DECEMBER 1966 ENGINEERING NOTES 1803
Table 3 Values of the correction factor /4 for increasing
pressure trace with initial pressure b
k = AP/5 n = 0.25 n = 0.50
0.5 0.999 0.998
1.0 0.996 0.995
2.0 0.992 0.989
In order to arrive at a closed-form solution, we shall examine
only the case for which m = 1. The result is
[(b
f = cPn (15)
(n + 1)AP(6 + AP/2)»
where AP = atb, i.e., the difference between the final and
initial pressure. An alternate representation of Eq. (15) is
[(I + - 1]
(16)
Downloaded by UNIVERSITY OF MICHIGAN on February 10, 2015 | http://arc.aiaa.org | DOI: 10.2514/3.28756
(n
where k s= AP/6, i.e., the ratio of the pressure change to the
initial pressure. The correction factor /4 is independent of a
and h, but dependent on their product, and it is dependent on
the initial pressure only as it affects k. Values of /4 are shown
in Table 3.
The values in Table 3, as compared to those in Table 2 for
m = 1, show that the difference between fobs and fcaic are de-
creased to approximately 1% because of the introduction of
the b term corresponding to the initial pressure, which would
generally be encountered in practical situations. The same
results are obtained for negative AP values, due to symmetry
considerations similar to those discussed for case 3.
Reference
1
Hewitt, E. and Stromberg, K., Real and Abstract Analysis
( Springer-Verlag Inc., Berlin, Germany, 1965), p. 191.
Electrostatic Generators in Space
Power Systems
DOMINIQXJE GlGNOUX*
Cosmic Inc., Washington, D. C.
Introduction
E LECTROSTATIC generators have been considered for
space power systems for a number of years. The first
models built, however, have been disappointing in many
respects. With the construction of newer designs, prospects
have developed which may change substantially our approach
to the space power problem. In this paper, we will review
the general operating principles of electrostatic generators,
analyze the features of models currently under development,
and describe the design which seems to be the most promising
for use in space power systems.
Principles of Operation of Electrostatic Generators
When an electric field E exists at the surface of a conduct-
ing material, there exists a force acting upon an element of
surface dS which is perpendicular to the surface. This is
described as a pressure called electrostatic pressure, P =
(e/2)#2, where e is the permittivity of the dielectric medium
Presented as AIAA Preprint 64-450 at the 1st AIAA Annual
Meeting, Washington, D. C., June 29-July 2, 1964; revision
received July 29, 1966.
* President. Associate Fellow AIAA.
Fig. 1 Electrostatic pressure on element of surface.
in front of the surface. If the surface involved is that of the
rotor of an electrostatic generator, the power produced per
unit area of rotor is equal to the scalar product of the two
vectors representing the velocity and the pressure, W/A =
P-V (See Fig. 1). Two methods can be used to explain the
operation of electrostatic machines or to calculate the power:
1) the method of field-theoretic analysis or 2) the method of
capacitance analysis. The first method consists in applying,
the formula
W = P-V (D
to the total surface of the machine. This would give the exact
amount of energy being converted at any given time. Such
analysis may lead to extreme complications, and it is usually
more convenient to use the method of capacitance analysis.
In an electrostatic machine using conducting charge car-
riers, one can use the concept of potential energy of each
charge carrier and describe its variation in terms of the varia-
tions of the capacitance of the charge carrier with respect to
the stator members. In this type of analysis no integration is
required since the energy conversion per cycle may be prop-
erly defined by the minimum and maximum capacitance of
the charge carrier with respect to the stator.
Oblique Field Generator
As we have seen, capacitance analysis is a convenient
method of calculating the performance of an electrostatic
generator. However, capacitance calculations do not show
the quantities which, in effect, limit the power of the electro-
static generator. Such quantities are determined, generally,
by the ability of the vacuum to withstand dielectric stresses.
Field-theoretic analysis, on the other hand, keeps in focus the
electric field on the surface of the electrodes and, therefore, is
useful in the determination of the best design. As a result of
field-theoretic analysis, the sharp edges of the foil used in the
Wimshurst machine were replaced by thick segments. The
same thinking led to generators with many charge carriers
and then to the constant oblique field generator.1
Figure 2 shows a cross-sectional sketch of a conducting-
carrier generator and the lines of force in the vicinity of a
charge carrier. It should be noted that P is tangent to the
lines of force. Therefore, it can be seen that on most of the
surface of the rotor blade, the product P-V, which is the
power/unit area of charge carrier, is very small or nil; only
the edges of the rotor blades contribute to energy conversion.
It appears, then, that the conducting-carrier generators we
have seen so far effectively utilize only a small percentage of
the rotor surface.
The first improvement which stems from the preceding con-
siderations is that of utilizing a multiple-capacitor design
and replacing the blades of the rotor by rods in the manner
shown in Fig. 3. One method for transforming thejliscon-
Fig. 2 Single-capacitor, two-element machine.
