Content uploaded by Francesco Romano
Author content
All content in this area was uploaded by Francesco Romano on May 10, 2016
Content may be subject to copyright.
SP2016 3124981
INTAKE DESIGN FOR AN ATMOSPHERE-BREATHING ELECTRIC PROPULSION
SYSTEM
Francesco Romano (1), Tilman Binder (1) , Georg Herdrich (1), Stefanos Fasoulas (1), Tony Sch¨
onherr (2)
(1) Institute of Space Systems (IRS), University of Stuttgart, Pfaffenwaldring 29, 70569, Stuttgart, Germany,
romano@irs.uni-stuttgart.de
(2) ESA/ESTEC, Keplerlaan 1, 2201 AZ Noordwijk, The Netherlands, tony.schoenherr@esa.int
KEYWORDS
Atmosphere-Breathing Electric Propulsion - ABEP
- RAM-EP - VLEO - Inductively Heated Plasma
Thruster - IPT - Intake - EFD
ABSTRACT
An atmosphere-breathing electric propulsion system
(ABEP) [1] ingests the particles of the residual
atmosphere and uses them as propellant for an
electric thruster to counteract the drag. The
system theoretically allows orbiting for unlimited
time without on-board propellant storage. A new
range of altitudes, e.g. 120-250 km in LEO, for
permanent orbiting can be accessed, thereby
enabling new scientific missions while reducing
costs. ABEP can be conceptually applied to any
planet with atmosphere. The intake is the device
that collects the atmosphere particles and drives
them to the thruster. Its nontrivial design requires
deep understanding of the flow physics. In this
paper, we present the outcome of our Balancing
Model (BM) [2] applied to intake designs of JAXA
and BUSEK. Optimization of the intake toward the
use of IPG6-S, a small scale inductively heated
plasma generator (IPG) as thruster candidate, has
been performed and results are hereby presented.
1 ATMOSPHERE-BREATHING
ELECTRIC PROPULSION
Very low Earth orbits are of great interest for many
scientific, civil, and military purposes. Recently
ESA’s mission GOCE ended; it provided detailed
information of the Earth’s geomagnetic field by
orbiting as low as 229 km [3] using ion thrusters to
compensate the drag. The amount of propellant
on board is a limiting lifetime factor for such a
mission, in particular if the S/C is orbiting very low
around a planet with atmosphere. The atmosphere
is indeed responsible for the drag, which slows down
the S/C and reduces its total mission lifetime. It
is also a limiting factor in terms of costs, as more
drag to be compensated for a longer time means
an increased amount of propellant to be carried
on-board, which again increases the total mass. The
lifetime of a S/C orbiting in LEO can be significantly
increased by applying an efficient propulsion system
that compensates the drag.
The basic idea of an ABEP is of using the particles
of the residual atmosphere as propellant and to
process them through a device for generating thrust.
This will decrease, ideally nullify, the on board
propellant requirement and will generate thrust to
partially or fully compensate the drag. A conceptual
scheme of a S/C with ABEP is shown in Fig. 1.
Inflow
Flight Direction Solar Array
Solar Array
Intake Exhaust
S/C Core
Figure 1: ABEP Concept
2 ATMOSPHERE COMPOSITION
In this section the atmosphere of Earth and Mars will
be analyzed as the environment dictates the design
of mission and propulsion system. Atmospheric
models and databases provides the composition of
particles collected by the propulsion system, their
density, temperature and pressure profile provides
information for the S/C and mission design.
