Journal of Electrical and Electronic Engineering

2015; 3(4): 76-86

Published online July 24, 2015 (http://www.sciencepublishinggroup.com/j/jeee)
doi: 10.11648/j.jeee.20150304.13
ISSN: 2329-1613 (Print); ISSN: 2329-1605 (Online)

Enhanced EHD and Electrostatic Propulsion Devices Based

on Polarization Effect Using Asymmetrical Metal Structure
Taku Saiki
Department of Electrical and Electronic Engineering, Faculty of Engineering Science, Kansai University, Suita, Japan

Email address:
[email protected]

To cite this article:

Taku Saiki. Enhanced EHD and Electrostatic Propulsion Devices Based on Polarization Effect Using Asymmetrical Metal Structure. Journal of
Electrical and Electronic Engineering. Vol. 3, No. 4, 2015, pp. 76-86. doi: 10.11648/j.jeee.20150304.13

Abstract: Electro hydro dynamic (EHD) and electrostatic propulsion devices were developed in the 1920s by Thomas
Townsend Brown. One such device, called a “lifter”, has no moving parts and, in the air, operates on electrical energy. It is a
fashionable device and has a very simple structure, basically consisting of a narrow wire electrode and a large, flat one. However,
it has a low ratio of propulsion force to unit electrical input power. According to theory, the propulsion force it generates depends
on the interaction between the ion density of the ionized air and the charges on the surface of the large electrode. EHD and
electrostatic propulsion models using the polarization effect are proposed to improve the ratio of the propulsion force to unit
electrical input power. The propulsion device generates propulsion force through the use of an asymmetrical metal structure with
charges generated by the polarization effect. The propulsion force the new devices generated for the same electric energy was 5.7
times higher than that of a basic type lifter owing to additional propulsion force being generated by the maximum polarization
effect in the experiments. It was found that combining other effects with this polarization effect results in the ratio of generated
propulsion force to electric power being close to 100N/kW when the electric power is high. This value is as high as that of a
helicopter. We also performed numerical analysis was also performed for capacitances and charges for various kinds of EHD and
electrostatic propulsion devices. An optimized system was developed and is discussed in this paper.
Keywords: EHD, Electrostatic Propulsion, Polarization, Electric Field, Charge, Electron

to explain the generated propulsion forces were proposed. One

1. Introduction view, that the effect can be explained by “generation of
A lifter that is made light in weight by using a high voltage momentum by ion wind”, was shown by the USA’s National
supply floats in the air. For such lifters, only electrical energy Renewable Energy Laboratory. However, recent rigid research
is needed to obtain the propulsion force in the air. Normal indicates that the principle of the propulsion can be explained
propulsion devices such as rockets need materials to be as the electrical forces between charges in ionic wind and the
propelled. Lifter-type propulsion devices have no moving electrons on the large flat electrode[6-8]. Thus, it has been
parts as helicopters do, and can be expected to be used as shown that the lifter propulsion is based on a principle that
Unmanned Aerial Vehicles (URVs) in the future [1]. differs from conventional principals. Rockets work on the
The Biefeld-Brown effect was discovered by Paul Alfred basis of the equation of motion.
Biefeld and Thomas Townsend Brown in the 1920s [2]. The objectives in developing the propulsion devices are
Brown also proposed and was awarded patents for a number of achieving 1) low propulsion force to electrical input energy, 2)
electro hydro dynamic (EHD) propulsion devices, such as low propulsion force per unit volume, and 3) low propulsion
lifters [1,3]. The principle, which is called ion craft, should be force per unit weight. Some papers [12,13] have previously
the same as that of the lifter. Many movies of lifters floating in reported experimental results for improving the ratio of
the air can be seen on the Internet, and a number of papers on generated propulsion force to electric energy. Thus, at present,
the theory of lifters to generate propulsion forces have been heavy, high voltage sources cannot be mounted on propulsion
published[4-12]. After the Biefeld-Brown effect was devices.
published, many discussions on it were held. Mainly, the In this paper, I propose an improved model for the structure
principles behind the effect were discussed and many theories of the rounded electrode used to improve the charge density of
ion wind. It is a model with a simplified small negative
 Journal of Electrical and Electronic Engineering 2015; 3(4): 76-86 77

