Heat Recovery Systems & CHP Vol. 15, No, 5, pp.

389-423, 1995
Pergamon Copyright © 1995.ElsevierScienceLtd
0890-4332(94)00055-7 Printed in Great Britain.All rights reserved
0890-4332/95$9.50+ .00

REVIEW P A P E R

E L E C T R O H Y D R O D Y N A M I C E N H A N C E M E N T OF H E A T
TRANSFER AND FLUID FLOW

P. H. G. ALLEN* and T. G. KARAYIANNISt~

*Thermo-Fluids Engineering Research Centre, City University, London, ECIV OHB, UK; and
tSchool of Engineering Systems and Design, South Bank University, London, SEI 0AA, UK

(Received 1 July 1994)

Abstract--The electrohydrodynamic (EHD) enhancement mechanism is first outlined in this paper. A

comprehensive review of past work on EHD single and two-phase heat transfer, as well as past work in
the related area of EHD-induced flow, is then presented. Correlation attempts are also reviewed. Recent
experimental results for EHD boiling and condensation in single- and multi-tube heat exchangers are
discussed. A description of the possible practical EHD electrode systems for applications in power
production cycles and refrigeration is also presented. The research work needed to clarify outstanding
questions in EHD and encourage its use in practical systems is discussed.

NOTATION

A dimensionless group (see equation (33))

B dimensionless group (see equation (50))
CA,CB, Cc empirical constants
C, specific heat (J/kg K)
d vapour film thickness (m)
D diameter (m) total differential operator
oc charge diffusion coet~cient (mS/s)
E electric field strength (V/m)
El electric influence number (defined in equation (31))
El" electric influence number (defined in equation (34))
f,,f~ defined in equations (47) and (49) (FV2/m2)
f frequency (Hz)
rb buoyancy force (N)
rd dielectrophoretic force on bubbles defined in equation (22) (N)
FD body force comprising second and third term of equation (I) (N/m 3)
FE electrohydrodynamic body force defined in equation (I) (N/m 3)
r: equivalent EHD force (defined in equation (39)) (N/m 3)
g acceleration due to gravity (m/s2)
G dimensionless group (see equation (50))
h surface height (m)
i current (A)
i~g enthalpy of evaporation (J/kg)
Ai" thermal energy term (defined in equation (40)) (J/kg)
J current density (A/m2)
K dimensionless group (defined in equation (42))
characteristic length (m)
L length (m)
IT empirical exponent
Ne dimensionless group (defined in equation (35))
N dimensionless group (defined in equation (52))
Nu Nusselt number (Nu = a l l 2 )
Nu" modified Nusselt number (defined in equation (37))
ANu change in Nu
P pressure (N/m 2)
eo electric power input (W)
Pr Prandtl number (Pr = ~%/2 )
Q thermal throughput (W)
q electric charge density (C/m~)

++Author to whom correspondence should be addressed.

389
390 P . H . G . ALLENand T. G. KARAYIANN1S

AQ change in Q (W)
4 thermal flux density (W/m ~-)
R.r radius (m)
f unit radial vector
Ra' modified Rayleigh number (defined in equation (38))
Re Reynolds number (Re = r / / v )
Re' dimensionless group (defined in equation (44))
Reo dimensionless group (defined in equation (53))
Re'e dimensionless group (defined in equation (45))
l time (s)
T temperature (C)
AT temperature difference, wall-saturated vapour, wall-bulk fluid (K)
F fluid velocity (m/s)
V volume flowrate (m~/s)
VR reactor power (kW)

Greek letters
C(, heat transfer coefficient (W/m2K)
corrected (condensation) heat transfer coefficient (W/m-'K)
f~ fluid coefficient of cubic expansion (I/K)
fluid permittivity (F/m)
N thermal diffusivity (m-'/s)
Ft dynamic viscosity (Ns/m)
0 contact angle (rad)
2 thermal conductivity (W/InK)
2* perturbation wavelength (m)
ion mobility (m-'/Vs)
kinematic viscosity (m2/s)
7I dimensionless group (defined in equation (43))
0 density (kg/m 3)
surface tension (N/m)
fie electrical conductivity (S/m)
"C charge relaxation time (s)
4) velocity (V)

Superscripts
mean, indicates vector quantity

Subscripts
c corona, critical
D dielectrophoretic plus electrostrictive
E due to electric field
g saturated vapour
gw vapour to solid
i inner
1 liquid
lg liquid to vapour
In logarithmic mean
max maximum
min minimum
o outer
0 of free space (permittivity), at zero electric field
w solid surface
wl solid to liquid

1. I N T R O D U C T I O N

Achieving higher heat transfer rates t h r o u g h various e n h a n c e m e n t techniques can result in

substantial energy savings, due b o t h to the increased p e r f o r m a n c e o f e q u i p m e n t a n d the design o f
smaller systems to meet required loads. A recent investigation into heat transfer e n h a n c e m e n t
initiated by the U K Energy Efficiency Office e n u m e r a t e d 20 different e n h a n c e m e n t techniques [1].
In general, heat transfer e n h a n c e m e n t m e t h o d s are categorized as " p a s s i v e " or " a c t i v e " , see
Bergles [2, 3]. The first c a t e g o r y includes the high heat-treated, r o u g h or extended surfaces,
displaced e n h a n c e m e n t devices, additives, etc. The use o f extended surfaces was the preferred
technique in h e a t transfer e n h a n c e m e n t in the past, b o t h in single a n d in t w o - p h a s e systems. In
t w o - p h a s e applications, m a n u f a c t u r e r s have recently i n t r o d u c e d high heat flux surfaces. In boiling
a p p l i c a t i o n s m a n u f a c t u r i n g techniques are e m p l o y e d to (a) d e f o r m the surface so as to create
 Electrohydrodynamicenhancement of heat transfer and fluid flow 391

re-entrant type cavities or (b) to deposit a matrix of metallic particles on the surface. Either gives
a porous layer which enhances boiling by providing sites (the pores) for bubble generation. Bergles
[2, 3] and Yilmaz et al. [4] provide experimental results on the performance of these high heat flux
surfaces in boiling. In condensation applications, in addition to low fin tubes, tubes with sharp
saw-tooth-type fins and trapezoidal-shaped extending fins are examples of enhanced tubes
introduced more recently. The second category of enhanced methods includes techniques such as
surface vibration, fluid vibration, vapour suction and a.c. or d.c. electrostatic fields.
The enhancement of single phase heat transfer processes, especially in gaseous systems, is an area
where researchers and designers are spending a great deal of effort. The need to improve the heat
transfer rates was made imperative by the ever increasing requirement for smaller, more
cost-effective thermal systems. In single phase heat transfer, the boundary layers that form on the
thermally-active surfaces offer a significant resistance to the flow of heat which in gaseous systems
can dominate the resistance offered by the solid walls. Enhancing techniques are therefore employed
to alter the boundary layer structure of the flow and, in laminar or free convective situations, to
increase the flow velocity.
The phase change that occurs during a boiling and a condensation process is generally a very
effective mode of heat transfer. However, there is further need to develop methods of enhancing
the heat transfer rates in both evaporators and condensers. A refrigeration/heat pump system
usually operates with heat sink and heat source at predetermined or not controllable (e.g.
atmospheric) temperatures. The working fluid operating temperature must be significantly lower
than the source and higher than the sink in order for the evaporator and the condenser to be of
reasonable, economic size. In a refrigeration/heat pump system these temperature differences,
which constitute external thermal irreversibility, reduce the coefficient of performance (COP) of the
refrigerator or heat pump. Similar arguments hold for the power producing cycles where the
efficiency of the cycle is reduced, as the greatest possible temperature difference is reduced by the
external thermal irreversibility mentioned above.
The beneficial result of heat transfer enhancement, i.e. the increase in the heat transfer coefficient,
would be:
(a) the reduction in the size of the heat exchangers for given ratings.
(b) the reduction in the temperature difference between the fluids exchanging heat and thus
greater thermal plant efficiency. (Alternatively, the transfer of greater rates of energy through a
given size of heat exchanger while maintaining moderate temperature differences.)
Reducing the temperature differentials at evaporator and condenser becomes particularly
important in the more recent development of power production using ocean thermal energy
conversion (OTEC) plants, geothermal energy plants, solar pond generation plants and similar
plants utilizing the organic Rankine cycle (ORC). In these plants the available temperature
difference (between source and sink) is already small, e.g. in OTEC plants this is the difference
between the temperature of surface (20-30°C) and deep sea water (4-7°C). Plants operating
between such small temperature differences must be significantly larger than high temperature
plants (nuclear or fossil fuel plants) in order to produce similar outputs. In addition, their
performance is more sensitive to the temperature differential at the evaporator-source and the
condenser-sink. A certain degree of superheat must be exceeded in the evaporator for nucleate
boiling to commence. Such plants may, therefore, face additional "start-up" problems if this degree
of superheat is not available. This places an additional limitation on the maximum possible
operating fluid temperature difference and thus reduces the plant efficiency.
The enhancing effect of a strong electric field on heat transfer rates (electrohydrodynamic--
EHD--enhancement) has been known for over 70 yr. Early work concentrated on the enhance-
ment of single-phase convective heat transfer. In the last 30 yr the greater potential of EHD in
enhancing two-phase heat transfer rates has been realized by more industrial and academic
researchers. This had led to the study of coupled electric field and pool boiling in the nucleate,
transitional and film boiling regimes. The effect of the electric field on the peak nucleate and the
minimum film boiling conditions has received particular attention. While work on EHD boiling
proceeded steadily, relatively little was published on EHD condensation until the last decade,
possibly due to the greater practical potential offered by EHD boiling. The present paper reviews
392 P.H.G. ALLENand T. G. KARAYIANNIS

the published work on E H D enhanced single- and two-phase heat transfer and presents the results
of recent work on both. A review of the past research on EHD-induced flow-pumping is also given
for completeness.

2. EHD E N H A N C E M E N T M E C H A N I S M

The physical basis of much electrically-enhanced heat transfer lies in the E H D force, FE per unit
volume, generated by an electric field, strength E, in a fluid of dielectric permittivity E, density Q,
at temperature T. This can be expressed as [5]

where q is the electric charge density in the fluid; q/7, its electrophoretic component, is the Coulomb
force exerted by an electric field upon the free charge in it. The third term comprises the
dielectrophoretic and electrostrictive forces on and within the fluid [6], due to the nature of • and
E and their spatial distribution.
Equation (1) should then be included in the Navier-Stokes equation which, for an incompressible
fluid, is
Dd
p ~ = pC + ~v _ W ' + #V2f, (2)

where D f / D t denotes the substantive acceleration consisting of the local contribution ~#/~t and
the convective contribution (,?'V)f, The vector pg is the gravitational force per unit volume; P is
the local fluid pressure. The term pV2f represents the viscous terms.
For an incompressible fluid, the continuity equation is
V. f = 0 (3)
and the energy equation, neglecting viscous dissipation effects, is
8T a~E 2 (4)
8-7 + e . V T = xV2T + pC---~

(The last term in equation (4) is the ohmic dissipation, which is usually neglected.)
For an ohmic dielectric, the electrostatic equations [7, 8] are
V'EE= q (5)

VxE=0 (6)
c3q
V.or + ~- = 0 (7/

and J the current density is

or = qf + a¢/7+ (f. V) (E/7) - D~Vq. (8)
Equations (2-8) will then formulate electroconvective problems.
In single-phase convective heat transfer, thermal gradients in the fluid produce non-uniformities
in the electrical conductivity (a¢) and the dielectric permittivity (E). If there is a change of a~ and
E in the fluid then free charge can build up, which can be acted upon by the electric field
electrophoretic component of E H D force. The dependence of the free charge generation on the
electrical conductivity and permittivity is given as follows.
Neglecting the movement of q, equation (8) becomes
J = a~/7, (9)
while in the steady state equation (7) becomes
v . J = 0. (10)
 Electrohydrodynarnicenhancementof heat transfer and fluid flow 393

Then
v . J = v.(.og) = g .v,7~ + ,7or-E. (ll)
which, for equation (10), gives
,~V" E = - / 7 . Vae, (12)
while, from equation (5),
E V . E = q - E.VE. (13)
Dividing equation (13) by (12)
E q -E.VE
(14)

Rearranging equation (14) gives

q = R.VE - E g.Vao. (15)

O"e
(Yabe [9] gives
E
q = ---~-'VtTe, (16)
o"e
which neglects the variation of the permittivity.)
Assuming that e and a~ are functions of temperature only, equation (15) gives

q ~V . V T & E - dae
(17)
= dr -
O"e
e.vr~

in a thermal field.
In the case of a.c. fields of frequency f, the magnitude of the fluid charge relaxation time z, where
= E/tre, determines whether free electric charge can build up in the fluid, which can then be acted
upon by the electric field. In convective flows with free charge present, the EHD effect becomes
important when z is of the same order as, or less than, the time needed for the fluid to traverse
the electric field [10].
If there is a spatial change of E then the second term of equation (l) becomes important. The
direction of this component depends on the direction of increase of E. This term can be expanded
as follows:

- ½ E:VE = - 1 e T~ + Vp . (18)
T
For non-polar fluids the first term in the brackets is zero and the equation above reduces to

1E2VE = 12 E2F(0E
~L~0p)T VP 1, (19)

which indicates that the resultant component force is towards the heated surface. The same can
be said in the case of polar liquids, although the resulting equation is more complicated. For
non-polar fluids the Clausius-Mossotti relationship:
I-O~] ( E - E0)(E--2E0)
p = (20)
T 3 '
can be used to simplify the third term of equation (1) to

½V[E2[~p]Tp]=½V[E:(E-
Eo)(;-2Eo).] (21)
(see Yabe [9]). (A similar equation can be derived for polar fluids.) This term of the EHD force
is a gradient and thus cannot cause any vorticity in the fluid and was usually neglected, possibly
394 P.H.G. ALLENand T. G. KARAYIANNIS

for this reason. It has, however, a significant effect at the interface and in two-phase heat transfer
[9].
In two phase heat transfer, namely boiling and condensation, EHD enhancement can be
attributed to the following factors which can act individually or in combination:

(a) Action at liquid/vapour interface

The EHD force acts to destabilize the layer of condensate or vapour forming during
condensation and peak nucleate and film boiling, respectively [7, 11, 12]. Film condensation can
reverse to pseudodropwise condensation and film boiling to nucleate boiling. In general, the
destabilization caused by the electric field results in the reduction of the thermal resistance at the
heat transfer surface.

