g., liquid being forced to the tube wall as flow
momentum increases). Additionally, if there is a small electric charge density present, then it
will contribute to the total EHD power as mass flux increases because of the convective current
term [i.e., the second term on the right-hand side of Equation (4)]. Note that the fluid velocity in
Equation (4) contributing to the current density is primarily in the radial direction for the elec-
trode design of this study. However, an increase in the axial fluid velocity also corresponds to an
increase in the radial velocity. Experiments were conducted to validate these statements. The
average heat transfer coefficient and the EHD power consumption, both versus mass flux, are
shown in Figure 8 for R-134a in the smooth tube at T = 25°C. As indicated in Figure 8, the
sat
quality range was held approximately constant through the range of mass fluxes. Therefore, the
EHD power consumption increases due to the increase in mass flux, which indicates the pres-
ence of some electric charge density.
The EHD power consumption relative to the total heat transferred (Q
/Q ) in all data
EHD total
shown in Figures 7 and 8 was less than 1.5%. The maximum Q
/Q
shown in Figure 7 is
EHD total

350
HVAC&R RESEARCH
1.27% and occurs in the 20 cm section at the lowest quality and heat flux. As the heat flux (heat
transferred) increases, Q decreases as described previously because the amount of vapor is
EHD
increasing. The maximum Q
mass flux of 101.4 kg/(m ·s). As the mass flux increases, Q
/Q
shown in Figure 8 is 0.094% and occurs at the lowest
EHD total
2
/Q
decreases because Q
EHD total total
increases at a greater rate than Q
.
EHD
Transient EHD Enhancement
The application of EHD forces to the convective boiling process enhanced and even sup-
pressed heat transfer when compared to heat transfer without EHD. All the results published in
the literature as well as the data given in this paper so far have been at steady state. Several inter-
esting observations were made when real-time temperature measurements were observed imme-
diately after applying the electric fields. The most profound behavior in the evaporator
temperatures was recorded with R-404A in the smooth tube at T = 25°C, when EHD forces
sat
were applied at 15 kV after the system had reached steady state without EHD. When EHD forces
were applied, the evaporator outlet refrigerant temperature began to rise rapidly. Then, after a
time lag, the same occurred for the evaporator inlet refrigerant temperature. This indicates that
EHD forces were initially enhancing heat transfer upon the application of 15 kV. However, the
enhancement eventually decayed and resulted in suppression, as indicated by a lower saturation
temperature after 50 min. This behavior was also influenced by mass flux, as shown in Figure 9.
2
The highest initial enhancement appears to occur between G of 248.0 and 299.6 kg/(m ·s) and it
2
decreases to the point of suppression upon application of 15 kV at G =102.4 kg/(m ·s), with the
change in quality of approximately 0.60. It is important to note that the change in quality was
measured after the system reached steady state, which occurred after the 80 min mark shown in
Figure 9. The quality and saturation conditions were changing substantially during the time
before the 80 min mark.
The water reservoir temperature, local surface temperatures, refrigerant temperature, refriger-
ant pressure, and the section pressure drop as a function of time for the 30 cm and 50 cm test
2
section locations are shown in Figures 10 and 11, respectively, for G = 299.6 kg/(m ·s). From
these figures, the initial enhancement can easily be detected on each test section location. The
initial enhancement is most pronounced on the top of the tube at each location. The tube surface
temperature at the bottom does not change significantly, and after 3600 s (60 min) there is still a
small enhancement. After 60 min at the top of the tube, suppression has started to occur, with the
greatest occurring in the 30 cm section. There is also suppression starting in the middle of the
tube in the 30 cm section after 60 min. The rapid decrease in the pressure drop in both the 30 and
50 cm test sections is misleading, because the refrigerant saturation pressure is rising rapidly
because of the sudden increase in absorbed energy. This is shown in both Figures 10 and 11
where the pressure-drop decay coincides with the saturation pressure rise. The fluctuation in the
two-phase pressure-drop signal is substantial. This is common for two-phase flow; however, it
increases the difficulty in measuring the pressure drop. The instantaneous pressure-drop mea-
surements in two-phase flow vary significantly because of the flow regime structure and the
application of EHD forces. The following conjectures are presented to explain the enhancement:
• Upon the application of 15 kV, the EHD forces start to influence the liquid-vapor interface.