electrode to make the ion source small. In addition, propulsion in improved propulsion force. However, the electrical charge
devices using the polarization effect, such as a cascading is not always proportional to the propulsion forces the EHD
model, were fabricated to improve the ratio of propulsion devices generate.
force to electrical input energy. In this work, I used a rounded electrode for the structure of
the EHD device in order to 1) use the model as a unit module,
2) achieve uniform divergence of the ionized air in the radial
2. Principle of EHD and Electrostatic direction, and 3) efficiently protect the ion density in the
Propulsion Devices rounded electrode from degradation.
I used metal structure for the device to enhance 1) the
Simply put, a lifter consists of a tin wire electrode and a charges on the electrodes owing to the polarization effect, 2)
large, flat electrode [1]. The principle for producing the the emission of ions from the negative needle electrode, and 3)
propulsion force of a lifter and EHD propulsion devices was the propulsion force produced by the electrical field between
proposed in recent research activities shown in the literature the ion wind and the large electrode or between the ion wind
[8,11]. An equation to calculate the generated propulsion force and the polarization electrodes.
of a lifter is shown next in references [8] and [11] For the idea for polarization effect in EHD devices, one [1]
of Brown’s patents proposed that an EHD device in which a
F = ∫ ρE ⋅ dv . (1)
v dielectric material is sandwiched between the electrodes
produces high propulsion force. The details for this
In the equation, E is the electric field generated from the experiment differed from those given in Brown’s patent in that
charges on the large flat electrode and v is the effective volume propulsion forces are generated by metal electrodes for which
area in which ionized air flows near the large electrode. The an asymmetrical structure is used to produce the polarization
equation for calculating ρ, the ion density near the large effect.
electrode in the air, is Here, the needle electrode was connected to the negative
and the large electrode to the positive voltage. The structure
Id
ρ= . (2) for generating ions was simplified.
µVA

Here, I is the current, d is the gap between the tin wire

electrode and the large flat electrode, µ is the ion mobility in
the atmosphere, V is the voltage between the electrodes, and A
is the profile area of ionized air floating near the large
electrode. It shows that all the generated ions do not interact
with the large electrode. Here, if we set E=V/d, eq. (1) is
simplified and we obtain

Id
F= . (3)
µ

Here, I describe the meaning of eq. (1). The charges on the

large electrode attract the ion wind by the electric field
between the generated ions and the large flat electrode. As a
result, propulsion force is produced. The observed current at (a)
the power supply results from the fact that ions exchange the
electrons on the large electrode (interaction between the ion
wind and the large flat electrode). Detailed numerical
estimations of EHD fluids, space charge, and potential in the
ionized air flow are shown in reference [5-9].
The ratio of generated power to electrical input power is
given by

F I
θ= = . (4)
P µE

Below, I simply describe the calculation method for the

capacitance between the electrodes and the charge on them.
The method is shown in eq. (1), according to the principle for
producing propulsion force. Evaluating the quantity of the
charge on the electrode when a high voltage is applied results (b)
78 Taku Saiki: Enhanced EHD and Electrostatic Propulsion Devices Based on Polarization Effect Using
Asymmetrical Metal Structure

For calculating model 1b, there is one metal rod electrode to

produce polarization connected directly to inner electrode. RG
is the gap length as shown in Fig. 1(b). RG is very much
shorter than L. Here, Ca2b2 is generated capacitance governed
by eq. (7) to set the space gap and Ca1b1 is the capacitance
between the negative needle electrode and the inner radial
electrode governed by eq. (8). The total capacitance for model
1b is determined by calculating the series connection of Ca1b1
and Ca2b2. The equation for calculating the capacitance for
model 1b is