(b) Action on vapour bubbles

Pohl [6] suggested that the dielectrophoretic force can act on an insulating sphere of radius R
and permittivity c2 in a surrounding medium of permittivity E, producing a force F given by
F d = 2~R3q [(E2 - q )/(E 2 + 2E, )]VE 2. (22)
If q < E~, which is the case of a vapour bubble in a liquid, i.e. in boiling, the bubble is attracted
to regions of lower electric field strength. (If E2 > E, the sphere is attracted to regions of higher
electric field strength.) A common geometry studied is one of a coaxial cylindrical electrode system.
The heat transfer surface is an inner cylinder at radius rj and at earth potential (~, = 0), see Fig. 1.
The outer cylinder, radius r2, is at high voltage 42. Assuming that such a system is two-dimensional,
i.e. a long cylindrical system, the electric field E, at any radius r, is given by
~2
E = - - ?, (23)
r ln(r2~
\r,/
where f is the unit radial vector (positive away from the axis) and

r.'--r, lln( 51 . (24)

L \r,/J
Then
2

L \r,/l
Therefore, the force on the sphere (or a bubble assumed to remain spherical) is

/ / /i
L \r,/j
In experiments with refrigerants the resulting electrodynamic force on the bubbles could be
comparable or significantly higher than the gravitational force. In the experiments of Karayiannis
Electrode at high
voltage (#2)
Earthed heated ., ...... -.
tube [ Iz =o) , ~C"
~ ~ ' ~ * ' ~ r ~ '., Vapourbubble

...7 u0u,0
"-.....--

Fig. I. Heated tube concentric with a cylindricalelectrode.

 Electrohydrodynamic enhancement of heat transfer and fluid flow 395

et al. [13] with R114, the outer cylinder had a diameter of 38 mm and the inner, earthed heat
transfer cylinder 19.1 mm. For 20 kV applied at the outer cylinder and dielectric constants of liquid
and vapour refrigerant of 2.17 and 1.0 respectively, the resulting magnitude of the electrodynamic
force on a bubble located at the heat transfer surface is
Fd= 5.13 × 104R3.
The buoyancy force is given as
Fb = 4 7 z R 3 ( p , _ P2)¢. (27)

For a liquid and vapour density of 1441 and 7.74 kg/m 3 respectively, the force magnitude is
Fb= 5.89 x 104R 3,

which is comparable to the EHD force. In heat transfer from smaller diameter cylinders (or in the
extreme, from wires) the EHD force can predominate, since it is proportional to 1/r 3.
In the experiments of Winer [14], again with R114, the ratio r2/r~ was 50.8 x 10-3/5.5 X 10 -3.
For 20 kV the resulting dielectrophoretic force on the bubble is
Fd = 19.7 x 104R3;

this is approximately 3.3 times higher than the buoyancy force.

The acceleration of the bubble can cause agitation of the neighbouring boundary layer liquid
and thus improve the heat transfer rates. However, the heat transfer rates can be greatly enhanced
if the bubbles are made to adhere and move along the heat transfer surface. In cases where the
tube outer surface is made of fins (low fin tubes), the local variation of the electric field and thus
the local dielectrophoretic force can be different both in magnitude and direction than the force
on a smooth surface. In the case of a low fin tube, the field is progressively weaker from the tip
to the root of the fin. The dielectrophoretic force can then move the generated bubbles to the
spacing between the fins. Columns of vapour can subsequently form which grow until the buoyancy
forces overcome the EHD force [15].
If the charge relaxation time is far greater than the bubble detachment period, the bubbles are
not affected by the electric field [16]. Differences in the charge relaxation time of R123 and R11
(0.89 x l0 -3 and 1.3 s [16]) respectively could account, among others, for the higher enhancement
ratio obtained for R123 in the experiments of Ohadi et al. [17]. It is therefore evident that the charge
relaxation time of fluids can be a very important parameter both in single and two-phase heat
transfer enhancement.

(c) Change in contact angle and surface tension

In the case of nucleate boiling, Rohsenow [18] has shown that the degree of superheat needed
for bubble growth at a given bubble size depends directly on the surface tension of the
liquid-vapour interface. Also, the contact angle, 0, is related to the surface tension by

COS 0 = -O'wl-- - ,O'gw (28)

~Ttg

where trw~, a~w and trig are, respectively, the surface tensions at solid-liquid, vapour-solid and
liquid-vapour interfaces [19]. Both 0 and surface tension have been observed to vary with electric
field strength [12, 20] and this would explain the initiation of ebullition by an electric field at reduced
superheat.

3. SINGLE PHASE HEAT T R A N S F E R

The three major ways in which EHD can enhance single phase convective heat transfer are:
(a) by flow of "corona (or electric) wind",
(b) by electrophoresis, with or without charge injection,
(c) by dielectric phoretic forces,
but (d) EHD-induced flow can be used to improve normal convective effects.
396 P.H.G. ALLENand T. G. KARAYIANNIS

(a) Corona wind

Accounts of this have varied from notes in technical journals, such as "developments to watch:
a noiseless fan, with no moving parts" [21] to a full-scale design study for its application to a power
station dry cooling system [22].
Generally, an array of sharp points or fine wires at high potential faces the surface to be cooled,
at earth potential. Table 1 lists descriptions of arrangements for which numerical values of heat
transfer enhancement are given and/or a practical application specified allowing for the fact that
it is not always easy to distinguish between corona wind in tubular systems and similar geometries
utilizing electrophoresis with charge injection.
Compared with other EHD enhancement phenomena, corona wind has relatively large energy
consumption. For example, Holmes and Basham [22] show optimum performance for a 25 by
25 mm needle array at 51 mm from the heat transfer surface when corona power input is 30% of
thermal throughput. Figure 2, from Moss and Grey [24], shows how an increase in the latter is
related to the former. The "unit gradient" lines indicate "break even" to which the effect tends
with increasing Reynolds number, Re.

(b) Electrophoresis
Electrophoresis would appear to give the best results but its effect depends critically upon the
purity or otherwise of the heat transfer fluid, on whether it is polar or non-polar and upon whether
or not charge is injected into it. Table 2 summarizes experimental work, most of it on annular
arrangements, from fine wires to narrow ducts, some with, and some without, deliberate charge
injection. Relatively large values of qE" in equation (1) can be attained with modest energy
expenditure. An ionization current i (A) from a source in a volume flow rate V (m3/s) gives
i
q =~ (C/m 3) (29)

and electrophoretic force

Ei
qE = --~ (N/m3); (30)

this can give rise to relatively large forces for little current consumption. For example, a gas flow
of 670 × 10 -6 m3/s over an ionizing electrode fed with 50/~A gives a charge density of 0.075 C/m 3
which, acted on by an electric field of strength 1 MV/m causes a force of 75 kN/m 3 which is at least
comparable with the pressure gradient causing flow (which depends on the nature of the duct
cross-section). The generation of electric charge, which occurs when the electrical conductivity
and/or permittivity have a spatial distribution, is described by equation (15).
Special mention should be made of two Russian contributions of the mid-1960s, by Baboi et al.
[59] and Semenov et al. [60] which provide a link between electrophoretic and dielectrophoretic
effects as well as, in ref. [59], with corona wind and between single- and two-phase heat transfer
effects. They reviewed earlier work and reported their own natural convection experiments with
air (70 #m dia. wire along axis of 17 mm dia. tube) and with transformer and castor oils (70/~m
dia. wire 10 mm below and parallel to 4 mm dia. cylindrical electrode). In these, with air, a.c. stress
gave ~E/0t0 increasing with E to a maximum (3.2 with wire temperature 52°C at 3.5 MV/m) then
decreased, even causing inhibition, the maxima increasing and shifting to higher E with increasing
temperature. By contrast, the oils gave unequivocal improvement but, whereas trE/tr o for castor oil
increased continuously with wire temperature, for transformer oil it peaked and, between 208 and
375°C, decreased from 5.8 to 1.9 at E = 39 MV/m. Below 98°C, the curves are concave upwards,
whereas with d.c. stress they tend to saturate, intersecting at ~E/0t0= 3.4 and E = 28.5 MV/m,
clearly showing that dielectrophoresis can give greater enhancement than electrophoresis.

(c) Dielectrophoresis
Dielectrophoretic enhancement was the first to be reported (e.g. by Senftleben [61], Senftleben
and Braun [62]) and, although less effective than (a) or (b), is the simplest to compute [63] and
correlate. It can assist the cooling of high voltage apparatus [64]. Table 3 reviews reported
experimental work that can be classified under this heading.
 Electrohydrodynamic enhancement of heat transfer and fluid flow 397

(d) E H D - i n d u e e d flow
Electrophoretic pumps can be classified as "induction" or "ion drag". The operation of the
former depends on the electrical conductivity of the working fluid providing dissociated positive
and negative ions, which are acted upon by the travelling electrostatic field. This field is set up by
interleaved electrodes fed in pairs from a high voltage three-phase supply. If the free charges follow
the apparent motion of this field, imparting their motion to the adjacent fluid, the pump is said
to be an "attraction" one. Its action is analogous to magnetic linear motion in that the fluid velocity
cannot exceed that of the travelling field. The simplest way of creating the free charges is by means
of a temperature---and hence electrical conductivity--gradient [see equation (16)] perpendicular to
the pumping direction. When heat is transferred from the fluid to the duct walls, the resulting
distribution of conductivity, and hence of charges, favours the attraction mode, but if the heat
transfer is in the opposite direction, the charge distribution gives flow against the field and the mode
of operation is "repulsion" [71]. In either case, the net electrical charge in the fluid is zero, whereas
in the ion drag pump charge is injected and then propelled by the electric field between injector
and collector electrodes [72].
All EHD pumps are characterized by low efficiencies; the highest reported in the work reviewed
here is 10%. However, low efficiency is considered offset by freedom from moving parts.
As has been shown, most recently by Sato et al. [73], dielectrophoretic pumping action in an
isothermal liquid can also provide convective flow to improve heat transfer conveniently and
efficiently. However, the work of Sayed-Yagoobi and collaborators [71, 72, 74-79] has concentrated
on induction ion drag pumps and therefore depends on electrophoresis. Nevertheless, to state, as
does ref. [77], that "the dielectric fluid being pumped must contain free charges" cannot be
accepted. Whether electrophoretic or dielectrophoretic, EHD pumping is possible and recent
experimental work is summarized in Table 4.

4. T W O - P H A S E HEAT TRANSFER

(a) Condensation
Publications on the EHD enhancement of condensation began appearing in 1965 with the work
of Velkoff and Miller [80] and Choi and Reynolds [81] using R113. This literature is reviewed in
Table 5. Several themes can be discerned. First, condensation has been studied on surfaces that
are flat or curved, horizontal or vertical. In all cases, the disposal of the condensate, once formed,
can be a problem. On a vertical surface, as the condensate film thickens with increasing distance
downwards, heat transfer coefficients reduce. In banks of horizontal tubes, the same effect results
from condensate dripping from those above. Smirnov and Lukanov [95] report a > 60% reduction
in heat transfer coefficient between rows numbers one and five for R11 condensing on a bundle
of low fin tubes.
As shown in Fig. 3, even the best EHD-assisted performance falls short of that given by using
a single low fin profile tube. EHD has little effect on low fin tube condensation as it and the Grigorig
effect act in opposition. However, in addition to destabilizing the condensate film, EHD can strip
condensate off the surface and overcome the effects described in the previous paragraph. Figure
4 shows that EHD enhancement of R114 condensing on a smooth, 506 mm long, 19.1 mm o.d.
brass tube with vapour to tube-wall mean temperature difference of 25 K is of much the same
magnitude whether the tube is horizontal or vertical. Values of ~t~/~0 from the measurements of
Cooper [96] are given as a function of voltage applied to a 38 mm i.d. cylindrical copper gauze
electrode surrounding the tube. (Measured mean heat transfer coefficients ~tE at electric field
strength E have been corrected to primed values ~t~ taking into account the increase due solely to
reduced AT [89].) The work described above relied on thinning the condensate film. Heat transfer
enhancement ratios (up to 4.5) were obtained using an electrode geometry that strips the condensate
from the surface [91], see Section 6. Higher heat transfer enhancement ratios---up to 20---were also
reported in refs [11, 83, 85]. Heat transfer coefficients are considerably reduced in the presence of
a non-condensable gas. Bologa et al. [92] reported that the damaging effect of the gas can be
nullified by the application of an electric field.
398 P.H.G. ALLEN a n d T. G . KARAYIANNIS

Table 1. Review of published experimental work on augmented single phase heat transfer using corona wind
Heat transfer (~)
Author(s) fluid(s) System m~, Notes
Holmes and Air Electrically heated >6 Compared with
Basham [22] vertical copper plate. (for 25 mm effect of air jets
plane or finned, facing by 75 mm normal and parallel
array of needle needle to surface (corona
electrodes, array at wind intermediate
51 mm between the two).
spacing). Better results with
finned plate.
Conceptual design
of 2GW system.
Velkoff [23] Air, oxygen, Horizontal isothermal 2 In laminar regime,
nitrogen heated tube, 31.75 mm i.d., (at R e = increased friction
with axial stainless steel 3000 wire factor expressed
corona wire, 127/zm dia., + ve. as increase of
starting lengths 6, 13 and Little effective viscosity
24 dia. difference (doubled at
between i~= I mA,
iv = 0.5 and greater with - v e
1.0 mA). wire).
Moss and Grey Nitrogen Horizontal isothermal 1.35 Part of a more
[24] heated tube, 305 mm long, (at R e = general programme
25.4 mm i.d. No details 1200, + ve of EHD research.
of axial wire. wire). Negative wire gave
inhibition.
Grosu and Air, argon, Heated 0.2 mm dia. nickel ~ 3 (air at Detailed analysis
Bologa [25] carbon dioxide, wire in coaxial high I atm and given for this
helium voltage cylinder. helium at geometry. Effect
37 mm dia. 21 atm). 2.5 increased with
(argon at pressure at ic
1.4 atm). 5 constant.
(carbon dioxide
at I atm).
Robinson [26] Air Heated platinum wire, 4.2 Corona discharge
508 u m dia., at high (with wire at heat transfer
voltage, along axis of 260'C and surface itself.
vertical steel tube, natural
53.2 mm i.d., with and convection).
without forced
convective cooling.
McDermott [27] Air Various arrangements for ~4 Schlieren
zero-field temperatures observation shows
130-590' C. High voltage mechanism. Power
applied with various consumed 3 W
sharp probes. compared with 14W by fan.
Mizushima, Air Vertical copper tubes, 4 (at Little effect at
Ueda, with 42 mm i.d. PVC water R e = 700). R e > 4000.
Matsumoto jacket, cooling hot air 2.5 (at
and Waga [28] passing downwards with R e = 2000).
axial nickel wire 300 #m
o.d. Tubes: 6.4 mm i.d.,
520 mm long and
15 mm i.d., 980ram long.
Windischmann Air Point above plane. 2 Effect decreased
[29] (locally) for temp. >200 K.
Yamaga Air Corona wind focused by -- Performance
and Jido [30] shaped intermediate assessed by tool
potential ring on to wear reduction.
(earthed) cooled object
(cutting tool).
Kibler and Air 10 mm dia. tubular assembly To be used in
Carter [31] incorporating point at arrays.
5 kV spaced 5 mm from
(earthed) grid.
Yabe, Mori Nitrogen at 40 rum dia. wire, 120 mm 12 Detailed
and Hijikata 105 Pa long, at 8.3 kV, 20 rnm below (locally) experimental
[32] 100 mm by 100 mm, flat, heated and theoretical
copper plate. analysis in ref. [33].
Velkoff and Air 152 ,urn dia. wires, spaced - 3.5 Improved heat
Godfrey [34] 12.7 mm parallel to, and transfer if stream
6 mm from, vertical velocity ~ 10 times
electrically heated, free convection
483mm by 251 mm, plate. velocity.
 Electrohydrodynamic enhancement of heat transfer and fluid flow 399