2.1 Earth’s Atmosphere
Earth’s atmosphere extends from ground to space
and it is composed of various gases, mostly
nitrogen and oxygen. However, depending on
2.2 Mars’ Atmosphere 3 INTAKE
the altitude, species concentrations vary, as it is
shown within Fig. 2. The altitude range of interest
for ABEP is below 250 km for being competitive
against conventional EP according to [4] and above
120 km due to heating effects [5], [6]. To gain
the most reliable values, a proper atmospheric
model has to be chosen. The chosen model
for Earth is the NRLMSISE-00, compared to the
common MSISE-90, it provides better estimation
of the density below 350 km of altitude and it is
the most accurate model for residual atmosphere’s
composition in LEO and VLEO [7]. The data have
been generated through the NRLMSISE-00 model
website [7]. Inputs are the date, geographical
coordinates and solar activity values, F10.7and Ap.
Care must be taken regarding the solar activity that
cycles every 11 years. At the maximum solar activity,
the atmosphere will be compressed and, therefore,
for the same altitude density will be greater, the
opposite happens at the time of minimum solar
activity. This effect is more evident on higher than
on lower altitudes [5].
h, km
100 120 140 160 180 200 220 250
n, m -3
10 11
10 13
10 15
10 17
10 20
NRLMSISE-00 Mean Solar Activity
F10.7 = F10.7 avg = 140, Ap = 15 He
O
N2
O2
Ar
H
N
Figure 2: Earth’s atmosphere composition
From Fig. 2 can be seen that for the altitude of
interest for ABEP the main components are atomic
oxygen O and nitrogen N2with a not-negligible
contribution of O2and Ar, the latter at low altitudes.
2.2 Mars’ Atmosphere
Mars’ atmosphere is much thinner compared to
Earth’s and it is mainly composed of carbon dioxide,
with addition of oxygen and other species depending
on the altitude. The pressure at the surface is 1%
of that on Earth and the low density results in a
much less inertia of the atmosphere, which leads in
a quick respond to any change. The altitude range of
ABEP is closer to the planet’s surface compared to
the Earth case. According to BUSEK a maximum of
180 km is set. A minimum of 80 km is chosen, that
is a value close to the Karman line for Mars and
also represents the altitude at which dust particles
have been detected in the highest dust storm activity
period by S/C [8]. The temperature profile itself is
different than that on Earth, temperature decreases
with increasing altitude until about 100 to 120 km,
where the temperature starts increasing again due
to extreme UV heating [8]. To perform the analysis,
the Mars Climate Database (MCD) Version 5.2 has
been selected [9]. It can simulate up to 350 km in the
thermosphere and accounts for dust storm and solar
activity. According to Fig. 3, the particles ingested by
the ABEP system will be mostly CO2and O. A first
important consideration of using CO2as propellant
is that the energy for ionization will be higher than
for other gases such as N2or O2due to frozen
losses. Another important consideration is regarding
the probable ingestion of dust particles by the ABEP
system, this might influence the performance of the
system and eventually deteriorate the S/C surfaces
at first [10] and the ABEP system.
80 100 120 140 160 180
h, km
1012
1015
1017
1020
n, m−3
MCD 5.2, Mean Solar Activity
CO2Ar N2CO O3OO2HH2
Figure 3: Mars’ atmosphere composition
3 INTAKE
The intake is the device that collects and delivers the
atmosphere particles to the propulsion system. Its
design is non-trivial because of the flow condition in
which the S/C is orbiting. The probability of collisions
between particles is very low and the velocity of the
flow is, with good approximation, that of the orbiting
S/C. Considering Earth orbit, this is about 7.8 km/s,
for Mars orbit about 3.5 km/s, in the considered
altitude ranges. The flow is dominated by collision
at the walls rather than inter-particle collisions, as
the Knudsen number Kn =λ/L is high. The most
advanced approach of intake design for ABEP is of
including an inlet structure that let the atmosphere
3.1 JAXA 4 INDUCTIVELY HEATED PLASMA THRUSTER
particles to flow in and stop them from escaping after
they are reflected back by the intake surfaces. A
conceptual scheme of an intake for ABEP is shown
in Fig. 4. This approach of atmosphere trapping has
been developed in the most advanced intake designs
presented by BUSEK [11], [12] and JAXA [13], [14].