Ca1b1 ⋅ Ca 2b 2
C2 = . (9)
Ca1b1 + Ca 2b 2

If the number of metal rods needed to generate polarization

(c)
is n, the capacitance is rewritten as
Fig. 1. Calculation models. (a) normal model 1a, (b) double ring model 1b, (c)
polarization pole model 1c C2 ' = nC2 . (10)
For the model shown in Fig. 1(a) I used a rounded electrode, In the case of model 1c, the capacitance between the upper
in contrast to the conventional lifter that uses flat plate and lower ring electrodes is
electrodes. Here, I neglected ion flow because its calculation
is very complicated. The equation to calculate the electrical 2πR
C0 = 2πε0 (11)
potential on the radial axis r for model 1a is ln ( 2R / a )


(
 L + L2 + R − r 2
) 
1/ 2

  ( ) 
The capacitance generated by using a center electrode to
 ln   generate polarization is the same as that for model 1b. The
R−r
−Q  1    capacitance C2 is derived by calculating the series connection
V (r) = ⋅ − . (5) of two Cab shown in eq. (7). The total capacitance C3 for model
4πε0  r L  1c, derived by calculating the parallel connection of C2 and C0,
 
  is
 
C3 = C 0 + C 2 . (12)
Here, L is the width of the large electrode and R is the radius
It is clear that eq. (12) and eq. (9) indicate increased
of the large electrode. Thus, the equation for the electrical
capacitances in each of the propulsion models.
potential difference Vab between both electrodes is
Here I return to the discussion about the propulsion force.

( )
 For model 1b, the generated propulsion force obtained by
 1/ 2

 R − a L + L + (R − b)
2 2
Q 1 1  reconsidering eq. (1) is
Vab = ⋅  − + ln  ⋅ 1/ 2   . (6)
4 πε 0 a b

 (
R − b L + L2 + R − a 2
( ) )  
 F2 = ∫ ( ρ1E1 + ρ2 E 2 ) ⋅ dv .
v
(13)

Here, 0＜a,b<R, and it is assumed that b is very close to R. Here, E1 is the electric field producing the charges on the
The equation for the capacitance between the electrodes for inner radial ring to generate polarization, E2 is the electric
model 1a is field producing the charges on the large outer radial electrode,
ρ1 is the ion density near the inner radial ring to generate
4πε 0
Cab = polarization, and ρ2 is the ion density near the large outer
 
( ) 
1/ 2

 R − a L + L + (R − b)
2 2 radial electrode.
1 1   . (7)
 a − b + ln  R − b ⋅
For model 1c, the generated propulsion force obtained by
1/ 2  


 (
L + L2 + ( R − a )
2
)  

reconsidering eq. (1) is

F3 = ∫ [ρ 3 E3 + ρ 4 (E2 + E4 )]⋅ dv . (14)

The third term in eq. (6) can be neglected because it is very v
small compared to the first term. If L>> R, eq. (7) is more
simplified and changed to Here, E3 is the electric field producing the charge –Q2 on the
upper ring, E4 is the electric field producing the charges Q2 on
b⋅a the lower ring electrode, ρ3 is the ion density near the upper
Cab = 4πε 0 . (8)
b−a radial electrode (the ions are produced at the edge of the
polarization rod at the center of the module), and ρ4 is the ion
 Journal of Electrical and Electronic Engineering 2015; 3(4): 76-86 79

density near the lower ring electrode (the ions are produced at 3. Experimental Setup
the edge of the upper ring electrode). Here, the polarity of ρ3 is
positive and that of ρ4 is negative. From eq. (13) and eq. (14), 3.1. Propulsion Model
comparing them to eq. (1), it is concluded that the propulsion Three types of model A units, which were actually
forces that both model 1b and model 1c generate should be fabricated in this work, are shown in Fig. 2. These models,
modified. Here, it should be noted that it is not always which produce propulsion force, can be set in series or parallel.
necessary to exchange the electrons between the electrodes I also fabricated two other types of models. In one, the ion
when the forces are produced. In other words, the current is wind flows downward in the ring electrode. In the other, the
not needed to generate the propulsion forces from the ion wind expands in the radial direction and flows outside the
polarization electrodes. ring electrode. In all the A model units, the ion wind flows
Finally, the ratios of generated power to electrical input inside the ring electrode (Fig. 2). The distance between the
power for model 1b and 1c are given by θ2=F2/P, andθ3=F3/P , negative needle electrode and the positive large ring electrode
respectively. in the vertical direction was set to be 4 cm in each case.