Tada, Takimoto Air Horizontal parallel (I) 3.7 600 < Re < 8000.
and Hayashi uniformly heated plates, Flow patterns
[35] 650 mm long, 225 mm wide, (2) 1.8 visualized and
spaced 30 mm apart with computed.
200pro dia. midplane (3) 2.5
ionizing wires, spaced
40ram at 0-12kV and
either: (I) parallel to
flow or (2) transverse to flow.
Also: (3) single ionizing
wire followed by mid-plane
electrode (with gas-solid
suspensions).
Nelson, Air Horizontal tube, 1590 mm 3.6 Peaks at 2000 < Re
Ohadi, Zia entrance, 1020ram test and <3000 and positive
and Whipple 410 mm exit sections, axial polarity. Pressure
[36] 250/zm dia. ionizing wire. measurements made.
(Also, dual wire arrangement Correlations use
giving slightly non-dimensional
greater effect.) variable, Ne (Section 5).

Secondly, from the earliest work, relationships have been suggested for correlating the results
of different experiments. This subject is dealt with in detail later.

(b) Boiling
By contrast with condensation, boiling appears to have benefited from EHD since 1916, when
a threefold increase in water evaporation rate was claimed by Chubb [97]. However, the present
era of such EHD usage began with the work of Bochirol et al. [98] and Bonjour et al. [99], published
in 1960 and 1962, closely followed by Watson [67], Choi [100], Jalaluddin and Sinha [101] and
Markels and Durfee [20, 102], all except ref. [102] for pool boiling. The work used a wide variety
of fluids and, with subsequent work, is reviewed in Table 6.
To define terms used in Table 6, Fig. 5 shows a typical boiling characteristic curve in terms of
thermal flux per unit area, 0, vs superheat, AT. Starting from AT = 0, the heat transfer process
proceeds along curve AB, by natural or forced convection, to C, where AT is sufficient to initiate
nucleate ebullition, thus giving greatly increased q (at D); q then continues to increase with AT until
(at E) it reaches a value qc known as peak nucleate boiling, critical or maximum, heat flux density
where the coalescence of the bubbles forming at the heat transfer surface into an unstable film
begins to impede energy transfer. This decreases q from E to F through a transition region of partial
film boiling until, at F, a minimum film boiling condition ((/mi,) given by a stable, blanketing,
vapour film. Increasing AT then gives stable film boiling, with increasing q, beyond F. If, from
beyond D, AT is decreased, nucleate boiling continues along DB' towards A, giving "boiling
hysteresis".

AQ/QoXl00

30
24
,
/ D.C.voltage,nitrogengas,
~ n e r w~'e positive. p

~roughput,
Re= Reynolds number
20 ,~o 10,00£

t I
oY , i 1
0 10 20 30 40 50 60 70 80 90
Pe/Qox 100
Fig. 2. Increase AQ in total thermal t h r o u g h p u t as a function of electrical power input P~ [24].
400 P.H.G. ALLEN a n d T. G. KARAYIANNIS

Table 2. Review of published experimental work on augmented single-phase heat transfer due to electrophoresis
Heat transfer D~,Do, L Em.~ {ctE'~
Author(s) fluid(s) (ram) (MV/m) ~ J ~ max Notes
Annular systems
Schmidt and Transformer 10, 18, 500 6.9 4 Forced convection
Leidenfrost oil (horizontal) values overall, water
[37, 38] to oil. (Oil side values
estimated in [391.)
(Relatively small) pressure
drop change measured.
Allen [39] Transformer oil 0.025, 2 2 , - - 8.4 ~ 1.2 Steady d.c. stress
to BS148 (vertical) gave inhibition which
changed transiently
to enhancement when
polarity reversed and
permanently when mixed
with a.c. Results of [37, 38]
analyzed assuming
mixed a.c./d,c, stress.
Mascarenhas, Fatty acids: (vertical) -- Effect deduced from
Mascarenhas, stearic 1.3 heating curves. A.c.
Ferreira De oleic 1.4 much less effect than
Souza and palmitic 1.4 d.c. Joulean heat shown
Rabello [40] to be negligible.
Senftleben Cyclohexane 0.05, 10, 20 2.2 Investigates and
and Schnabel (horizontal) explains d.c. inhibition
[41] due to space charge at
low E and AT in good
insulators.
Care and Transformer 0.025, 25,-- 35 -- Followed up [39] by
Swan [42] oil (horizontal) investigating:
(I) boundaries between
inhibition and
enhancement at much
higher d.c. E, finding
oscillatory regime (2)
transient due to polarity
change as function
of pre-stressing
and relaxation times.
Porter and Transformer 1.58, 31.8, 6.7 1.9 Forced convection.
Poulter [43] oil 2800 Negative voltage gave
(horizontal) greater, 50 Hz zero,
effect. Much greater
pressure drop needed
for same heat
transfer increase
without elect, field.
Newton and Transformer 344, 364, 500 2 2 Forced convection. No
Allen [44] oil (vertical- consistent difference
insulated and between positive and
uninsulated- negative polarity but
heat transfer uninsulated heat transfer
surfaces) surface shows somewhat
greater improvement.
Poulter and Avtur, hexane 3.2/1.6, 12.4, 18.3 ~ 20 Forced convection.
Miller [45] (injection 600 Defines injection
strength (horizontal) strength parameter as
estimated 0.7, q//~E, and correlation
0.05) parameter as ~\'~p,'l t.
Fernandez and Transformer 1.59, 25.4, 6.8 6 Forced convection.
Poulter [46] oil 1000 Greater improvement
using negative polarity.
1.59, 19.1, 7.6 23 Pressure drop increased
1000 by up to 66%. Current
(horizontal) drawn :~ 101~A. Radial flow
pattern observed
Gap between horizontal plates (Heat trans]er downwards)
D#'nen$ions
Mascarenhas Oleic acid 2 mm thick 1.25 1.36 Used steady state
[47] layer, thermal conductivity
apparatus. Joulean heat
shown to be negligible.
 E l e c t r o h y d r o d y n a m i c e n h a n c e m e n t o f heat transfer and fluid flow 401

Borbulya, Transformer (a) Disks 70 mm

Kojuhar and oil dia. 1.5 3
Bologa [48] (b) Cylinders, M mm dia.
by 100 mm long
(spacing unspecified).
Gross and Transformer 50 mm dia. cell. 0.3 -- Study of convection
Porter [49] oil patterns due to d.c.
fields. No a.c. effect.
Turnbull [10] Corn oil 102ram by 76mm ~3 I0 Part of an extensive
aluminium plates study into the instability
separated 25 ram. due to an electric field.
castor oil ~3
Rectangular duct (forced convection)
Porter and Transformer Horizontal, ~ I 1.9 at (Relatively small)
Smith [50] oil 180ram by 17mm Re = 95, friction factor
section, 1230 mm =9.1 increases measured.
long, upper Effect on heat transfer
wall heated. kW/m 2. coefficient decreased with
Re and increased
with heat flux.
Fujino, RI13 Vertical, 120mm 1.5 8 Mechanism of effect
Yokoyama and by 2 mm section, explored by
Mori [511 200 mm high, one shadowgraphs.
wall heated.
Ishiguro, "'Furonsorubu Horizontal, 20 at Heat transfer and
Nagata, Yabe AE", (96% R113, 150 mm by 20 mm Re = 1470, pressure drop
and Nariai 4% ethanol) section, test 3 at increase factors
[521 section, 600 mm Re = 27300, decreased with Re.
follows 1000 mm entry wires at Heat transfer
with 14 off300/zm dia. 10 kV. increases greater
stainless steel wires, than pressure drop
20 mm apart, 5 mm from increases.
wall, both walls heated.
Nuclear reactor applications
Stach [53] Air Reactor 1.2 For average velocity
cooling (f) range 4-23 m/s,
channel: no reactor power (V a)
details given. 200-2000 kW and
voltage (~b) I-3 kV, Nu
increased linearly with
VRfb/f 2. Flow of partially
ionized gas analyzed
(in two dimensions).
Berger and (1) (a) air, Annuli 0.56 Up to 50% enhancement
Derian [54] (b) CO 2 (120- (1)41 mm o.d., 1.5 and 8% inhibition
190"C) at low 28 mm i.d., 485 mm depending on
(up to 2 atm) long. pressure, polarity
pressures. and (HV power)/(gas
(2) CO 2 (35-75'C) at high (2) 42 mm o.d., 5.28 velocity)-'.5.
(up to 30 atm) pressures. 28 mm i.d., 500 mm long.
Berger and Derian [55] More experimental details of ref. [54] above
Miscellaneous systems
Coulson and Liquid Heated tubes, 1.8 Results for d.c.
Porter [56] paraffin 3.2, 4.7, 6.4, 9.5 and (polarity not
12.7mm dia., 152mm specified), effects
long, parallel to 1 mm of a.c. "'much less".
dia. HV electrode.
Schnurmann Perfluoromethylcyclohexane 127/zm dia., ~60 2.4 both Part of study
and Lardge (non-polar) 260 mm long, + ve and including boiling.
[571 heated platinum -- ve. Clarified importance
wire parallel of electrical
to, 27.7, 25.4 nature of fluid.
20% isopropyl and 50.8 mm ~48 8.7 aE/a0 graphs suspect
alcohol in from, 254 mm by electrode as they converge
n -heptane, 50.8 mm plate positive. at origin.
(polar) pure electrode. ~ 60 4.4
n-heptane electrode
(polar) negative.
Cooper and Transformer Facsimile 3.6 2.3 Positive electrode
Allen [58] oil to BS148 shell/tube heat has much less effect.
exchanger with Effect decreases with
array of three Re, very sensitive to
earthed, water- conta_mination by air and
heated tubes, to AT. Circumstances
12.6 mm o.d., spaced giving inhibition found
16.6 mm vertically, but turned to enhancement
between electrodes with low frequency
with 4 mm clearance. alternating polarity.
402 P.H.G. ALLEN and T. G. KARAYIANNIS

Table 3. Review of published experimental work on augmented single-phase heat transfer due to dielectrophoresis
Heat transfer Emax CtE)
Author(s) fluid(s) Di, Do, L MV/m ~oo m~, Notes
Annular systems
Senftleben Argon, oxygen 30 or 50/am, 10.7 1.5 Cylinder in cryostat
and Braun [62] (non-polar), ~ 4 5 mm, 70 mm, (ethyl or oil bath at temp.
carbon horizontal, chloride), between 90 and 400 K.
monoxide Pressure 0.116-0,986 bar.
(slightly A.c. and d.c.
polar), used, no polarity
ethyl chloride effect. Experimental
(polar). data given in full.
Ahsmann and Toluene, 20/am, 40 mm, 9.2 1.5 Cylinder in melting
Kronig [65] n-heptane (pure 70 mm, (carbon ice. High voltage
and commercial) horizontal, tetra- 40 Hz square wave and
carbon tetra- chloride), d.c.
chloride,
ammonia gas.
De Haan [ 6 6 ] n-heptane, 22, 55 and 5.0 1.3 Cylinder in
toluene. 99.5/am, 40mm, (n-heptane). thermostat, 40-80 C.
160mm, Greater variation of
horizontal. ~ and D i than [65].
High voltage as [65].
Watson [67] n-hexane. 100/am, 30 mm, 24.5 2.6 D.c. stress, no
100 mm. polarity effect.
Vertical ducts
Weber and "'Transil" Rogowski 9.6 3.6 60Hz stress.
Halsey [68] insulating (uniform field)
oil electrodes,
25.4 mm dia.,
one heated,
spaced
5,1-20.3 ram.
Savkar [69] Light mineral Duct 762 mm 7.3 2.2 60 Hz stress. One or
insulating oil long, 305 mm both sides heated.
wide, up to Longitudinal roll
12.7 mm thick cells predicted from
(with equal analytical solution
length entry of governing equations,
section). including EHD forces,
confirmed
experimentally.
• E/~t0 decreases with
Re.
Newton and Transformer Facsimile 6 1.25 50 Hz stress.
Allen [44] oil to BSI48 transformer
winding ducts,
12.7mm by 57ram
by 339 mm long,
6.3 mm by 28 mm
by 267 mm long
(with entry
length
sections).
Wang, Collins Transformer As ref. [44]. 6 1.25 Governing equations
and Allen oil to BSI48 solved numerically
[63] to include EHD forces
confirmed by
experiments of ref. [44].
Miscellaneous system
Oliver [70] Sulphur -- 8 Gas pressure l0 bar.
hexafluoride Conductor dia. 33 mm.
No effect observed.