This influence can result in liquid attraction toward the energized electrode at the center of the
tube and/or instabilities at the liquid-vapor interface. The vapor bubbles, as they are formed,
will also be forced to the tube wall.
• Initially, the instabilities and liquid and vapor motion can lead to heat transfer enhancement.
The instabilities and liquid motion can thin the liquid layer near the wall, promoting greater
evaporation.

VOLUME 8, NUMBER 4, OCTOBER 2002
351
Figure 9. Transient Refrigerant Temperature Response After Application of
15 kV in Smooth Tube with R-404A for Various Mass Fluxes
• After a certain time, the polarization forces can redistribute the liquid and vapor completely,
attracting much of the liquid to the energized electrode and repelling the vapor to the wall. As
more vapor is forced to the tube wall, the wall temperature will increase because of the
increase in the resistance caused by an increasing vapor barrier. This will eventually lead to
suppression if the tube surface is not sufficiently replenished with liquid.
From these results, heat transfer enhancement can be obtained if EHD forces are pulsed
(on/off ) at a certain frequency. For the conditions shown in Figures 10 and 11, the EHD forces
should be on for no more than approximately 200 s. Further research is needed in this area to
determine the effect of electric field pulsation on heat transfer enhancement.
Applicability of EHD to Convective Boiling
Application of EHD based on the extraction phenomenon has been explored by only a limited
number of researchers. Because the extraction phenomenon tends to move the liquid away from
the electrically ground heated evaporator wall, it can result in enhancement as well as suppres-
sion of heat transfer. However, significant enhancements in heat transfer can be achieved with
the EHD extraction phenomenon at low vapor quality and heat flux levels. Regardless of the
enhancement or suppression of heat transfer, the EHD extraction phenomenon will always result
in an increase in pressure drop in the evaporator section.

352
HVAC&R RESEARCH
Figure 10. Transient Water Temperature, Refrigerant Temperature, and
Refrigerant Pressure for All Test Sections, and Transient Surface Temperature
and Pressure-Drop Responses in 30 cm Section After Application of 15 kV in
2
R-404A in Smooth Evaporator Tube at G = 299.6 kg/(m ·s)
Heat transfer enhancements can be further improved and suppressions can be potentially
prevented by proper design of the electrode. Suspending an electrode in the center of the
evaporator along the axis is not necessarily the optimum design. Better results can potentially
be achieved with a hollow electrode (concentric tube design) installed along the tube axis at a
close proximity to the tube wall. However, such an electrode design will likely result in a
higher pressure drop. In microchannels (e.g., for aerospace applications), EHD enhancement
may provide significant heat transfer enhancements in the convective boiling process at much
higher pressure-drop penalties. The EHD extraction phenomenon shows promise for enhanc-
ing the transient heat transfer with relatively long frequencies. However, more work is needed
in this area.
In summary, the application of the EHD extraction phenomenon to convective boiling is not a
simple matter and might be limited to low quality and heat flux levels (especially for HVAC&R
applications) unless innovative electrode designs prove otherwise.

VOLUME 8, NUMBER 4, OCTOBER 2002
353
Figure 11. Transient Water Temperature, Refrigerant Temperature, and
Refrigerant Pressure for All Test Sections, and Transient Surface Temperature and
Pressure-Drop Responses in 50 cm Section After Application of 15 kV in
2
R-404A in Smooth Evaporator Tube at G = 299.6 kg/(m ·s)
CONCLUSIONS
Electrohydrodynamically enhanced convective boiling of alternative refrigerants, based on
the extraction phenomenon, was experimentally investigated. Enhancement as well as suppres-
sion of heat transfer was observed with EHD, depending on flow quality and heat flux levels.