Inflow
Flight Direction
Thruster
Inlet structure
Figure 4: Intake Concept
A very important parameter needed to evaluate
the performance of the intake is the collection
efficiency ηc, see Eq. 1, and it represents the number
of particles that finally flows through the thruster,
˙
Nthr divided by incoming number of particles ˙
Nin.
ηc=˙
Nthr
˙
Nin
(1)
According to the analysis in [2], ηcfor the JAXA case
is of about 40% and of 20% for the BUSEK case, but
it has to be considered that ˙
Nin is reduced for the
JAXA case, as only a fraction of the total front area
is open (see Sec. 3.1).
3.1 JAXA
Fujita’s study [13] considers a by-pass-like design
in which atmosphere particles enter through a ring
section, as shown in Fig. 5. Particles reach the
back of the intake hitting a 45◦steep surface,
the diffuser/reflector, and are afterwards scattered
on the back of the satellite core and to the
thruster’s acceleration grids. An Electron Cyclotron
Resonance (ECR) device ionizes the particles in
the ionization chamber, that is the region on the
back of the satellite core. Ionized particles are
afterwards extracted through accelerating grids to
produce thrust.
Figure 5: JAXA’s intake design, [13]
3.2 BUSEK
BUSEK Inc. [12] studied the MArs Breathing Hall
Effect Thruster (MABHET), a S/C with an ABEP
system. In this design, the intake is a long tube of
3.7 m length and 0.6 m diameter, with a honeycomb
structure in the front composed of many small ducts.
The intake is designed as a long tube in order to
achieve a higher density region at the back part due
to the presence of an assumed ”collision cascade”
phenomena. A total pressure increment of 100 was
observed in DSMC [12]. The design is shown in
Fig. 6.
Figure 6: BUSEK’s intake design, [12]
4 INDUCTIVELY HEATED PLASMA
THRUSTER
An inductively heated plasma thruster (IPT), is a
concept of electric propulsion system that is based
on inductively coupled plasma sources (ICP). It is
mainly composed of a cylindrical discharge channel,
in which the propellant is fed and, afterwards, ionized
by induction. The induction is provided by an RF AC
current fed coil, as shown in Fig. 7.
5BALANCING MODEL
Figure 7: Inductive plasma
In detail, an axial time-varying magnetic field
is generated together with an azimuthal electric
field. They cause the gas particles to excite and
to release free electrons. A chain reaction is
established and plasma is formed. This has to
be accelerated and expelled from the propulsion
system to produce thrust. Some acceleration is
provided by Lorentz force, however, a proper stage
for plasma acceleration is needed to increase the
thruster efficiency and amount of thrust to a useful
level. The advantage of such a device in ABEP
application, when compared to the conventional
electric propulsion systems such as GIT and HET,
is the absence of electrodes, components that are
in direct contact with the plasma. Lifetime of
current EP technologies is driven by decrease of
performance due to degradation of acceleration grids
(GIT) and erosion of discharge channels (HET) in
time. This issue amplifies when using aggressive
propellants, such as atomic O, highly present in
ABEP Earth’s orbit altitude range. Moreover, an
IPT does not need a neutralizer, as the plasma
leaving the discharge channel is already neutral.
Therefore, the use of an IPT is an attractive
solution for ABEP application, first because much
longer mission are expected, especially due to
the possibility of full drag compensation while at
the same time requiring no propellant on-board,
and second because of the extended flexibility in
propellant utilization. The Institute of Space System
(IRS) of the University Stuttgart in Germany, has
decades of experience on inductively heated plasma
generators that have been, and are, used for reentry
condition simulations [15]. IPG6-S is a small scale
inductively heated plasma generator available at the
laboratories of IRS, and it has been selected for
ABEP application as thruster candidate, due to its
size and power level. IPG6-S is made of a 5.5 turns
coil wrapped around a discharge channel made of
quartz, that has an outer diameter of 40 mm and a
length of 180 mm, see Fig. 8. The power supply
delivers up to 20 kW power at 4 MHz. This has been
successfully operated with a mass flow of air and O2
from ∼0.2−120 mg/sfor ABEP analysis at input
power levels up to 3.5 kW [5].