EHD eliminates the q decrease between E and G and, in addition, increases q¢. These phenomena
have received very much more attention than EHD's other action, namely the initiation of nucleate
boiling at smaller AT and hence the virtual elimination of the hysteresis portion (ABCDB'A) of
the characteristic, e.g. by applying volts at AT values between b and c. This initiation process has
been reported only by Karayiannis et al. [13], Jalaluddin and Sinha [101], Basu [104] and Allen
and Cooper [15] who have also made a video record of it [117]. In this, observation is made of
a tube, immersed in R114. Hot water passes along the tube, losing heat to the refrigerant. Thus,
near the inlet, the tube surface gives sufficient superheat for bubbles to evolve at a well-defined set
 Electrohydrodynamic enhancement of heat transfer and fluid flow 403

Table 4. Review of recently published experimental work on EHD pumping

Author(s) Fluid(s) System Flow rate Notes
Induction (attraction)
Seyed-Yagoobi, Sun No. 4 oil Total loop length 5.2 m, 50 mm/s (4 x Heated section of
Chato, Crowiey with added vertical limbs 2.1 m, 16 mm natural return limb gave
and Krein [74] Shell ASA-3 i.d., 3-phase electrode convection natural convection
antistatic system formed by rate) at (forward) flow from
additive to alternate metal and glass optimum frequency top to bottom of
vary a~. rings forming tube ~ I m ( ~ 5 Hz) and pump. Some flow
long. 30 kV 3-phase oil ~c. detected without
supply, 0-100 Hz,
reversible. Natural heating or cooling.
convection generated by
heating or cooling.
Bohinsky and 4 synthetic Total (vertical limb) Maximum Higher than
Seyed-Yagoobi petroleum loop length 2.2 m, 50 mm 385 mm/s synchronous
[75] fluids, 4 i.d., 430 mm long 3-phase obtained velocities observed
petroleum oils, electrode system formed with doped but not explained.
I polyol ester, by 18 gauge copper wire aromatic
I fluorinated in grooves. 12 kV 3-phase hydrocarbon
fluid, I supply, 0-13 Hz, sine, (n-hexane)
aromatic square or triangular at 9 Hz.
hydrocarbon, waveform.
I silicone
fluid, some
with added
over-based
calcium
sulphonat¢ to
vary a c.
Ion-drag
Castafleda RI I Total (horizontal) loop Maximum Current consumption
and Seyed- length 2.68 m, 50 mm i.d., 240 mm/s and velocity linear
Yagoobi [78] pumping section 360 mm with all functions of number
long. 13 electrode pairs, electrodes of energized
8 mm apart, spaced 15 mm, energized electrodes.
formed by 20 gauge tin- at 28.5 kV
copper wire in grooves,
using no sharp points.
50 kV supply.
Bryan and 3 synthetic Total (vertical) loop 220 mm/s Efficiency (;~ 5.5%)
Seyed- petroleum length 1.92 m, 70mm i.d., obtained increased with
Yagoobi [77] fluids, 4 pumping section 330 mm with charge relaxation
petroleum oils, long. 10 electrode pairs, undoped time.
I fluorinated 10 mm apart, aluminium synthetic
fluid, I electrodes spaced 5 mm, petroleum
aromatic upper (emitter) electrode fluid
hydrocarbon, fitted with 90 stainless (dodecyl-
1 silicone steel needles, pointing benzine)
fluid, at collector electrode, at 20 kV.
50 kV supply.
Bryan and Dodecylbenzine Two pump sections: 334 mm/s Increasing ¢¢
Seyed-Yagoobi (a) as ref. [77] but with with reduced flow rate.
[79] total loop length 1.86 m. undoped Higher efficiency
(b) with one electrode fluid and and slightly higher
pair only, giving loop 10 electrode velocity with plain
length 1.33 m and also pairs collector electrode.
having modified collector at 25 kV.
electrode (with stainless
steel wire web extending
6 mm radially inward).
50 kV supply.
Dielectrophoretic
Sato, Yabe and "Furonsorubu 24 mm square, about 160 mm 143 mm/s Effect of flow on
Tak¢tani [73] AE" (96% R II3, long, horizontal pump with convective heat
4% ethanol), section containing 61 mm toroidal transfer estimated.
R I 13, R 123. long, 20 mm square hollow electrode Other (less
high voltage electrode at 20 kV. effective) electrode
adjacent to various arrangements tested.
(toroidal, rod,
spherical) earthed
electrodes.

of sites. However, towards the outlet, AT is only sufficient to give convective heat transfer and thus
a well defined "front" is seen between boiling and non-boiling regions. On application of electrode
voltages greater than about 8 kV there is immediate--and more vigorous--ebullition everywhere
on the tube surface. In at least one case (Baboi et al. [111]), the phenomenon was deliberately
HRS I $/~--B
404 P.H.G. ALLEN and T. G. KARAYIANNIS

Table 5. Review of published experimental work on EHD augmented condensation heat transfer

Author(s) fluid(s) System m~ Notes

Velkoff and R I 13 Vertical cooled cooper 3, 4, 5, 6
Miller [80] plate, 152 mm high,
229 mm wide, facing:
(1) 102,am dia.
horizontal wire ~ 1.4
(2) Mesh screens ~3
(3) Aluminium plate. ~ 1.7

Choi and R113 Vertical stainless 2 1, 2, 3

Reynolds [81] steel tube, 1346 mm
long, 23.8 mm i.d.,
cooled on outside and
coaxial with inner
electrodes 6.3, 12.7
and 19.1 mm o.d..

Choi [821 As for Choi and Reynolds [81].

Holmes and R114 Flat silver-plated Up to 6, 7

Chapman [83] copper cooled and times
electrode plates (each 10
152 mm--horizontal--by (overall).
38 mm--sloping)
at different angles to
give variable wedge
shaped enclosure.

Seth and R113 Horizontal copper water 1,6 5, 8

Lee [84] tube with coaxial outer
electrode.

Bologa and R113, hexane, Vertical copper plate, 10 (for 2

Didkovskiy diethyl ether 220mm by 120mm, spaced RI13 &
[ill 7 mm from parallel (plain hexane)
or horizontally slotted) 20 (for
electrode, ether).

Didkovskii As for ref. [11]. As for ref. [11] together with As for ref. [11]. 2,5,6
and Bologa vertical annulus,
[851 cooled tube 300 mm high
21 mm o.d., electrode
tube 35 mm i.d.

Didkovsky and As for Bologa and Didkovskiy [11]. 2,3,7,9

Bologa [86]

Smirnov R113, Vertical annulus, cooled 3.6 2, 3, 6, 7

and Lunev diethylether outer jacket, 215 mm high
[871 30 mm i.d., around
central electrode with
smooth, screw-threaded
or perforated surface,
gaps 2.5 to 4ram.

Yabe, Kikuchi, (a) Water Needle electrode above 10

Taketani, Mori liquid surface.
and Hijikata
1881 (b) R113 Cooled brass plate, 2.24 3, 11
600 mm high, 100 mm wide
with shaped copper wire
electrode systems.

Cooper and R12 and R114 Single cooled horizontal 2.9 3, 5, 7, 12, 13
Allen [89] tubes 514 mm long with
smooth (19.1 mm o.d.) and
low fin (19.0 mm o.d.)
surfaces. Electrode
system: (1) concentric
copper gauze (38 mm i.d.)
(also vertical with
R114) (2) Parallel brass
plates with auxiliary
6.25 mm o.d. rod (to
simulate (I)).

Trommelmans RI I, R113 Horizontal smooth copper 1.1 3, 14

and Berghmans and R114 cooled tube 1100ram long,
[901 18 mm o.d., concentric
with copper wire spiral
electrode, 38 mm i.d.
 E l e c t r o h y d r o d y n a m i c enhancement o f heat transfer and fluid flow 405

Yabe, Taketani (a) Silicone Thin liquid film on and 3,5

Kikuchi, Mori oil between electrically
and Maki [91] conducting glass sheets.
(b) R l l 3 Vertical smooth brass 4.5 12
cooled tube 540 mm high,
18 mm o.d. with copper
helical (condensate
removal) wire. 0.5 mm
o.d. and curved (stress
applying) plate
electrodes, 21.2 mm o.d.
Bologa, R113-helium Flat plate 200 mm high 1.8
Savin and Rll3-air 120mm wide with parallel 2.6
Didkovsky [92] R ll3-carbon plate electrode (having 3.0 2,3,4,5,8
dioxide slits) or wire electrode
Hexane-air (0.18 mm wires spaced 5 mm 5.0
apart).
Damianidis R114 Horizontal bundle of nine 1.08
Collins, smooth surface tubes,
Karayiannis 19.1 mm o.d.+ 360mm long.
and Allen Perforated plate and
[93] rod electrode.
Yamashita, R114, Verticle bundle of 102 6 2,3,5,15,16
Kumagai, n-perfluorohexane smooth surface stainless 4
Sekita, steel tubes, 19.05 mm
Yabe, o.d. with combined
Taketani, lattice/helical wire
and Kikuchi electrode system.
[94] Condensing temp.: 90 C
(RII4) and 150C
(n-perfluorohexane).
Notes2
(1) Analysis in terms of perturbation wavelength.
(2) Results correlated.
(3) D.c. electric field.
(4) Corona discharge gave electrophoresis.
(5) Condensate film perturbation observed.
(6) Condensate spraying observed.
(7) A.c. electric field.
(8) Inhibiting effect of non-condensable gas partly neutralized by EHD.
(9) Condensate film thickness measured.
(10) Gas-liquid interface stability studied theoretically and experimentally.
(I I) Improved heat transfer due to condensate removal by electrodes measured.
(12) ~E Values corrected for reduced AT.
(13) EHD effect with low fin tube <10%.
(14) Effect of EHD on shape of pendant condensate droplets discussed.
(15) Tube length effect studied.
(16) Electrode power consumption increased by oil contamination.

avoided by reducing the thermal loading to the experimental value required (i.e. along EDB' in
Fig. 5) "so that most of the vapour-formation centres were active". Also, as remarked by Markels
and Durfee [20], "It appears likely however that differences exist between voltage effects when a
relatively large diameter heat transfer surface is used and when the boiling takes place on a thin
wire". A fine wire fed by a constant electric current cannot instantaneously provide the additional
thermal energy demanded by suddenly promoted ebullition. The energy could only come from a
considerable decrease in temperature, and hence in AT, thus suppressing ebullition before it became
established. By contrast, the more usual industrial case of a source heated by fluid flow maintains
a much more nearly constant temperature, as can a high thermal capacity source (see Table 6
notes).
EHD boiling heat transfer investigations included tests with a variety of tubular surfaces
including smooth, low fin and enhanced surfaces such as Thermoexel. It was found that the nucleate
boiling heat transfer enhancement obtained on a smooth tube was very moderate compared with
the low fin and the Thermoexel tubes. The most recent work in this area (based on data from
refs [13, 96, 118]) is summarized in Fig. 6 on a "tube rating" (kW per metre of tube vs AT) basis
which, for a given tube diameter, governs heat exchanger design.
Although, as shown in Table 6, a variety of geometric configurations was examined, only recently
have researchers concentrated their efforts on tube bundles in heat exchangers like the shell and
tube. Damianidis et al. [! 16] reported experiments on nine tube bundles consisting of either smooth
or low fin tubes. The working fluid was R114. In the case of the low fin tube bundle, the
406 P.H.G. ALLEN and T. G. KARAYIANNIS

Table 6. Review of published experimental work on EHD augmented pool boiling heat transfer
Heat transfer Healedsurface (~q~.E (ate)
Aatho.) diam0ter,mm) . . . . . . .- , . . Notes
Coaxial eylindriealsystems
Bochirol, Benzene,
Bonjour and hexane, 0.2 2.7 3,6
Weil [981 trichlorethylene, (trichlorethylene
toluene, ethyl ether, and
liquid nitrogen ethyl ether)
Pure water,
methyl ethyl ketone, 0.1 6
acetone, methyl alcohol (methyl alcohol)
Bonjour, As for Bochirol, Bonjour and Weil [98]
Verdier and Weil [99]
Watson [67] n-hexane 0.1 2.6 2,6
Choi [100] R113 0.254 or 0.508 >2 1, 4, 6
Markels and isopropanol, 9.525 6 2, 5, 7
Durfee [20] distilled water >1.4
Winer [14] R114 5.57 1.6 1,2, 6
Lovenguth R21, R113, 0.51 2.9 I, 2, 6
and carbon tetrachloride,
Hanesian [103] chloroform
Basu [104] Carbon tetrachloride 0.040 2, 3, 6, 8
Jones and R I 13 0.127 or 0.406 1,4,6
Schaeffer [105]
Jones and R113, 0.812, 1.15 4,5,6
Hanock [106] R113/methanol or 1.83
Zheltukhin, Acetone, 5.0 to 6.5 1.82 5, 6
Solomatnikov, benzene, (acetone)
Mikhaylov and n-diethyl
Usmanov [107] ether
Allen and R114 18.6 low up to 60 2,7,8
Cooper [151 fin o.d.
Kawahira, R 11 22.4 4 2,3,6,8
Kubo and 10,11
Yokoyama [108]
Karayiannis, R114 19.1 4 2,7,8
Collins and Allen [13]
Ohadi, RI23, R123/oil 12.7 o.d. 5.5 (R123) I1
Papar, Ng, RI I, RI I/oil 1.7 (RIO
Faani and
Radermacher [17]
Parallel plate systems Electrode
gap (ram)
Olinger and Deionized 6.35 or 25.4 2,3,6,9, 10
Colver [12] water
Blachowicz, Benzene 5.0 to 20.0 2 3, 7, 9
Brooks and
Tan [1091
Kawahira RII 3.0 4.4 2, 3, 6, 10
Kubo and II
Yokoyama [108]
Other systems System(s)
Jalaluddin Methanol, Heated sphere 2, 3, 6, 8, 9
and Sinha isopropanol, and plane
[101] methyl ethyl electrode.
ketone, benzene
Asch [1101 RII3 Heated wire 2,3,6,11
(0.356 mm dia.)
between 25.4 mm dia.
discs
spaced 25.4 mm
or (asymmetrically)
12.7 mm alia.
spheres with
60.3 mm gap.
Baboi, Benzene, Heated wires 1,2,3,6,11
Bologa and toluene, 0.05, 0.07
Klyukanov liquid argon 0.10 and 0.50ram dia.
11111 parallel to 6.0 mm dia.
cylindrical electrode.
 E l e c t r o h y d r o d y n a m i c e n h a n c e m e n t o f heat transfer a n d fluid flow 407