Application of EHD also increased pressure drop. The EHD power consumption relative to the
heat transferred in this study was less than 1.5% and primarily attributed to the electric conduc-
tion mechanism. The transient application of EHD, in the case that resulted in the suppression of
heat transfer under steady-state conditions, provided transient enhancements for convective
boiling of R-404A in a horizontal smooth tube. More work on the transient application of EHD
is needed. Innovative electrode designs may lead to the successful application of the EHD
extraction phenomenon to the convective boiling processes relevant to HVAC&R industries.

354
HVAC&R RESEARCH
ACKNOWLEDGMENTS
The authors gratefully acknowledge the financial support of this research by ASHRAE and
NASA Lyndon B. Johnson Space Center (JSC). The authors also gratefully acknowledge
DuPont for providing the refrigerant used in the experiments and Wolverine Tube, Inc., for pro-
viding the boiling tubes.
NOMENCLATURE
c
D
E
specific heat, J/(kg·K)
boiling tube diameter, m
electric field strength, V/m
ε
electric permittivity, F/m
dielectric constant
p
κ
µ
µ
ρ
ρ
dynamic viscosity, kg/ms
electric mobility
3
f
EHD force density, N/m
e
e
3
2
density, kg/m
G
h
h
I
J
k
mass flux, kg/(m ·s)
3
2
electric charge density, C/m
electric conductivity, S/m
surface tension, N/m
heat transfer coefficient, W/(m ·K)
latent heat of vaporization, J/kg
electric current, A
e
σ
e
fg
σ
st
2
electric current density, A/m
Subscripts
thermal conductivity, W/(m·K)
length of test section and electrode, m
mass flow rate of water, kg/s
mass flow rate of refrigerant, kg/s
Prandtl number
avg
e
E
EHD
G
H
i
in
l
m
o
out
ref
s
sat
σ
average
electric
L
·
m
electrode
electrohydrodynamic
ground
homogeneous
inside
inlet
liquid
mixture
without EHD
outlet
refrigerant
separated
saturation
based on conduction current term
temperature
·
mref
Pr
∆P
Q
pressure drop, kPa
power consumed or heat transferred, W
2
q″
heat flux, kW/m
r
interface radius, m
electrode radius, m
inside tube radius, m
temperature, °C
velocity, m/s
applied voltage, V
quality, kg/kg
i
r
e
r
o
T
u
V
x
∆x
change in quality, kg/kg
e
T
v
Greek Symbols
vapor
α
void fraction
w
wall or surface
REFERENCES
ASHRAE. 2001. 2001 ASHRAE Handbook—Fundamentals. Atlanta: American Society of Heating,
Refrigerating and Air-Conditioning Engineers, Inc.
Bryan, J.E. 1998. Fundamental Study of Electrohydrodynamically Enhanced Pool and Convective Boiling
Heat Transfer. Ph.D. dissertation, Texas A&M University, College Station.
Bryan, J.E. and J. Seyed-Yagoobi. 1997. Influence of Flow Regime, Heat Flux, and Mass Flux on Electro-
hydrodynamically Enhanced Convective Boiling. Proceedings of the ASME Heat Transfer Division,
ASME, New York, pp. 187-196 (also accepted for publication in the ASME Journal of Heat Transfer,
2001).
Bryan, J.E. and J. Seyed-Yagoobi. 2000. Electrohydrodynamically Enhanced Convective Boiling: Rela-
tionship Between Electrohydrodynamic Pressure and Momentum Flux Rate. ASME Journal of Heat
Transfer 122:266-277.
Cotton, J.S., M. Shoukri, J.S. Chang, and T. Smith-Pollard. 2000. Electrohydrodynamic Flow and Con-
vective Boiling Augmentation in Single-Component Horizontal Annular Channels. Proceedings of
ASME–IMECE, Orlando, Florida, HTD-366-4:177-184.