(a) (b) (c)

Fig. 2. Enhanced EHD and electrostatic propulsion devices. Model A units: (a) normal Aa, (b) with added inner ring electrode Ab, (c) with added radial
electrode Ac.

Table 1. Model A parameters. 1.5g, and 1.8g, respectively.

The devices were set in series to enhance the propulsion
Diameter of
Diameter Height Height of Inner force for the same electrical input power. In Fig. 3 depicting
Model Inner Ring
D (cm) L(cm)
D1(cm)
Ring L1(cm) model B1, the cascading of two devices is shown at the left
Aa1 8 3 － － side and the cascading of n devices is shown at the right.
Aa2 16 5 － － Model at the left side in Fig. 3 was used in this experiment. For
Ab 16 5 8 5 the model at the left side, the distance between the upper ring
electrode and the lower ring electrode was set to be 3 cm. A
The aluminum foils used as large electrodes were 10 µm in wire was connected on the upper ring electrode and the end
thickness. To improve the capacitance of the devices, various was closed to the lower ring electrode with a gap of 1 cm. It
kinds of electrodes were added inside the large ring electrodes. was normally difficult to connect the devices in series because
The Aa model has a single ring electrode, the Ab model has in some cases the propulsion forces disappeared due to ion
another ring electrode inside the outer ring electrode, and the flow or the positive and negative electrode settings. The upper
Ac model has a star electrode inside the outer ring electrode. ring electrode generates polarization. Negative charges are
These models were used to conduct an experiment to produce collected at the edge of the wire connected on the upper ring
high propulsion forces. The size parameters are shown in electrode. The weights of the model B1 was1.1g.
Table 1. The weights of the model Aa1, Aa2 and Ab were 0.6g,
80 Taku Saiki: Enhanced EHD and Electrostatic Propulsion Devices Based on Polarization Effect Using
Asymmetrical Metal Structure

Fig. 3. Enhanced EHD and electrostatic propulsion devices. Model B1.

Model B2 (Fig. 4) has a metal wire at the center of the propulsion forces, which work on both the upper ring
module to generate polarization. Positive ions are produced electrode and the lower ring electrode, are generated. The
from the upper edge of the wire, generating a propulsion force weights of the model B2 was 3.2g. The model is adapted to
to the upper ring electrode by the positive ion flow. Thus, two Fig. 1(c).

(a) (b)
Fig. 4. Enhanced EHD and Electrostatic propulsion devices Model B2. (a) Structure, (b) photo image.

In model B3 (Fig. 5), the ion wind flows in the radial ring electrode and finally the current is supplied from a high
direction. Enhanced propulsion force is produced by the voltage source. The structure is proposed because an EHD
polarization effect. The model is adapted to Fig. 1(b). Four generator should be set in the center of the device. In the future
small plate electrodes are set in the radial direction inside the such generators will use a weakly ionized plasma (such as a
outer large ring electrode. Negative ions generated from the flame) to generate an ultra-high voltage [15].The weights of
upper negative needle electrode are trapped the outer large the model B3 was 3.8g.
 Journal of Electrical and Electronic Engineering 2015; 3(4): 76-86 81

(a)

(b)
Fig. 5. Enhanced EHD and Electrostatic propulsion devices Model B3. (a) Structure, (b) photo image.

The structure parameters of model B are shown in Table 2.

Table 2. Model B parameters.

Diameter of Ring Height L Diameter of Inner Height of Inner (Upper) Distance between
Model Gap (cm)
D(cm) (cm) (Upper) Ring D1(cm) Ring L1(cm) Top and Bottom Rings(cm)
B1 8 3 8 3 1 3
Upper 2
B2 16 5 8 3 9
Lower 3
B3 32 5 8 3 1 －