Zhorzholiani Acteone, Heated 3 1,2,3,6,12

and benzene horizontal (acetone)
Shekriladze [I 12] n-pentane tubes, 6.0 and
8.0 mm dia., parallel
to plane electrode.
Vertical
semi-cylindrical
heated channel
facing plane electrode.
Rutkowski [113] Liquid nitrogen Heated horizontal 3,6
wire, 0.10ram dia.,
parallel to 3.0 mm
cylindrical electrode.
Yabe and R113/ethanol Heated 2. I 2,7,8,13
Maki [114] ("Furonsuburu plane beneath
AE') toroidal EHD
"'pump" electrode.
Schnurmann n-heptane, Heated ~7 2,6, II
and Lardge 20% solution earthed wire ~7
[57] of isopropyl 0.127 mm dia.
alcohol in and 260.0 mm
n-heptane, long electrode
perfluoromethylcyclohexane between ~2
two plates
254.0 x 508.0 mm
one earthed and
one at high potential.
Tube bundles
Damianidis, R114 Low fin tubes, 2.5 2,7,8,11
Karayiannis, dia. 19.05 mm,
AI-Dadah, fin pitch 0.7 ram.
James, Perforated plate
Collins and rod
and Allen [115] electrode.
Ogata, RI23 5-tube smooth 7 2,6
Iwafuji, tube bundle,
Shimada and 22 mm dia. 0.5 m
Yamazaki long. Steel
[16] rod electrodes
6 mm dia. 3 mm
from tubes.
50-tube smooth 7 2,6
tube bundle,
22 mm dia. 0.5 m
long. 12 electrode
wires 3 mm from tubes.
80-tube smooth 2,7
tube bundle.
22 mm dia. 2 m
long. Steel
rod electrodes
6 mm dia. 3 mm
from tubes.
Damianidis, Rll4 9 low fin 2.5 2,7,8
Karayiannis, tubes, fin
Collins and root dia. 17.6mm,
Allen fin height
[116] 0.5 mm, fin
pitch 0.88 mm.
9 smooth tubes 1.3 2,7,8
o.d. 19.1 mm.
Perforated brass plate
and rod electrode
Notes:
(I) Dieleetrophoresis analyzed.
(2) D.c. electric field.
(3) A.c. electric field.
(4) Effect of EHD on minimum film boiling.
(5) A.c. electric field frequency varied.
(6) Electrically heated source.
(7) Fluid heatcd source.
(8) Ebullition initiated by EHD.
(9) High thermal capacity source.
(10) Fine wire mesh or array high voltage electrode.
(I I ) Observations of bubble dynamics.
(12) Correlation suggested.
(13) Enhancement by liquid jet (but bubble dynamics also affected).
408 P. H. G. ALLENand T. G. KARAYIANNIS

O~'( kW/mZK)

I B'=rct°"~'~'="~z [ i [

L L i

1 i
1 3 6 10 30
(T,-To) K
Fig. 3. EHD enhancement of R12 condensation with electrode voltages 20, 25, and 30 kV at outer surface
of smooth horizontal tube compared with END-free smooth and low fin surface tubes (T s = 30°C) [96].

enhancement was significant, see Fig. 7. No enhancement was obtained for the smooth tube bundle
except in the convective regime. This was attributed to the electric field distribution. The electrode
geometry used in ref. [115] resulted in radial electric field with the resultant E H D force pushing
the bubbles away from the heated tube surface instead of on to the surface. This was not the case
in the work reported by Ogata et al. [16]. They tested three different tube bundles boiling R123
and obtained significant E H D improvement in the heat transfer coefficient (seven times at 18 kV).
Ogata et al. used steel rods and wires placed circumferentially and 3 m m away from the tubes as
the electrodes. The non-uniformity of the resulting electric field moved the bubbles towards the
tube surface and hence enhanced the heat transfer. The bubble movement in a low fin geometry,
which has a beneficial effect on the heat transfer surface, was discussed earlier. Both research groups
reported that E H D enhancement of boiling heat transfer in bundles is as effective as on a single
tube.
E H D boiling inside tubes was studied by Yabe et al. [119] using a 1 0 m m inside diameter
cupronickel tube and a coaxial cylindrical perforated steel electrode of 5 m m diameter. The E H D
behaviour of the non-azeotropic mixture of R123 and R134a was studied. An enhancement of
almost three times was obtained in the heat transfer coefficient. A change in the flow pattern from
stratified to annular was also observed. The enhancement ratio decreased as the mass flux through
the tube increased. The researchers also verified that for a given thermal duty, the pressure drop
across the evaporator is the same, with or without EHD.

t
ag 3.0
a 0
2.5

2.0

i I
1.5

.--j Tube horizontal

[] Tube vertical

1"015 20 25 30 35 40
Electrode voltage (kV)
Fig. 4. EHD enhancement of RI14 condensation at outer surfaces of vertical and horizontal tubes
(AT = 25 K, TO= 90'C) [961.
 Electrohydrodynamic enhancement of heat transfer and fluid flow 409

G
Heat
flux
4z
(W/m)

Critical
heat flux~, E

/
Nucleate
b:i/:;ag:i°r I / Transition
boiling
',.
--~,,
/~

''. J b o i l l n g

b c ~T : (Tw -TO ) (K)

Fig. 5. The boiling curve showing heat flux density vs degree o f AT.

5. E H D E N H A N C E M E N T CORRELATIONS

Correlation of heat transfer data in terms of non-dimensional quantities for specific geometries
has long proved a valuable tool for heat exchanger design.

(a) Single phase

Correlations of experimental EHD results for systems with geometric similarity in terms of
non-dimensional variables have been used for over 40yr, since Kronig and Schwartz [120]
postulated the superposition of the effects of normal natural convection and those of dielec-
trophoresis, expressed as Nusselt number, Nu, and its change, ANu, respectively. ANu gave a linear

31111 IIIII
,ooo !!!!l I ,1 / t~ ~
!!!!!
~1111
I lit o'o...J i r, tllll
till " _~'k 1 0 I/liJ

..... : Then'noexc~HE (20kV)

xxxxx ThermoexceI-C (20kV)
I.ow-IIn (27kV)
100 I III
. . . . • . . [. i.l i l l . ::;;; Sm~lh (2CkV)
I I IN I .J J Jill 0600@ Themoexce-C (27kV)
I Ill ~ 71 /t IIII a ,,, ,,,,,,,,,, ~ E (OkV)
5'''' / ""/ ', Ill 04004 "rhe~moex=ot-C ( Ok~
41111 / I I I II .,,tree Low-fin (OkV)
i. ~ i, ~,~ Smooth (OW)
31111 " IIIII
Illl " IlllJ
5678~ 2 3 4 5678 2
1 10
S u p e r h e a t AT (K)

Fig, 6. Single tube rating vs degree of AT for various tube surfaces and R114 (Tube o,d. ~ 19 ram).
410 P.H.G. ALLEN and T. G. KARAYIANNIS

r-rr-r~ 0 kV heaUng
:;::: 0 W
•'~AA*~ 10 kV
00~ 10 kV ¢oolin9
~ kV
kV
heotJng
heating/cooling
5-

i*O!'--
+ i

l l * ~
Ill*'ilP
~ 4ffJ II I
~A IP

J
2 3 4 5 6 7 8 910
Superheat txT (K)
Fig. 7. EHD heat transfer enhancement in a nine low fin tube evaporator with RII4 (Ts= 30°C).

log-log plot against El, an electric influence number, multiplied by the fluid Prandtl number, Pr.
El is defined as
dE
P ~-~ f2ATE2
El = /a2 (31)

and is closely related to the Grashof number, Gr, used to correlate natural convective heat transfer.
In fact
~E2 d~ 1
El = Gr L~- ~~-] (32)
p/JgE
El = Gr A (33)
say. A later modification due to Senftleben and Biiltmann [121] proposes
Gr Pr A (1 + 0.00008A )
as the correlation variable. The correction to El is, however, small and usually of the same order
as, or smaller than, the experimental error.
To allow for the effect of fluid electrical conductivity on electrophoresis, Turnbull [10] proposed
an alternative electric influence number:
1 da e
pE ~ -dT :2ATE2
El' = #2 (34)

He warned, however that successful correlation with respect to either El or El" is not a sufficient
criterion to distinguish between electrophoretic and dielectrophoretic effects; this is best done by
comparing a.c. and d.c. results.
Figure 8 shows results from ref. [45] for "Avtur" and from ref. [122] for transformer oil,
correlated in terms of Nu, Pr and a voltage parameter,
 Electrohydrodynamicenhancementof heat transfer and fluid flow 411

_.!
NuPr ~

10

6 . j , ¢ " -
•/ "
Y Av'mtionkerosine (AVTUR)
500 "Re -2500
Constant heat flux

• ..~ ~ Transformer oil

I00 "Re ,=1600
Constant wall temperature of 100"C

'to' 6 81b 3 2 " /. 6 8 4 o "

Voltage parameter ~.Ev'~
P
Fig. 8. Correlationof EHD singlephaseheat transferfor aviation kerosineand transformeroil [45].

To correlate corona wind results, Davidson et al. [123] have suggested using a dimensionless electric
number:

Ne=~-Dpf 2, (35)

where i is the ionization current, r/is the ion mobility, D is the tube i.d. and p and f are fluid density
and mean velocity, respectively, and this has been used successfully by Nelson et al. [36].

(b) Condensation
In the case of EHD-assisted condensation, correlation has been attempted from the earliest work.
Having analyzed the disruption of the condensate film in terms of a perturbation wavelength 2",
Choi and Reynolds [81] and later Choi [82] used this as the characteristic length in the Nusselt
number, Nu', and modified Rayleigh number, Ra', for the correlation
Nu' = CA (Ra') °'25, (36)

where CA is an empirical constant and

Nu" = ct - - (37)
2
Ra' = F~P(~'*)3Ai'
2# AT,, ' (38)

where 2 and # are fluid thermal conductivity and viscosity, respectively. ATe° is the logarithmic
mean value at AT,

F¢ = ~ 1-~) EgL,J, (39)

a being liquid surface tension, Eg and E~ permittivities of vapour and liquid, respectively, Eg the
electric field strength in the vapour and
Ai' = rig
• + epgAT + g3 CplA T1n, (40)

HRS 1 5 / ~
412 P.H.G. ALLEN and T. G. KARAYIANNIS

where i~g is the enthalpy of evaporation and cpg and cpt are specific heats of vapour and liquid,
respectively. By contrast, Bologa and Didkovskiy [l l] in their earlier work used the simpler
relationship:
~E E 2.5
o 8[ cl ,
where Ec is the electric field strength at which the EHD effect becomes significant (Ec = 1.3 MV/m).
They also reported greater enhancement for polar fluids (diethyl ether) than non-polar (hexane,
R113). They attributed this to the presence of free charges and the promotion of the electrophoretic
force. However, in later work [85] they used the relatively complex correlation
FK'.:~o.:]
lo,L j,
where

K = %E~[ (42)
Gig
(the measure of the electric forces to the surface tension ratio),
[2
= -- (43)
zv
(the measure of the time ratio between damping of mechanical disturbances and relaxation time)
and

Re'= oh (44)
ilgpv '
the characteristic length [ being the annular gap, while h is the surface height and z the charge
relaxation time for the liquid. Subsequently [86], they correlated their own results together with
those of Velkoff and Miller [80] and Smirnov and Lunev [87] in this way, but with the index of
n decreased to 0.175, see Fig. 9. Smirnov and Lunev [87] themselves proposed the correlation of
NUE - C8 Nuo .

LDoJ

3 I I I I Iii i /

. • . m~,~, _

v 7

5 , R113
•% I Z ~ o o Diethyl ether
• . ~ o • _n-hexane
•~ o • D a t a of [80]
3

• I 1 I I I
? ~oz 2 3 S ? IO 3

K~.= lC0.17s
1
(Re/) -~
Fig. 9. Correlation of E H D condensation heat transfer [86].
 Electrohydrodynamic enhancement of heat transfer and fluid flow 413

Nusselt numbers using heat transfer surface height as characteristic vs Re~, where
Re'E= h (,, -___¢g)Ev,,~/P/'° [ I - ( E c DgDo
i ~ 2/] ._]
\ (45)

In spite of these later refinements, Cooper [96] found the original [8 l, 82] correlation valid for his
own results, as well as for those of Didkovsky and Bologa [86]. With CA = 0.35 (rather than 0.5
as originally proposed) the scatter is about _+30% for F¢ > 60,000, which is no worse than with
the correlations suggested in refs [86] and [87].

(c) Boiling
In boiling, the effect of the electric field upon the maximum (qc) and minimum (point F in Fig.
5) heat fluxes and the correlation of this effect has received particular attention. In the case of qc
this is important since it could represent the maximum allowable heat transfer. Increasing the heat
flux to a value greater than q~ gives rise to very high AT that can lead to burn-out. Johnson [124]
considered the influence of the electric field on hydrodynamic stability. He amended the
hydrodynamic theory of boiling heat transfer to include the effect of a perpendicular electric field
across the vapour-liquid interface that exists at the maximum and minimum heat flux and
correlated the ratio (q¢)E/(qc)0 as a function of fluid properties and electric field strength. One
resulting equation for maximum heat flux condition was

(c)¢)E
('c)0 ~ [l+~.~3(p'~f~p')ga-]]'/2
x/ --pg)ga J
= --3(p, - - - - - - ' (46)

where

f ' = (%)
~ [E'c ,-+ qq]2 E2" (47)

A similar relationshi p was recommended for the minimum heat flux.

r~+ 2(Pl-~--Pg)ga ]1/2
(#min)E 13 f2 + x/f~ + 3 ( p , - pg)ga.J (48)
(qmin)0 [4 (p,_ pg)ga],/4 ,
where
El) [El -- Cg]2 E2. (49)

These relationships were compared both in Johnson [124] and Winer [14] with the limited
experimental results of Winer (for a radial field a cylindrical heat transfer surface and R114) and
found to be in good agreement, Fig. 10. In addition, Johnson stated that electric fields of the order
of ,-~ 1 MV/m are required for any significant effect.
Jones and Schaeffer [105] modified the Zuber hypothesis [125, 126] for the minimum film boiling
point to account for a strong E H D coupling at the vapour-liquid interface forming on a fine wire.
They identified three electroquasistatic regimes based on the relation of the electric field frequency
and the fluid charge relaxation time. The first regime was that of an insulating liquid where the
frequency is much greater than the inverse charge relaxation time. Following the intermediate
regime the third regime refers to a sufficiently conductive fluid. Jones and Schaeffer obtained
relationships similar to those above for insulating and conducting fluids. Their theoretical results
for (qmin)E/(qmin)0 were somewhat below the experimental results for R113 boiling on a fine wire.
Berghmans [127] considered a flat heater located in a conducting liquid and performed a stability
analysis including the effect of a uniform d.c. field. This analysis resulted in
q¢ m hi/21 ~ I (1)1/4
~'g ' l g ~ ~
rG 2 (G3--ff~+l
(aplg)l/4L-~+ 4 )1/211/2, (50)
414 P.H.G. ALLENand T. G. KARAYIANNIS

((1'~), 7

6'

0
2 4 6 8
1.00E+06 1.00E+07
E (Vim)
Fig. 10. Comparison of the predicted effect of an electric field on maximum heat flux with experimental
results [124].

where G 2 EgE2d/6 is the ratio of the electric forces to the surface tension forces and B 2 = p~gd2/a,
=

B 2 being the Bond number which is the ratio of inertia to surface tension forces; d is the vapour
film thickness which can be obtained from experiments. This equation compared well with the data
of Markels and Durfee [20]. Markels and Durfee used isopropyl alcohol and distilled water, both
being conducting liquids. A correction was made to allow for the cylindrical surface used in ref.
[20]. The choice of d which varies in the experiment was considered to be a possible reason for
any deviations of the theory from the experiment.
Zhorzholiani and Shekriladze [112] have successfully correlated their own results and those of
Markels and Durfee [20] on a log plot of
({lc)E FD + gP
vs
((lc)o gP
This is seen in Fig. 11. In their analysis, Zhorzholiani and Shekriladze considered F D as both the
second and third components of the E H D force in equation (1). It is the volume force density
obtained from their summation. In their experiments they considered both neutral (n-pentane,
benzene) and polar (acetone) fluids.