VOLUME 8, NUMBER 4, OCTOBER 2002
355
Eckels, S.J., T.M. Doerr, and M.B. Pate. 1994. In-Tube Heat Transfer and Pressure Drop of R-134a and
Ester Lubricant Mixtures in a Smooth Tube and a Micro-Fin Tube: Part I—Evaporation. ASHRAE
Transactions 100(2):265-282.
Kandlikar, S.G. 1990. A General Correlation for Saturated Two-Phase Flow Boiling Heat Transfer Inside
Horizontal and Vertical Tubes. ASME Journal of Heat Transfer 112:219-228.
Kline, S.J. and F.A. McClintock. 1953. Describing Uncertainties in Single Sample Experiments. Mechani-
cal Engineering 75:3-8.
Melcher, J.R. 1981. Continuum Electromechanics. Cambridge, Massachusetts: MIT.
Norris, C., J.S. Cotton, M. Shoukri, T. Smith-Pollard, and J.S. Chang. 1997. Inlet Quality Effects on Hori-
zontal Convective Boiling Under the Electrohydrodynamic (EHD) Effect. Proceedings of ESA-IEJ
Joint Symposium on Electrostatics 2:76-94.
Ohadi, M.M., M. Salehi, S. Dessiatoun, and A. Singh. 1995. EHD-Enhanced Pool Boiling of R-123 in Par-
allel Plate Configuration—Application to Compact Heat Exchangers. Proceedings of Fourth ASME/
JSME Thermal Engineering Joint Conference, Maui, Hawaii, 2:225-232.
Salehi, M., M.M. Ohadi, and S. Dessiatoun. 1996. Minimization of Pressure Drop for EHD-Enhanced In-
Tube Boiling of R-134a. Advances in Energy Efficiency, Heat/Mass Transfer Enhancement Proceed-
ings of 1996 ASME International Mechanical Engineering Congress and Exposition, HTD-338:25-31.
Seyed-Yagoobi, J. and J.E. Bryan. 1999. Enhancement of Heat Transfer and Mass Transport in Single-
Phase and Two-Phase Flows with Electrohydrodynamics. Advances in Heat Transfer 33:95-186.
Singh, A. 1995. Electrohydrodynamic (EHD) Enhancement of In-Tube Boiling and Condensation of Alter-
native (NON-CFC) Refrigerants. Ph.D. dissertation, University of Maryland, College Park.
Singh, A., M.M. Ohadi, S. Dessiatoun, and W. Chu. 1994. In-Tube Boiling Enhancement of R-123 Using
the EHD Technique. ASHRAE Transactions 100(2):818-825.
Singh, A., M.M. Ohadi, S. Dessiatoun, and M. Salehi. 1995. In-Tube Boiling Enhancement of R-134a
Utilizing the Electric Field Effect. ASME/JSME Thermal Engineering Joint Conference 2:215-223.
Taylor, J.R. 1982. An Introduction to Error Analysis. Mill Valley, California: University Science.
Thom, J.R. 1964. Prediction of Pressure Drop During Forced Circulation Boiling of Water. International
Journal of Heat and Mass Transfer 7:709-724.
Torikoshi, K. and T. Ebisu. 1993. Heat Transfer and Pressure Drop Characteristics of R-134a, R-32, and a
Mixture of R-32/R-134a Inside a Horizontal Tube. ASHRAE Transactions 99(2):90-96.
Wattelet, J.P. 1994. Heat Transfer Flow Regimes of Refrigerants in a Horizontal-Tube Evaporator. Ph.D.
dissertation, University of Illinois, Urbana-Champaign.
Wattelet, J.P., J.C. Chato, A.L. Souza, and B.R. Christoffersen. 1994. Evaporative Characteristics of R-12,
R-134a, and a Mixture at Low Mass Fluxes. ASHRAE Transactions 100(1):603-615.
Yabe, A., T. Taketani, H. Maki, K. Takahshi, and Y. Nakadai. 1992. Experimental Study of Electrohydro-
dynamically Enhanced Evaporator for Nonazeotropic Mixtures. ASHRAE Transactions 98(2):455-461.