generate a sinewave with a high output voltage and 25 KHz

3.2. Measurement of Propulsion Force frequency. The output voltage was evaluated by using a
The instrument used to measure the propulsion forces of the voltage tester (CD731a, SANWA, Japan) and a high voltage
EHD devices is shown in Fig. 6. probe (HV-60SANWA, SANWA, Japan). The maximum
The propulsion devices were connected to a DC high measured output voltage of the Cockcroft Walton circuit was
voltage generator. A six stage Cockcroft Walton circuit was 48 kV. The current was evaluated by connecting 22k ohm
used to generate the voltage. A sinewave generator was resistors in series to the propulsion devices and measuring the
connected to the Cockcroft Walton circuit. In this experiment, voltage. A dielectric plastic rod with a diameter of 1.0 cm was
a DC power supply (PR36-3A, TEXIO, Japan) was used to set on a gravimeter (ACS-20, ASONE, Japan). The propulsion
force was measured by reading the decrease in the weight.
82 Taku Saiki: Enhanced EHD and Electrostatic Propulsion Devices Based on Polarization Effect Using
Asymmetrical Metal Structure

The calculated results of the capacitance for each model are

shown in Fig. 8. The dashed line shows the results for model
1a, the dotted line shows the results for model 1b, and the solid
line shows the results for model 1c. a and a1 are all set to be
0.005 [m]. b was set to be R-0.005[m]. To simplify the
calculation, L was set to 0.03 [m], assuming that R1 is
adequately longer than R. n was set to be 4 and RG was set to
be 1 cm in the calculation. The capacitances for model 1a and
1b degraded as R increased. However, the capacitances for
model 1c increased as R increased due to the structure.

Fig. 6. Instrument for measuring propulsion force of the enhanced EHD

propulsion devices.

4. Results
4.1. Calculated Capacitance and Charge on Electrodes

The calculated results on the capacitance of model 1a are

shown as a function of L in Fig. 7. Here, R is set to be 4 cm and
is fixed at that length. It was assumed that the calculating
space is in the vacuum. Fig. 9. Calculated charge results.
The results clarified that the capacitance for model 1a,
shown in Fig. 1(a), the capacitance increased gradually and The calculated results for the total charges on the large
saturated as L increased. electrode are shown in Fig. 9. The dashed line shows the
results for model 1a. The dotted line shows the results for
model 1b, and the solid line shows the results for model 1c.
For the input voltage to the propulsion devices, I consider the
discharge property between the needle and flat electrodes, i.e.,
the limit of 15 [kV] per cm for the discharge determined [14].
Thus, the maximum input voltage was given by an equation of
15 x R [kV]. For model 1c, the calculated charge was obtained
by adding the charge on the upper ring electrode to that on the
lower ring electrode. For model 1b, it was obtained by adding
the charge on the inner ring electrode to that on the outer large
ring electrode. The charges for model 1a, 1b, and 1c increased
as R increased. The charges for model 1c were the largest for
all the models.
Fig. 7. Calculated capacitance for model 1a as a function of L.
4.2. Measured Propulsion Force

The results obtained for the measured propulsion force for

model Aa1, Aa2, and Ab are shown in Fig. 10. The black
circle dots show the results for model Aa2. The white circle
dots shows the results for model Aa1. The white square dots
shows the results for model Ab. The data shown by the dots
in Fig. 10(a) is consistent with that shown by the dots in Fig.
10(b). The calculated electrical input power in Fig. 10(b) was
obtained by multiplying the measured current data. When the
input voltages were 22 and 26 kV, the measured current
levels were 0.15 and 0.26 mA, respectively.
When the electrical input power was 8 W, it can be seen
that the force that model Aa2 generated was 2.5 times higher
than that of model Aa1. The maximum generated force was
Fig. 8. Calculated capacitance results for each propulsion model.
1.8 g at an input voltage of 43 kV. However, the generated
 Journal of Electrical and Electronic Engineering 2015; 3(4): 76-86 83

force was gradually saturated as the input voltage and the input voltage of 43 kV. In spite of using a long L, the
electrical input power increased. In another experiment, the generated force was only slightly improved.
force measured for model Aa2 for L= 10 cm was 2.0 g at an

(a) (b)
Fig. 10. Measured propulsion force results for model Aa1, Aa2, and Ab. (a) Voltage-force property, (b) Power-force property.