5C i ! I ~ II I
I i ~ I / .I ~L
(~=)= ' .I l J ..... ~ L I
Data scattedtl ~,.. • 4 - L ~ ~ I E .L.i. ~
(¢r=), • ''-r" ' , i I- - ' j " i I

.-.* ......

10 -~'.i-iII .... ~L-,

I
ii ]
Ili~,[i
i 1 tEip;
I
IJ !i tlii~
; ~ i l l
] ] I lille I I I ; I I;

Co~re4a~ion ~ On: Acetor~e a n d E ~ z e n e

(homogeneo~ l n d mhQrno~e~eous fleh:~s)
~ from [29] for i ~ r o ~ l and
d i = i l l e d water.

'1 l iiiLkl I ii'

10 100 1000
to* gp_
gP

Fig. 1I. Correlation of EHD increase in critical heat flux [I 12].

 Electrohydrodynamicenhancement of heat transfer and fluid flow 415

More recently, Cooper [128] has proposed a model for EHD nucleate boiling that leads to the
relationship

OrEN -"/2 = 0.3Reo °16, (51)

O~o

where n is found experimentally,

N=I+ 1.5fi(Es -- Ej) E2 (52)

and

Reo = . q a
t]sPl (Pl-- Pg)" (53)
Equation (51) correlates the results of refs [20, 67, 99, 100, 111] to within + 10% for the majority
of 37 observations graphed with only one outside ___30%, see Fig. 12.

6. ELECTRODE SYSTEMS

Little of the work reviewed here claims to include proposals for the practical implementation
of EHD enhancement in industrial plant, although most commend its potential for material
savings, etc. However, in Japan, Yabe et al. [88, 91] describe recent work on the development of
electrode systems for vertically aligned shell/tube condensers with particular applications in heat
pump systems. The electrode took the form of varying pitch helices. These use EHD to remove
the condensate film at points along the tube surfaces and to spray it on to the high voltage
electrodes down which it trickles. Figure 13 shows one such arrangement, optimized with respect
to extraction and removal angles (defined in the figure) and dimensions.
This arrangement has been incorporated into a 50 kW, 102 tube, EHD condenser for a prototype
heat pump and has shown heat transfer coefficient improvements compared with no EHD of about
six times for Rl14 [94]. (This work has also confirmed the validity of applying EHD results
obtained in single-tube rigs to multi-tube heat exchangers.) To enhance boiling, the Japanese have
also employed EHD pumping [114] or fine wire electrode systems [108] but neither has been applied
in the prototype heat pump mentioned above. By contrast, in the U.K., the electrode geometry
developed by Allen and Cooper [129] specifically for shell/tube heat exchangers and shown in
Fig. 14 is mechanically simpler and more compact. It has been shown to be effective for both
condensation and boiling. It comprises a combination of rod and perforated plane electrodes,
giving a fair approximation to the field that can be produced if a concentric cylinder was placed

• Baboietal [111] o Bonjouretal [99]

=--!~H-'~ • OtOi [100] • MarkelsandDurfee [20]
EO
n Watson [67]

0.5

8
8

0.1

0.01 0.1 1 10
Re o

Fig. 12. Correlation of END nucleate boiling heat transfer rate [128].
416 P . H . G . ALLEN and T. G. KARAYIANNIS

Heat
transfer tube

faction angle

spacer

Helical
wire

an

'Support rod

Fig. 13. Electrode system for EHD enhancement of vertical condenser tubes [88,91].

Electrically
insulating-
Mesh or
locating
pierced plate
plate

plate
~ f electrode

,., .t_

)
......... iiii...7

1 2 ....... l ........ i ~.... b (D

Fig. 14. Electrode system for EHD enhancement of vertical and horizontal shell/tube condensers and
boilers as proposed by ref. [129].
 Electrohydrodynamic enhancement of heat transfer and fluid flow 417

Shell of Rod
heat x ~ Electrode

Plane
Bectrode

Stress
relieving
rod

~attransfer
EqulpotenUals at surface
10% intervals
Fig. 15. Distribution of electric field strength for rod and plane electrode system [130].

around each tube, as shown in Fig. 15 from ref. [130]. Figure 16 shows an arrangement described
by Poulter and Allen [131] for applying an electric field at the inner surface of tubes.

7. D I S C U S S I O N

Clearly, in single phase heat transfer, certain geometries (e.g. inside tubes with axial electrodes)
and certain fluids (e.g. "polar" ones) are especially amenable to EHD enhancement. Although
many of the geometries investigated may seem to give little engineering design data, the fluids

Fig. 16. Electrode system designed and used to apply an electric field at the inner surface of tubes [131].
418 P.H.G. ALLENand T. G. KARAYIANNIS

investigated in them give an indication of their EHD potential. Evidently, all common gases are
amenable to corona wind generation. Fluids shown to have been amenable to low energy
consumption EHD enhancement are included in Tables I-3.
What direction should single-phase EHD enhancement research and development take? Two
clues are embedded in the data presented here. First+ the possibility noted in ref. [39] that a mixture
of d.c. with superimposed a.c. stress can give a more reliable and repeatable electrophoretic effect.
Secondly, that, as mentioned in ref. [58], low frequency "square wave" a.c. stress can have the same
effect. This is borne out by optical evidence of the behaviour of dielectric fluids in an electric field.
Many publications [24, 49, 5 l, 68] give this and probably the most apposite to the present argument
is ref. [51] which, using shadowgraphs, proves the region disturbed by the field developing from
the (positive) earthed electrode across the duct width.
What next in two-phase EHD heat transfer? The clues to the answer lie in the work of Ogata
et al. [16], Damianidis et aL I! 15] and the more recent work of Xu et al. [132]. An optimization
strategy in the design of E H D multi-tube evaporators has to be identified. Cornwell [133, 134]
examined the existence of sliding bubbles in boiling outside tube bundles and reported on their
influence on heat transfer. He stated that the heat transfer mechanism involves the process of
bubbles sliding on the surface disrupting the boundary layer and continued by saying that this effect
is more important than evaporation. He offered as evidence, among other observations, the
variation of the local heat transfer coefficient on a single tube, i.e. higher ~ on the tube sides.
Cornwell concluded that the mechanism of heat transfer at the sliding bubbles is dominated by
the influx of liquid to the layer under the bubble as it slides along the surface. The authors believe
that the above arguments must be considered, if one is to optimize the electrode/heat transfer
surface. The electrode/tube geometry has to be such that bubbles are directed along the heat
transfer surface. In the current work of Xu et al. different electrode/heat transfer geometries
are examined. A finite element model has been developed to calculate the E H D force on a bubble
using equation (22). The findings of the mathematical simulation will be verified in a smooth
tube bundle using R l l and R123. In the case of low-fin or other tubes with enhanced surfaces,
the local non-uniformity of the electric field will enhance heat transfer rates significantly, as
discussed earlier.
The isolated response of the heat exchanger has been examined and, as indicated above, will still
attract the attention of researchers. The integrated response of the thermal system and the beneficial
effects of E H D (size, cost) should also be considered. Ogata et al. [16] modelled an organic Rankine
cycle (ORC) power generation system with waste energy as the heat source and a heat pump using
waste heat from urban rivers as the heat source. Both systems used R123. The required heat transfer
area of the ORC, and accordingly the cost, decreased by 20% on the inclusion of the E H D
enhancement at the evaporator. In the heat pump system, at constant heat transfer, the cost of
the E H D evaporator was about ! 0% lower than a conventional (low-fin) one. The COP of the E H D
system was 7.9% higher. Damianidis et al. [l 15] reported on a model of a typical water chiller using
R22. They found that the use of an enhancement ratio of 1.8 gave a 15% reduction in the required
heat transfer area of the evaporator. Currently, the research team at South Bank University is
working on a prototype water chiller (modelled in ref. [115]) to assess experimentally the benefits
of the use of EHD. (R22 was found not to support the high voltage [135]. The plant currently
utilizes R12 and R123 will be used in a second stage.)
Obstacles to the development of improved heat transfer systems such as practical E H D heat
exchangers have been reviewed by Butterworth [136]. He points out that plant construction is
organized so that "people who can see the advantages of improved heat transfer systems, and
indeed benefit from them, may not be the same people charged with designing those systems. The
latter have no incentive to make changes and will therefore design the system the same way as they
did the last time". He concludes that "there are, however, possible ways to improve the shell and
tube, and we should use as much ingenuity to making further developments in the shell and tube
as we have to developing alternative types".
One must, of course, consider the relative merits of E H D and "passive" enhancement techniques,
remembering that the latter are effective only in the nucleate boiling region (up to E in Fig. 5).
By comparison, E H D can enhance every part of the boiling curve. Kajikawa et aL [137] gave results
for tubes comparable in size to those used by the present authors [13, 15, 89, 115, 116, 135] coated
 Electrohydrodynamicenhancementof heat transferand fluidflow 419

with a "Highflux"-type porous copper coating and with fine metal fibre used in an R114 boiler.
Apart from a graph showing a step increase of about three times in tube row heat transfer
coefficients as the warm water temperature increased to 30°C and above (no refrigerant saturation
temperature is quoted), which is described as "peculiar", there is no data on hysteresis. Maximum
values of about 11 kW/m:K can be construed from their results. Results for other refrigerants
are given by Czikk et al. [138] (R1 l, maximum • 42.5 kW/m2K) for coated tubes and by Arai et
al. [139] (Rl2, maximum ~ 24kW/m2K) for "Thermoexcel-E" with no mention of hysteresis.
Hysteresis with a porous metallic coating has, however, been thoroughly investigated by Bergles
and Chyu [140] for the case of R113. They found that about 8 K superheat was needed for
ebullition, which then increased ~ by about 20 times, a degree of boiling curve hysteresis that they
describe as "dramatic". They conclude that although greatly enhancing heat transfer, a porous
metallic matrix surface is to be expected to give hysteresis and that this effect had not been presented
previously in the literature.
Regarding enhanced condensation, Arai et al. [139] report improvements using R12 and
"Thermoexcel" tube amounting to between 25 and 70%. This, however, could well be no better
than the improved performance of tube banks given by EHD condensate stripping. The excellent
boiling performance of EHD, combined with a low fin surface, is probably connected with
Cornwell's observation [134] that "this may become churn flow where bubbles coalesce to form
elongated vapour regions". The electric field combines with buoyancy and the tube profile shape
to produce the elongation.
Lastly, the many other possibilities inherent in EHD enhancement should not be ignored. For
example, the work of Rutkowski [113] suggests that the evaporation of recalcitrant cryogenic
liquids may well be assisted by EHD. Bologa and Savin [141] described how an electric field can
be applied for intensifying and controlling the process of heat and mass transfer in zones of
evaporation, condensation and transport processes in heat pipes. In another paper, Savin et al. [142]
found that the rate of evaporation can increase significantly with the application of electric fields.
They reported that at 30 kV the rate of evaporation of liquid nitrogen increased by 6.5 times its
rate at zero-field. Higher increases were obtained for R l 13. Thornton [143] gives a general account
of the application of electrical energy to a variety of chemical and physical rate processes, including
heat and mass transfer.

8. CONCLUSIONS

The authors hope that they have justified the following conclusions:

I. Single phase heat transfer rates can be significantly enhanced using electric fields. Corona
wind and electrophoresis give the highest enhancement ratios, although consideration must be
given in the former case to the power consumption and in the latter to the purity and polarity of
the medium and the possibility of charge injection.
2. EHD pumps, although not very efficient, can find practical use due to the absence of moving
parts.
3. EHD can be applied effectively to the enhancement and control of condensation and boiling
of very many fluids.
4. In condensers, the EHD effect upon smooth tubes cannot compete with the degree of
enhancement offered by a "low-fin" surface. Nevertheless, it is expected to improve the perform-
ance of banks of low fin tubes.
5. In boilers, the EHD effect can increase 0c and eliminate both boiling hysteresis and decreases
in the boiling heat transfer coefficient due to transition and film boiling.
6. Electrode systems have been developed for the application of EHD in shell/tube heat
exchangers, although optimization is yet to be completed.