 1804 J. SPACECRAFT
ISTATOR
I ROTOR —
~_ STATOR
INDUCTOR
Fig. 3 Constant oblique field machine.
tinuous forces in the gap into continuous ones consists of em-
bedding the rods in an insulating material as shown in Figs.
3 and 4. If the rods are in sufficient number, the field in the
rotor-stator gap approaches a uniform field. The main limi-
tation of electrostatic machines is the breakdown in the rotor-
stator gap. If one considers only the gap, it appears that the
generator with embedded rods may be treated in a fashion
similar to that of an insulating-carrier machine. For example,
let <j be the average charge density/rotor area (this charge is,
in fact, located on the rods) and N and T, the normal and
tangential components of the field at a point just off the rotor
Downloaded by UNIVERSITY OF MICHIGAN on February 10, 2015 | http://arc.aiaa.org | DOI: 10.2514/3.28756
surface (see Fig. 3). In the rotor-stator gap we then have
o- = 2eN (2)
and the power/unit area becomes
W/A = 2eNTV (3)
with a force/unit area having a tangential component
F/A = 2eNT (4)
where A is the rotor area, not the charge-carrier area.
The determination of the maximum field is a complicated
one, since the components N and T of the field should be con-
sidered separately. As a practical matter, it appears that the
normal field in pressurized gases or vacuum may be as high as
50 kv/mm, whereas the tangential field is usually limited to
about 3 kv/mm. Under these circumstances, the maximum
theoretical value of F/A, defined by the preceding equations,
is 2650 newtons/m2. In the next section, F/A will be defined
from the measurement of the total power of a generator, and it
will be seen that the value of the product NT obtained by ex-
periments accomplished thus far is still two orders of magni-
tude below the theoretical maximum.
Normalized Power
In general, electrostatic energy conversion can be described
as the work that results from rotating a force/unit area about
an axis. The power is, therefore, the product of three dif-
ferent factors for a rotor of radius R, rotating at n rpm: 1)
the average force/unit area, i.e., a characteristic of the gen-
erator design; 2) the surface on which the force is applied,
which is proportional to E2; and 3) the velocity of such a
surface, which is proportional to Rn. One can therefore
assume that the power of a generator is proportional to R*n
and that the coefficient of proportionality is the normalized
power times a constant. In other words, the total power of a
generator is obtained by inserting a parameter descriptive of
Fig. 4 Exploded view of
constant oblique field
generator.
VOL. 3, NO. 12
the design (normalized power) in a formula that depends on
the geometry. For instance, the power of a disk generator is
as follows:
W - (5)
where p = normalized power. Conversely, the normalized
power p is calculated from experimental data by the general
formula
total power
p =
area X average linear velocity
Therefore, the normalized power is an index for rating
electrostatic generator designs. It is based on "raw power"
as obtained from experiments, but it enables useful informa-
tion to be derived from small-scale, inexpensive experiments.
The generator that would seem to have the highest normal-
ized power is the one described by Vorobyev,2 which is a
multiple-capacitor generator operated in air at 10 atm and
for which p = 260 newtons/m2. Oblique field generators
built by Cosmic Inc. have a normalized power of 25 newtons/
m2 in air and 200 newtons/m2 in pressurized hydrogen.
Vacuum tests have so far reached only 30 newtons/m2. This
last figure is abnormal since power in vacuum should be higher
than power in pressurized hydrogen, and is due to difficulties
in commutation. Single capacitor generators have a nor-
malized power of about 15 newtons/m2 in vacuum.3'4 These
values are summarized in Fig. 5.
The constant oblique field generator thus far described
utilizes brushes for commutation.5 Its performance in a
vacuum is limited, therefore, as there exists a voltage drop
between the brushes and the charge carrier. A new design
has been evolved which utilizes rectifiers as commutation
means. In the design shown in Fig. 6, each charge carrier is
connected to the input and the output terminals by rectifiers.
These rectifiers are so oriented that when a charge carrier is
adjacent to the input inductor, the excitation voltage induces
VACUUM PRESSURIZED GAS
(AIR AT 10 ATM)
OBLIQUE FIELD
AT 15 AT
. 5 0 SINGLE
CAPACITOR
(SF6 )
27 []
£3 20
16 ( „' SINGLE
CAPACITOR
4.7
HH Constant Oblique.Field Generator (Ref. 1 "& 5)
\) Russian Multiple-Capacitor Generator (Ref. 2)
/\ Single-Capacitor, Three-Element Machine (Ref. 4)
O Single-Capacitor Generator (Ref. 3)
Fig. 5 Comparison of normalized power of four different
generators.
 DECEMBER 1966 ENGINEERING NOTES
Output Inductor
Input Inductor
Input Terminal Output Terminal
Fig. 6 Principle of operation of the brushless constant
oblique field electrostatic generator.