(a) (b)
Fig. 11. Measured propulsion force results for model B1. (a) Voltage-force property, (b) Power-force property.

When the electrical input voltage was 44 kV, it can be seen The measured propulsion force results obtained for model
that the propulsion force that model Ab generated was 1.6 B2 are shown in Fig. 12. If the generated power values shown
times higher than that of model Aa1. The maximum generated in Fig. 10(a) are compared with those shown in Fig. 12(a),
force was 2.8 g at an input voltage of 44 kV. However, the when the electrical input voltage was 43 kV, it can be seen that
generated force was not saturated as the input voltage and the the force that model B2 generated was 2 times higher than that
electrical input power increased. In another experiment, in of model Ac. The maximum generated force was 3.6 g at an
which a star electrode and two inner multi-rings were added to input voltage of 43 kV. The generated force increased as a
model Aa2, the maximum generated forces for them were 2.3 function of a value close to the square of the input voltage as
and 3.2 g, respectively. shown in Fig. 12(a).
The measured propulsion force results obtained for model Measured propulsion force results obtained for model B3
B1 are shown in Fig. 11. If the generated power values shown are shown in Fig. 13. The maximum generated force was 5.1 g
in Fig. 10(a) are compared with those shown in Fig. 11(a), at an input voltage of 48 kV. The generated force increased as
when the electrical input voltage was 46 kV, it can be seen that a function of a value close to the square of the input voltage as
the force that model B1 generated was 3 times higher than that shown in Fig. 13(a). If the generated power values shown in
of model Aa1. The maximum generated force was 2.3 g at an Fig. 10(a) are compared with those shown in Fig. 13(a), it can
input voltage of 44 kV. The generated force markedly be seen that the maximum generated force that model B3
increased at an input voltage of 38 kV as shown in Fig. 11(a). generated was 6.4 times higher than that of model Aa1.
84 Taku Saiki: Enhanced EHD and Electrostatic Propulsion Devices Based on Polarization Effect Using
Asymmetrical Metal Structure

(a) (b)
Fig. 12. Measured
easured propulsion force results for model B2. (a) Voltage-force property, (b) Power-force
force property.

(a) (b)
Fig. 13. Measured
easured propulsion force results for model B3. (a) Voltage-force property, (b) Power-force
force property.

the surface of the large flat electrodes,

electrodes it can be concluded that
5. Discussion the polarization effect improves the forces.
I proposed modified EHD and electrostatic devices to The calculated results lead one to expect that the force that
improve the ratio of the propulsion force to electrical input model B3 generates will be 7 times higher
high than that of model
power. In the numerical analysis, the results of the calculated Aa1 owing to eq. (1) and eq. (13).
(13 If the devices are operated
total charge on the surface of the large electrode were enlarged at a high voltage, the gap used will be longer, the stored
due to the polarization effect. From the he propulsion force charges on the large and flat electrodes
electrode will be larger, and the
results obtained for the devices governed by eq. (1) and eqs. quantity of the charges will be enlarged proportional to the gap
(13) (14),
), because the forces are proportional to the charges on length approximately.
Table 3. Comparison of measured and calculated propulsion force.

Model Voltage（kV） Current （mA） Power (W) Fexp(g) d (m) Fcal(g) Eq.(3) α Fexp/ Fcal θ (N/kW)
Aa2 42 0.15 9.9 1.8 0.09 1.1 1.6 1.8
Ab 45 0.39 19 2.8 0.09 1.7 1.6 1.4
B1 46 0.31 14 2.2 0.066 0.94 2.4 1.6
B2 43 0.42 18 3.8 0.09 1.8 2.1 2.1
B3 48 0.26 12 4.6 0.066 0.81 5.7 3.8