Acknowledgements--The results givenin Figs 3, 4, 6, 7 and 12 are fromwork supported, successively,by the Scienceand
Engineering Research Council and the BritishTechnologyGroup. The authors wish also to acknowledgethe assistance
of their research students R. K. AI-Dadah,Y. Yah and Y. Xu in preparing the figures.
420 P . H . G . ALLEN and T. G. KARAYIANNIS

REFERENCES

1. D. A. Reay, Heat transfer enhancement--a review of techniques and their possible impact on energy efficiency in the
U.K. Heat Recover), Systems & CHP !1, 1-40 (1991).
2. A. E. Bergles, The challenge of enhanced heat transfer with phase change. Trans. VII Congresso Nazionah, sulla
Transmissione del Calore, Florence, 1989, pp. 1 12.
3. A. E. Bergles, Techniques to augment heat transfer. In Handbook of Heat Transfer Applications (Edited by W. M.
Rohsenow, J. P. Hartnett and E. N. Ganic), 2nd Edn. McGraw-Hill, New York (1985).
4. S. Yilmaz, J. J. Hwalek and J. W. Westwater, Pool boiling heat transfer performance for commercial enhanced tube
surfaces. ASME, Paper No. 80-HT-41, 1980.
5. U D. Landau and E. M. Lifshitz, Electrodynamics of Continuous Media. Pergamon, New York (1963).
6. H. A. Pohl, Dielectrophoresis--The Behavior of Neutral Matter in Nonuniform Electric Field. Cambridge University
Press, Cambridge (1978).
7. T. B. Jones, Electrohydrodynamically enhanced heat transfer in liquids--a review. Adv. Heat Transfer 14, 107-148
(1978).
8. F. P. Grosu and M. K. Bologa, Similarity criteria for convective heat exchange in an electric field. Appl. Elec. Phenom.
(USSR) 20, 120 125 (1968).
9. A. Yabe, Active heat transfer enhancement by applying electric fields. Proc. A S M E / J S M E Thermal Engng Joint Conf.,
1991 (Edited by J. R. Lloyd and Y. Kurosaki), pp. xv-xxiii.
10. R. J. Turnbull, Electroconvective instability with a stabilizing temperature gradient If. Experimental results. Phys.
Fluids" 11, 2597-2603 (1968).
11. M. K. Bologa and A. B. Didkovskiy, Enhancement of heat transfer in film condensation of vapors of dielectric liquids
by superposition of electric fields. Heat Transfer--Soviet Res. 9(1), 147-151 (1977).
12. J. L. Olinger and C. P. Colver, A study of the effect of a uniform electric field on nucleate and film boiling. Chem.
Engng Prog. Symp. Ser. 67, 19-29 (1971).
13. T. G. Karayiannis, M. W. Collins and P. H. G. Allen, Electrohydrodynamic enhancement of nucleate boiling heat
transfer in heat exchangers. J. Chem. Engng Commun. 81, 15 24 (1989).
14. M. Winer, An experimental study of the influence of a nonuniform electric field on heat transfer in a dielectric fluid,
TRW Systems Report, Report No. EMI7-14, 1967.
15. P. H. G. Allen and P. Cooper, The potential of electrically enhanced evaporators. Proc. 3rd Int. Syrup. on the Large
Scale Application of Heat Pumps, Oxford, U.K., 1987, pp. 221 229.
16. J. Ogata, Y. Iwafuji, Y. Shimada and T. Yamazaki, Boiling heat transfer enhancement in tube-bundle evaporators
utilizing electric field effects. ASHRAE Trans. 98(2), 435-444 (1992).
17. M. M. Ohadi, R. A. Papar, T. L. Ng, M. A. Faani and R. Radermacher, EHD enhancement of shell-side boiling
heat transfer coefficients of R-123/oil mixture. ASHRAE Trans., ibid, 427-434 (1992).
18. W. M. Rohsenow, Nucleation with boiling heat transfer. Ind. Engng Chem. 58(1), 40-47 (1966).
19. W. M. Rohsenow, A method of correlating heat-transfer data for surface boiling of liquids. Trans. Amer. Soc. Mech.
Engrs 74, 969-976 (1952).
20. M. Markels and R. L. Durfee, The effect of applied voltage on boiling heat transfer. A1ChE J. 10(1), 106-109 (1964).
21. Anon. Developments to watch: a noiseless fan, with no moving parts. Product Engng, 10 October, 29 (1966).
22. R. E. Holmes and S. J. Basham, A dry cooling system for steam power plants. Proc. 6th b~tersociety Energy Conversion
Eng. Con£, Boston, MA, 1971, pp. 1218 1228.
23. H. R. Velkoff, The effects of ionization on the flow and heat transfer of a dense gas in a transverse electrical field.
Proc. Heat Tran,~fer and Fluid Mechanics Institute, University of California, Berkeley (Edited by W. H. Giedt and S.
Levy), pp. 260-275. Stanford University Press (I964).
24. R. A. Moss and J. Grey, Heat transfer augmentation by steady and alternating electric fields. Proc. Heat TransJer
and Fluid Mechanics Institute, University of Santa Clara, 1966 (Edited by M. A. Saad and J. A. Miller), pp. 210-235.
25. F. P. Grosu and M. K. Bologa, The influence of electric fields on heat-exchange processes in gases. Appl. Elec. Phenom.
(USSR) 5(23), 350-356 (1968).
26. M. Robinson, Convective heat transfer at the surface of a corona electrode, Int. J. Heat Mass Transfer 13, 263-274
(1970).
27. J. McDermott, High-voltage ionic discharges provide silent, efficient cooling. Electronic Des. 20 30 September, 22-24
(1971).
28. T. Mizushina, H. Ueda, T. Matsumoto and K. Waga, Effect of electrically induced convection on heat transfer of
air flow in an annulus. J. Chem. Engng Japan 9, 97-102 (1976).
29. H. Windischmann, Investigation of corona-discharge cooling (CDC) of a horizontal plate under free convection,
AIChE Syrup. Series 138, 23--30 (1974).
30. J. Yamaga and M. Jido, Corona discharge cooling technique, U.S. Patent No. 3,938,345, 1976.
31. K. G. Kibler and H. G. Carter, Locatised corona discharge. U.S. Patent No. 4,015,658, 1977.
32. A. Yabe, Y. Mori and K. Hijikata, Heat transfer augmentation around a downward-facing flat plate by non-uniform
electric fields. Proc. 6th Int. Heat Tran~$fer Cot~f., 1978, Paper M-20, Vol. 3, pp. 171 176.
33. A. Yabe, Y. Mori and K. Hijikata, EHD study of the corona wind between wire and plate electrodes. AIAA J. 16(4),
340 345 (1978).
34. H. R. Velkoff and R. Godfrey, Low-velocity heat transfer to a flat plate in the presence of a corona discharge in air.
J. Heat Transfer 101, 157 163 (1979).
35. Y. Tada, A. Takimoto and Y. Hayashi, Heat transfer enhancement in a convective field by applying ionic wind. Proc.
A S M E / J S M E Thermal Engng Joint Con/i, 1991 (Edited by J. R. Lloyd and Y. Kurosaki), Vol. 3, pp. 9-14.
36. D. A. Nelson, M. M. Ohadi, S. Zia and R. L. Whipple, Electrostatic effects on heat transfer and pressure drop in
cylindrical geometries, ibid, pp. 33 39.
37. E. Schmidt and W. Leidenfrost, Der Einfluss elektrischer Felder auf den W/irmetransport in ffiissigen elektrischen
Nichtleitern, Forsch. auf dem Geb. des Ing.-wes 19(3), 65 80 (1953).
38. E. Schmidt and W. Leidenfrost, Der W~irmetransport in fliissigen elektrischen Nichtleitern unter dem Einfluss
elektrischer Felder, Chemie-lng.-Techn. 26, 35-38 (1954).
 Electrohydrodynamic enhancement of heat transfer and fluid flow 421

39. P. H. G. Allen, Electric stress and heat transfer, Brit. J. Appl. Phys. 10, 347-351 (1959).
40. S. Mascarenhas, Y. Mascarenhas, M. Ferreira De Souza and R. F. Rabello, Thermal conduction of liquid dielectric
under the influence of electric fields (fatty acids). An. Acad. Brasil Ci. 28, 95-98 (1956).
41. H. Senftleben and P. Schnabel, Der Einfluss yon Raumladungen in hochisolierenden Fl/issigkeiten auf den
W/irmeiibergang unter Wirkung elektrischer Felder. Z. Phys. 170, 82-92 (1962).
42. J. M. Care and D. W. Swan. Some transient phenomena in heat transfer resulting from electric stress. Brit. J. Appl.
Phys. 14, 263-266 (1963).
43. J. E. Porter and R. Poulter, Electrothermal convection effects with laminar flow heat transfer in an annulus. Proc.
4th Int. Heat Transfer Conf., Paris, 1970, Paper FC3.7, Vol. 2.
44. D. C. Newton and P. H. G. Allen, Senftleben effect in insulating oil under uniform electric stress. Lett. Heat Mass
Transfer 4, 9-16 (1977).
45. R. Poulter and I. A. Miller, Heat transfer enhancement in shell/tube heat exchangers employing electrostatic fields.
Proc. 1st U.K. Nat. Heat Transfer Conf., Leeds, 1984, VoL 2, pp. 707-716.
46. J. L. Fernandez and R. Poulter, Radial mass flow in electrohydrodynamically enhanced forced heat transfer in tubes.
Int. J. Heat Mass Transfer 30, 2125-2135 (1987).
47. S. Mascarenhas, The action of homogeneous electric fields upon the heat transmission of liquid dielectrics. An. Acad.
Brasil. Ci. 28, 99-105 (1956).
48. U.T. Borbulya, I. A. Kojuhar and M. K. Bologa, Electroconvection heat transfer with dielectric fluids. Arch. Academy
ofMoidavian S.S.R. 2, 90-93 (1965).
49. M. J. Gross and J. E. Porter, Electrically induced convection in dielectric liquids. Nature 212(5068), 1343-1345
(1966).
50. J. E. Porter and R. B. Smith, The effect of a transverse electrostatic field on laminar flow heat transfer in a rectangular
duct. Proc. 5th Int. Heat Transfer Conf., Tokyo, 1974, Paper FC 5.4, Vol. 2, pp. 198-202.
51. T. Fujino, Y. Yokoyama and Y. H. Mori, Augmentation of laminar forced-convection heat transfer by the application
of a transverse electric field. Trans. A S M E J. Heat Transfer 111, 345-351 (1989).
52. H. Ishiguro, S. Nagata, A. Yabe and H. Nariai, Augmentation of forced-convection heat transfer by applying electric
fields to disturb flow near a wall. Proc. A S M E / J S M E Thermal Engng Joint Conf., 1991 (Edited by J. R. Lloyd and
Y. Kurosaki), Vol. 3, pp. 25-31.
53. V. Stach, Influence of electric field on the cooling gas flow in a nuclear reactor. Int. J. Heat Mass Transfer 5, 445-456
(1962).
54. F. Berger and L. Derian, The influence of an electric field on the heat transfer to CO2 at high and low pressure in
a nuclear reactor. Proc. 3rd Int. Conf. on the Peaceful Uses of Atomic Energy, Geneva, 1964, Vol. 8, pp. 355-363.
55. F. Berger and L. Derian, The influence of an electric field on the heat transfer to gaseous coolants in a nuclear reactor.
Acta Technica CSAV 3, 312-335 (1965).
56. J. M. Coulson and J. E. Porter, The effects of electric fields on transport phenomena. Trans. Inst. Chem. Engrs 44,
T388-T394 (1966).
57. R. Schnurmann and M. G. C. Lardge, Enhanced heat flux in non-uniform electric fields. Proc. R. Soc. Lond. A 334,
71-82 (1973).
58. P. Cooper and P. H. G. Allen, The Senftleben effect in cross-flow heat exchange and the part played by electronic
conduction. PhysicoChem. Hydrodynam. 4, 85-101 (1983).
59. N. F. Baboi, M. K. Bloga and K. N. Semenov, The influence of electric fields on heat transfer in liquids and gases.
Appl. Elec. Phenom. (USSR) 1, 46-57 (1965).
60. K. N. Semenov, N. F. Baboi and M. K. Bologa, Effective influence of electro-magnetic field on heat transfer in gases
and liquids. Academic News o f the Moldavian SSR 2, 82-89 (1965).
61. H. Senftleben, Die Einwirkung elektrischer und magnetischer Felder auf des W/irmeleitvermSgen von Gasen. Phys.
Z. 32, 550 (1931).
62. H. Senftleben and W. Braun, Der Einfluss elektrischer Felder auf den W/irmestrom in Gasen. Z, Phys. 102, 480-506
(1936).
63. S. Y. Wang, M. W. Collins and P. H. G. Allen, Prediction of the effect of electrohydrodynamic (EHD) enhancement
of fluid flow and heat transfer. Proc. 1st Int. Conf. on Advanced Computation Methods in Heat Transfer, Portsmouth,
1990, Vol. 2, pp. 231-238.
64. P. H. G. Allen, Heat transfer at high voltage. Elec. Times 139, 321-322 (1961).
65. G. Ahsmann and R. Kronig, The influence of electric fields on the convective heat transfer in liquids. Appl. Sci. Res.
A 2, 235-244 (1950). (Correction: 3, 83/84, 1951.)
66. H. J. De Haan, The influence of electric fields on the convective heat transfer in liquids II. Appl. Sci. Res. A 3, 85-88
(1951).
67. P. K. Watson, Influence of an electric field upon the heat transfer from a hot wire to an insulating liquid. Nature 189,
563-564 (1961).
68. K. H. Weber and G. H. Halsey, Free convection in electric fields. Proc. Heat Transfer and Fluid Mechanics Institute,
pp. 97-100. Stanford University Press (1953).
69. S. D. Savkar, Dielectrophoretic effects in laminar forced convection between two parallel plates. Phys. Fluids 14,
2670-2679 (1971).
70. A. J. Oliver, The effect of electric fields on heat transfer and its relevance to the CEGB SF 6 power cable, CERL Lab
Note TPRD/L/2398/N82, 1983.
71. J. Seyed-Yagoobi, J. C. Chato, J. M. Crowley and P. T. Krein, Induction electrohydrodynamic pump in a vertical
configuration: Part 1--Theory. Trans. A S M E J. Heat Transfer 111, 664-669 (1989).
72. J. Seyed-Yagoobi, J. A. Castafieda and J. E. Bryan, Theoretical analysis of ion-drag pumping. Proc. IEEE-IAS Annual
Meeting, 1992.
73, M. Sato, A. Yabe and T. Taketani, Heat transfer enhancement by applying an electro-hydrodynamical pump utilizing
dielectrophoretic force. Proc. A S M E / J S M E Thermal Engng Joint Cot~f., 1991 (Edited by J. R. Lloyd and Y. Kurosaki),
Vol. 3, pp. 3-8.
74. J. Seyed-Yagoobi, J. C. Chato, J. M. Crowley and P. T. Krein, Induction electrodynamic pump in a vertical
configuration: Part 2--Experimental study. Trans. A S M E J. Heat Transfer I I I, 670~74 (1989).
422 P . H . G, ALLEN and T. G. KARAYIANNIS