Downloaded by UNIVERSITY OF MICHIGAN on February 10, 2015 | http://arc.aiaa.org | DOI: 10.2514/3.28756
charges from the input terminal onto the charge carrier.
Conversely, when the charge carrier is adjacent to the out-
put inductor, charges flow from the charge carrier to the out-
put terminal.
The principle of operation may also be easily understood
from the point -of view of capacitance variation. It will suf-
fice to consider the expression "adjacent to the input induc-
tor'7 as meaning "in a position of minimum capacitance with
respect to the input inductor and of maximum capacitance
with respect to the output inductor/7
In the construction utilizing brushes, the charge carriers are
mounted on the stator. This construction is no longer prac-
tical here, since it would entail rotating the rectifiers and using
slip rings to connect the rectifiers to the input and output ter-
minals. It is simpler to rotate the inductors and keep the
charge carriers stationary. Therefore, in the following we
shall always refer to the inductor holder as the rotor and to
the charge carrier holder as the stator.
The only moving contacts in this configuration are slip
rings connecting the excitation power supplies to the induc-
tors. These slip rings are not subject to any current except
that required to compensate for the leakage between inductors,
and are not absolutely necessary. In a large-scale machine,
built for long duration, it may be more advantageous to use
radioactive batteries to supply excitation power.
Conclusions
The single-capacitor generator, although an interesting re-
search and training tool, does not seem capable of competing
with homopolar magnetic generators in the production of
space power. The constant oblique field generator, on the
other hand, shows promise of being efficient, light in weight,
and capable of a power-to-weight ratio of the order of 5 to 10
kw/lb. Tests with this type of generator are confirming
earlier estimates. Further work is needed in the area of com-
mutation and material selection. Experiments conducted
so far have shown that designs utilizing metal surfaces which
are at a high voltage and exposed to a vacuum environment
have low reliability. To obtain high power densities, insu-
lating or semiconducting dielectric surfaces must be used.
The brushless design will be built in the near future with
the goal of providing power at high voltage in the 5 to 40 kv
range.
References
1
Gignoux, D., "Constant oblique field electrostatic generator,'7
ARS Preprint 2556-62 (1962).
2
Vorobyev, A. A., Vorobyev, G. A., et al., Vysokoye ispytatel
'noye oborudovanive i izmereniya (State Publishing House of
Energetics, Moscow-Leningrad, 1960) p. 110.
3
Denholm, A. S., Trump, J. G., and Gale, A. J., "High volt-
age generation in space: The parametric electrostatic machine,77
ARS Progress in Astronautics and Rocketry: Energy Conversion
1805
for Space Power, edited by N. W. Snyder (Academic Press Inc.,
New York, 1961), Vol. 3.
4
Pauli, D. C. and Rebman, J. A., "Tests of a conducting
carrier-type electrostatic generator," Contract AF 33 (616),
Pneumodynamics Corp., Systems Engineering Division, Bethesda,
Md., TR 6636-1 (January 1962).
5
Anton, H. F., Gignoux, D., and Shea, J. J., "Constant
oblique field generator," Contract AF 33(657)-7769, Cosmic Inc.,
ASD-TDR-63-87 (February 1963).
A Method for the Preliminary Design
of Radiative Heat-Shield Panels
SANTOS SHIELDS*
LTV Aerospace Corporation, Dallas, Texas
A PRELIMINARY design method has been developed
which permits the calculation of bulk insulation and in-
sulative washer thicknesses for a heat shield (with cooled sub-
structure) subjected to a step thermal environment and with
specified temperature limits for the materials. A primary
object is to include the thermal resistances of standoff at-
tachment members (Fig. 1), which fix the bulk insulation
thickness. The bulk insulation thickness in turn affects the
height, and hence the bending moments, of the standoff mem-
bers. An iterative procedure, combined with minimization
techniques, produces a family of preliminary thermal design
optimization curves. These curves are supplemented by
structural design curves which permit some flexibility in the
panel design without exceeding the maximum stress limita-
tions for the standoff attachment members. The optimum
design may be obtained on the basis of either minimum system
weight or the minimum weight for panel sizes larger than
those required for minimum system weight. Panel size is re-
stricted by the bending loads introduced by thermoelastic
effects on the standoff members which connect the hot ex-
terior skin to the cool substructure
The steps are outlined in Fig. 2. Three design conditions
may result: both high-temperature and low-temperature
Fig. 1 Schematic of uncooled heat-shield panel with
attachment.
Presented as AIAA Preprint 66-507 at the AIAA 4th Aero-
space Sciences Meeting, Los Angeles, Calif., June 27-29, 1966;
submitted July 19, 1966; revision received September 12, 1966.
The author wishes to thank J. E. Medford and V. C. Igou for
their contributions in the thermal and structural equations de-
velopment. This work was supported by Air Force Flight Dy-
namics Laboratory, Research and Technology Division, Wright
Patterson AFB, Ohio, under Contract AF33(615)-1261.
* Engineering Specialist, Astronautics Division.