In the experiment on the forces that model A and B The gap length d is the value for which 1 cm was added to
generated, the models showed high performance with respect the gap length from the edge of the radial electrode to the
to the ratio of force to electrical input power. negative needle electrode in Table 3. I used eq. (3) to calculate
Especially for model B3, A large propulsion force are the theoretical values of the propulsion force the devices
generated whenever ion flows in the horizontal direction. The generated.. The ion mobility was set to be 2.15 x 10-4 (m2/Vs).
ion wind cannot be contribute to the enhanced propulsion The ratio of generated force to electrical input power of a
force. This should prove that the lifter work by the basic lifter with a single narrow electrode and a single flat
electrostatic propulsion as mentioned in the reference[6,7]. aluminum electrode, as shown in reference [1,2], was
The electrical
lectrical input power, measured generated power, estimated to be 1.3 N/kW.
calculated generated force, and ratio are shown in Table 3. It has been clarified from this experiment that the total
 Journal of Electrical and Electronic Engineering 2015; 3(4): 76-86 85

amount of ions generated from the small electrode in the the experiment was 10 µm; this thickness should have an
conventional lifter model is large. For a large, flat electrode, optimized value.
however, the ion density is low. If the aluminum foil is very thin, the potential that confines
The 0.4 mA current for model Aa2 with 16-cm outer the charge will degrade and thus charges on the metal plate
diameter is the same as that of the basic lifter at the input cannot be confined efficiently. Basically, the structure of the
voltage of 46 kV. However, the generated force for model Aa2 large electrode must be maintained to store a lot of charges. In
is 1.6 times higher than that of the basic lifter as shown in the future, the optimized thickness should be studied.
Table 3. The model A structure with downstream ion flows Modifications should also be made to obtain a higher ratio
was able to confine ions generated from the negative electrode of the generated force to electrical input power. Using multiple
and prevent the ion density from being low close to the large large, flat electrodes as positive electrodes should make it
radial electrode. possible to improve the ratio, since it should enhance the
The forces that the models B1, B2, and B3 generated are electrical field intensity. The experimental data in reference
rigidly proportional to the square of the input voltage because [12] shows the effect to improve the ratio of generated force to
the electric field intensity increases when the number of ions electrical input power clearly. However, the effects of
generated increases at the same time. enhancing the electric field intensity due to the edge and flat
The forces that model Aa generated were rigidly generated surfaces of electrodes are mixed. Thus, we cannot divide them.
in proportion to the input voltage because the electric field In reference [12], a 6.5 N/kW ratio was obtained when the
intensity is constant when the number of generated ions small electrode was negative s in a high voltage and high
increases. power regime. The models I showed may be optimized, in
The forces that models Aa1 and Aa2 generated saturated which case it should be able to obtain a higher ratio than this.
when the input voltage was 43 kV. This is because the It is hard to set B2 modules in cascade but it is possible to set
generated ions were saturated in the air, the ion density them in parallel. Models B1 and B3 can be set in series.
consequently reached the space charge limit, and the ions were Setting devices in series, as shown at right in Fig. 3, will result
repelled near the large electrode. in an increased ratio of generated force to electrical input
For model Ab, the inner ring electrode was set in the center power. Using a multi-electrode structure comprising a large
of the large ring electrode and the measured generated force flat electrode, multi-stage polarization electrodes, and using
was more than twice that measured for model Aa2. The reason long gap between electrodes and more high input voltage [13]
is that the surface area of the positive large electrode increased. should make it possible to raise the ratio to over 100N/kW at a
However, a model having too many ring electrodes will not low electrical input power and less 100N/kW at a high
result in an increase in the generated force. electrical input power. Using long gap between electrodes
For model B1 using the polarization effect, the ring results in improving the ratio of generated force to electrical
electrodes were cascaded and the force measured for it was 3 input power as shown in eq. (4) because the electric field is
times as large as that measured for model Aa1. weak for the same input voltage [13]. The ratio at a high
For model B2 using the polarization effect, there was a pole electrical input power would be higher than that of a
to generate polarization. The force measured for model B2 conventional helicopter using a motor, for which the ratio is
was close to 2 times larger than that measured for model Ac. around 70N/kW.
The force measured for model B3 was 4.6 g at an input Combining model B3 with model B1 will result in
voltage of 46 kV and electrical input energy of 12 W. The ratio generating a force of over 10 g. Moreover, using weakly
of the generated force to electrical power was 4N/kW. The ionized plasma to produce ultra-high voltage, as in EHD
value was maximum in all the proposed models. The current generators [15], we should use the optimized system to reduce
of model B3 was 50% lower than that of model Aa1 for the the weight of the high voltage power supply and improve the
same input voltage of 24 kV. The ratio of the generated force ratio of generated force to electrical input power.
to electrical input power was 15N/kW at 23 kV input voltage.
This is one order higher than that of the basic lifter, for which 6. Conclusion
the ratio was 8.3N/kW at 28 kV input voltage. The generated
force for input voltage was not saturated. The property is ideal This paper discussed various kinds of proposed EHD and
and well consistent with the theory. The polarization plate electrostatic propulsion devices using asymmetrical
electrodes should be set to be tilted for the large rounded polarization electrodes that I propose as means to improve the
electrode. The maximum angle for generating force should ratio of generated propulsion force to electrical input power.
exist. In these experiments, I was able to modify the ratio of The levels of force that the devices generated were measured
the generated force to the electrical input power. in experiments. Electrical charges on the surfaces of
However, the ratio of the generated force to unit weight has electrodes were estimated numerically with an aim to
not been modified yet. Using a high voltage generator with improving the generated force. The models substantially
higher output voltage and longer gap length will enable the improved the generated force; for the same electric energy, the
ratio of generated force to unit weight to be improved. force they generated was 5.7 times higher than that of a basic
Specially, the ratio of the models B1, B2, and B3 should be type lifter. This was due to additional propulsion force being
substantially modified. The aluminum foil thickness used in generated by the polarization effect.
86 Taku Saiki: Enhanced EHD and Electrostatic Propulsion Devices Based on Polarization Effect Using
Asymmetrical Metal Structure