75. B. J. Bohinsky and J. Sayed-Yagoobi, Induction electrohydrodynamic pumping. Selecting an optimum working fluid.
Proc. IEEE-IAS Annual Meeting, 1990, pp. 795 801.
76. J. Seyed-Yagoobi, A theoretical study of induction electrohydrodynamic pumping in outer space. Trans. ASME J.
Heat Transfer 112, 1095 1097 (1990).
77. J. E. Bryan and J. Seyed-Yagoobi, Experimental study of ion-drag pumping using various working fluids. IEEE Trans.
Electrical Insulation 26, 647 655 ( 1991 ).
78. J. A. Castafieda and J. Seyed-Yagoobi, Electrohydrodynamic pumping of Refrigerant 11. Proc. IEEE-IAS Annual
Meeting, 1991, pp. 500- 503.
79. J. E. Bryan and J. Seyed-Yagoobi, An experimental investigation of ion-drag pump in a vertical and axisymmetric
configuration. IEEE Trans. on Industry Applications 28, 310-316 (1992).
80. H. R. Velkoff and J. H. Miller, Condensation of vapour on a vertical plate with a transverse electrostatic field. Trans.
ASME J. Heat Transfer 87, 197-201 (1965).
81. H. Y. Choi and J. M. Reynolds, Study of electrostatic effects on condensing heat transfer, Air Force Technical Report
AFFDL-TR-65-51, Air Force Flight Dynamics Laboratory, Wright-Patterson Air Force Base, Ohio, 1965.
82. H. Y. Choi, Electrohydrodynamic condensation heat transfer, Trans. ASME J. Heat Transfer 90, 98102 (1968).
83. R. E. Holmes and A. J. Chapman, Condensation of Freon-114 in the presence of a strong nonuniform alternating
electric field. Trans. ASME J. Heat Tran~fer 92, 616~620 (1970).
84. A. K. Seth and L. Lee, The effect of an electric field in the presence of noncondensable gas on film condensation heat
transfer. Trans. ASME J. Heat Tran,~['er 96, 257-258 (1974).
85. A. B. Didkovskii and M. K. Bologa, Intensification of heat exchange upon condensation of a vapour in an electric
field. High Temp. 16, 490~496 (I978).
86. A. B. Didkovsky and M. K. Bologa, Vapour film condensation heat transfer and hydrodynamics under the influence
of an electric field. Int. J. Heat Mass Tramfer 24, 811-819 (1981).
87. G. F. Smirnov and V. G. Lunev, Heat transfer during condensation of vapour of dielectric liquids in electric fields.
Appl. Elec. Phenom. (USSR) 2, 37 42 (1978).
88. A. Yabe, K. Kikuchi, T. Taketani, Y. Mort and K. Hijikata, Augmentation of condensation heat transfer by applying
non-uniform electric fields. Proc. 7th Int. Heat Transfer Conf., 1982, Vol. 5, pp. 189-194.
89. P. Cooper and P. H. G. Allen, The potential of electrically enhanced condensers. Proc. ~~d Int. Syrup. on the Large
Scale Application of Heat Pumps, Oxford, U.K., 1984, pp. 295--309.
90. J. Trommelmans and J. Berghmans, Influence of electric fields on condensation heat transfer of non-conducting fluids
on horizontal tubes. Proc. 8th Int. Heat Transfer Conf., 1986, Vol. 6, pp. 2969-2974.
91. A. Yabe, T. Taketani, K. Kikuchi, Y. Mort and H. Maki, Augmentation of condensation heat transfer by applying
electro-hydro-dynamical pseudo-dropwise condensation, ibid, pp. 2957-2962.
92. M. K. Bologa, I. K. Savin and A. B. Didkovsky, Electric-field-induced enhancement of vapor condensation heat
transfer in the presence of a non-condensable gas. Int. J. Heat Mass Transfer 30(8), 1577.1585 (1987).
93. C. Damianidis, M. W. Collins, T. G. Karayiannis and P. H. G. Allen, EHD effects in condensation of dielectric fluids.
Proc. 2nd Int. Symp. on Condensers and Condensation, Bath, U.K., 1990, pp. 505--518.
94. K. Yamashita, M. Kumagai, S. Sekita, A. Yabe, T+ Taketani and K. Kikuchi, Heat transfer characteristics of an EHD
condenser. Proc. ASME/JSME Thermal Engng Joint Con[]+ 1991 (Edited by J. R. Lloyd and Y. Kurosaki), Vol. 3,
pp. 61 67.
95, G. F. Smirnov and 1. I. Lukanov, Study of heat transfer from Freon-I 1 condensing on a bundle of finned tubes. Heat
Transfer--Sot'iet Research 4(3), 51 56 (1972).
96+ P. Cooper, Electrically enhanced heat transfer in the shell/tube heat exchanger, PhD Thesis, University of London,
I986.
97. L. W. Chubb, Improvements relating to methods and apparatus for heating liquids. U.K. Patent No. 100,796, (1916).
98. L. Bochirol, E. Bonjour and L. Weil, Exchanges thermiques-Etude de l'action de champs electriques sur les transferts
de chaleur dans les liquides bouillants. C.R. Heb. des Seances de I'Acad. des Sciences (Paris) 250, 76 78 (1960).
99. E. Bonjour, J. Verdier and L. Weil, Electroconvection effects on heat transfer. Chem, Engng Progr. 58(7)+ 63-66 (1962).
100. H. Y. Choi, Electrohydrodynamic boiling heat transfer, PhD Thesis, Dept. of Mech. Eng., M.I.T., 1962.
101. A. K. Jalaluddin and D. B. Sinha, Effect of an electric field on the superheat of liquids, Nuovo Cimento 26 (supplement),
Series X, 234--237 (1962).
102. M. Markels and R. L. Durfee, Studies of boiling heat transfer with electric fields. Part I. Effect of applied a.c. voltage
on boiling heat transfer to water in forced convection. AIChE J, 11(4), 716-719 (1965).
103. R. F. Lovenguth and D. Hanesian, Boiling heat transfer in the presence of nonuniform, direct current electric fields.
Ind. Engng Chem. Fundam. 10(4), 570-576 (1971).
104. D. K. Basu, Effect of electric field on boiling hysteresis in carbon tetrachloride. Int. J. Heat Mass Tran,~ter 16,
1322 1324 (1973).
105. T. B. Jones and R. C. Schaeffer, Electrohydrodynamically coupled minimum film boiling in dielectric liquids. AIAA
J. 14(12), 1759-1765 (1976).
106. T. B. Jones and K. R. Hallock, Surface wave model of electrohydrodynamically coupled minimum film boiling. J+
Electrostatics 5, 273-284 (1978).
107. V. A. Zheltukin, Y. K. Solomatnikov, D. M. Mikhaylov and A. G. Usmanov, Effect of electric-field frequency on
heat transfer. Heat Transfer--Soviet Research 10(6), 10-13 (1978).
108. H. Kawahira, Y. Kubo and T. Yokoyama, The effect of an electric field on boiling heat transfer of Refrigerant 11.
Proc. IEEE/IAS Annual Meeting, Atlanta, 1987, pp. 1446-1454.
109. R. A. Blachowicz, B. W. Brooks and K. B. Tan, Boiling heat transfer in an electric field. Chem. Engng Sci. 35, 761-762
(1980).
110. V. Asch, Electrokinetic phenomena in boiling "Freon-113". J. Appl. Phys. 37(7), 2654 2658 (1966).
111. N. F. Baboi, M. K. Bologa and A. A. Klyukanov, Some features of ebullition in an electric field. Appl. Elec. Phenom.
(USSR) 20, 126-141 (1968).
112. A. G. Zhorzholiani and 1. G. Shekriladze, Study of the effect of an electrostatic field on heat transfer with boiling
dielectric fluids. Heat Transfer--Soviet Research 4(4), 81-98 (1972).
113. J. Rutkowski, The influence of electric field on heat transfer in boiling cryogenic liquid+ Cryogenics 17, 242-243 (1977).
 Electrohydrodynamic enhancement of heat transfer and fluid flow 423

114. A. Yabe and H. Maki, Augmentation of convective and boiling heat transfer by applying an electro-hydrodynamical
liquid jet. Int. J. Heat Mass Transfer 31('Q, 407-417 (1988).
115. C. Damianidis, T. G. Karayiannis, R. K. Al-Dadah, R. W. James, M. W, Collins and P. H. G. Allen, EHD boiling
enhancement in shell and tube evaporators and its application in refrigeration plants. A S H R A E Trans. 98(2), 462-472
(1992).
116. C. Damianidis, M. W. Collins, T. G. Karayiannis and P. H. G. Allen, EHD enhancement of heat transfer in
evaporators and condensers. Proc. l lth Int. Syrup. o f Heating, Refrigerating and Air Conditioning, Interklima 91,
Zagreb, 1991, pp. I0-26.
117. Electrohydrodynamic (EHD) Effects on Condensing and Evaporating Freons. Video held by ASME Heat Transfer
Films Library, Mechanical and Aerospace Engineering, The University of Tennessee, Knoxville, TE, 37996-2210.
118. Y. Y. Yah, T. G. Karayiannis, P. H. G. Allen, M. W. Collins and R. S. Neve, EHD enhancement of boiling heat
transfer at low superheat surfaces in heat exchangers. Proc. Eurotherm Seminar No. 33, Recent Developments in Heat
Exchanger Technology, Paris, 1993.
119. A. Yabe, T. Taketani, H. Maki, K. Takahashi and Y. Nakadai, Experimental study of electrohydrodynamically (EHD)
enhanced evaporator for nonazeotropic mixtures. ASHRAE Trans. 98(2), 455-461 (1992).
120. R. Kronig and N. Schwartz, On the theory of heat transfer from a wire in an electric field. Appl. Sci. Res. A l, 35-46
(1947).
121. H. Senftleben and E. Bfiltmann, Die Einwirkung elektischer Felder auf den W/irmefibergang in Gasen. Z. flit Phys.
136, 389-401 (1953).
122. J. L. Fernandez, Electrohydrodynamic enhancement of forced convective heat transfer in tubes, PhD thesis, University
of Bristol, U.K., 1975.
123. J. H. Davidson, F. A. Kulacki and P. F. Dunn, Convective heat transfer with electric and magnetic fields. In Handbook
of Single-Phase Convective Heat Transfer (Edited by S. Kakac, R. K. Shah and W. Aung), Chap. 9. Wiley, New York
(1987).
124. R. L. Johnson, Effect of an electric field on boiling heat transfer. AIAA J. 6(8), 1456-1460 (1968).
125. N. Zuber and M. Tribus, Further remarks on the stability of boiling heat transfer, UCLA Report No. 58-5, UCLA,
Los Angeles, California, 1958.
126. N. Zuber, Hydrodynamic aspects of boiling heat transfer, ACE Rept No. AECU-4439, Atomic Energy Commission,
1959.
127. J. Berghmans, Electrostatic fields and the maximum heat flux. Int. J. Heat Mass Transfer 21, 791-797 (1978).
128. P. Cooper, EHD enhancement of nucleate boiling. Trans. A S M E J, Heat Transfer 112, 458-464 (1990).
129. P. H. G. Allen and P. Cooper, Improvements in or relating to heat exchangers, U.K. Patent No. 8522680, 1985.
130. P. Cooper, Practical design aspects of EHD heat transfer enhancement in evaporators. ASHRAE Trans. 98(2), 445-454
(1992).
131. R. Poulter and P. H. G. Allen, Electrohydrodynamically augmented heat and mass transfer in the shell/tube heat
exchanger. Proc. 8th Int. Heat Transfer Conf., San Francisco, 1986, Vol. 6, pp. 2963-2968.
132. Y. Xu, R. K. A1-Dadah, P. H. G. Allen and T. G. Karayiannis, Incorporation of EHD enhancement in heat
exchangers. I C H M T Int. Syrup. on New Developments in Heat Exchangers, Portugal, 1993.
133. K. Cornwell, The influence of bubbly flow on boiling from a tube in a bundle. Proc. Eurotherm No. 8, Advances in
Pool Boiling Heat Transfer, Paderborn, Germany, 1989, pp. 177 184.
134. K. Cornwell, The role of sliding bubbles in boiling on tube bundles. Proc. 9th Int. Heat Transfer Conf., Israel, 1990,
Vol. 3, pp. 455-460.
135. T. G. Karayiannis, R. K. A1-Dadah, R. W. James and P. H. G. Allen, Electrohydrodynamic boiling heat transfer
enhancement in heat exchangers. Paper 93-WA/HT-35, ASME Winter Annual Meeting, New Orleans, U.S.A., 1993.
136. D. Butterworth, Changes in heat exchangers for process applications: incentives and barriers. Heat Transfer Engng
8(4) 19-22 (1987).
137. T. Kajikawa, T. Agawa, H. Takazawa, M. Amano, K. Nishiyama and T. Homma, Studies on OTEC power system
characteristics and enhanced heat transfer performance. Proc. 6th OTEC Conf., Washington USA, Paper 11-5, 1979.
138. A. M. Czikk, C. F. Gottzmann, E. G. Ragi, J. G. Withers and E. P. Habdas, Performance of advanced heat transfer
tubes in refrigerant-flooded liquid coolers. ASHRAE Trans. 76(1), 96-119 (1970).
139. N. Arai, T. Fukushima, A. Arai, T. Nakajima, K. Fujie and Y. Nakayama, Heat transfer tubes enhancing boiling
and condensation in heat exchangers of a refrigerating machine. ASHRAE Trans. 83(2), 58-70 (1977).
140. A. E. Bergles and M. C. Chyu, Characteristics of nucleate pool boiling from porous metallic coatings. Trans. A S M E
J. Heat Transfer 104, 279-285 (1982).
141. M. K. Bologa and I. K. Savin, Electrohydrodynamic heat pipes. 7th Int. Heat Pipe Conf., Minsk, 1990, BI4P.
142. I. K. Savin, M. K. Bologa and V. P. Kor~vkin, Effect of an electric field on the rate of heat and mass exchange during
evaporation. Electronnaya Obrabotka Materialov 6, 52 54 (1986).
143. J. D. Thornton, The applications of electrical energy to chemical and physical rate processes. Rev. Pure Appl. Chem.
18, 197-218 (1968).