Future subjects in this research will include using 1) a high [8] R. Ianconescu, D. Sohar, and M. Mudrik, “An analysis of the
voltage supply with reduced weight achieved by using a Brown- Biefeld effect,” J. of Electrostatics, 69, pp. 512–521
(2011).
flame-jet or low-density-plasma EHD generator and 2)
multiple polarization electrodes as means to improve the [9] M. Chen, L. Rong-de, Y. Bang-jiao, “Surface aerodynamic
generated propulsion force. model of the lifter,” J. of Electrostatics, 71(2), pp.134–139
(2013).
[10] F. X. Canning, C. Melcher, and E. Winet, “Asymmetrical
References Capacitors for Propulsion,” NASA, NASA/CR—2004-213312
(2004).
[1] T. T. Brown, “Electrokinetic Apparatus,” U.S. Patent
N°2949550, 1960. [11] J. Wilson, H. D. Perkins, and W. K. Thompson, “An
Investigation of Ionic Wind Propulsion,” NASA,
[2] http://www.jlnlab.com/ NASA/TM—2009-215822 (2009).
[3] http://www.blazelabs.com/ [12] M. Eiant and R. Kalderon, AIP advances, 4, 077120-1-20
(2014).
[4] M. Tajmar, “Biefeld–Brown Effect: Misinterpretation of
Corona Wind Phenomena”, AIAA Journal 42(2) (2004). [13] K. Masuyama and S. R. H. Barret, “On the performance of
electro hydrodynamic propulsion”, Proc. of Royal Soc. A, 469
[5] L. Zhao and K. Adamiak, “Numerical analysis of forces in an (2154), 20120623-1-16 (2013).
electrostatic levitation unit,” J. of Electrostatics, 63, pp.
729-734 (2005). [14] T. Kouno, “High Voltage Engineering”, Asakura Publishing,
Tokyo (1995) (Chapter 2, in Japanese).
[6] L. Zhao, K. Adamiak, “EHD gas flow in electrostatic levitation
unit”, J. of Electrostatics, 64, pp. 639–645 (2006). [15] T. Saiki, "Study on High Voltage Generation Using Flame
Column and DC Power Supply", J. of Electrostatics, 70,
[7] A. A. Martins, M. J. Pinheiro, “Modeling of an EHD corona pp.400-406 (2012).
flow in nitrogen gas using an asymmetric capacitor for
propulsion”, J. of Electrostatics, 69 (2), pp. 133–138 (2011).