Figure 8: IPG6-S operating with air
5 BALANCING MODEL
In the following, an analytical model [2] for the
evaluation of a generic ABEP intake configuration
is presented. The generic design is made of an
inlet section, and a chamber section. In the latter
all particles are assumed to have impacted the walls
with complete accommodation, therefore proceeding
only with thermal movement. These particle flows
are the backflow through the intake, and the flow
through the outlet. The outlet can be represented by
thruster’s acceleration grids, as in the JAXA design,
an injection device or a further stage of compression.
By balancing these flows, the conditions in the
separate sections can be estimated. The basic
assumptions for the analytical model are following
the nomenclature of Fig. 9.
Twall
Inflow
pin, nin ,
Tin, vin
Intake Control Volume, Chamber
pch, nch ,
Tch, vch
˙
Nin
Ain
Θint.1,˙
Nint.1
Θint.2,˙
Nint.2
Θout,˙
Nout
(˙
Naccel.)
Aout
Figure 9: Balancing Model Scheme
Ain and Aout are the respective cross sections
5BALANCING MODEL
for the inflow and the outflow representing those of
the chamber section. The parameters of the inflow
are known from the atmospheric model: number
density nin, pressure pin, temperature Tin and the
S/C velocity vin.Θis the transmittance into a specific
direction through a single structure. It is defined
as the fraction of particles that passes through the
exit section against the amount of particles which
have originally entered the structure at the start
section, as part of them are scattered. For the
model, three transmittances are to be set: one for
the incoming flow, one for the backflow coming from
the chamber (accounting for two values through the
intake part), and a third for the outflow. Based
on these transmittances, the respective particle
flows can be defined. ˙
Nint.1is the flow of
particles passing through the intake section to the
chamber section, ˙
Nint.2is the backflow that goes
back to the atmosphere after having reached the
chamber section, and ˙
Nout is the net outflow. Main
assumptions of the model in its here presented
version are:
•Free molecular flow - no collisions;
•Ideal gas and single species;
•Complete diffusive accommodation, α= 1;
•Fixed temperature, Tch =Twall;
•Only thermal velocity inside the chamber.
The particle flow ˙
Nin into the intake which can be
collected is calculated using free stream conditions
and inflow area; the actually collected flow is reduced
by the transmittance Θint.1.
˙
Nint.1=˙
NinΘint.1(2)
Based on the for-mentioned hypotheses, the
macroscopic velocity of the collected particles in ˙
Nin
will be zero and a superposed backflow will not
influence the inflow as it is a free molecular flow.
Starting from the temperature of the particles inside
the chamber, the thermal mass flux Γ, defined in
Eq. 3, according to [16, p.151], can be calculated.
Γ(n, T )xi=nrmpkBTch
2π=mpn¯vxi(3)
Therefore, it is possible to apply Γto determine
backflow and outflow in the chamber as following:
˙
Nint.2=Γ(nch, Tch )
mp
AinΘint.2(4)
˙
Nout =Γ(nch, Tch )
mp
AoutΘout +˙
Naccel. (5)
The continuity equation, see Eq. 6, can be
applied which states that the net amount of particles
flowing through a control volume, in this case the
chamber section, must be zero. ˙
Naccel. is the
accelerated particle flow actively extracted by the
thruster. This value depends on the operation point
of the thruster and its acceleration process. It is
expected that a minimum nis needed inside the
chamber for ignition. Therefore, the focus is at the
situation before ignition, ˙
Naccel. = 0.
˙
Nint.1=˙
Nint.2+˙
Nout (6)
The assumption of no macroscopic velocity in the
chamber is as in Eq. 7:
˙
NinΘint.1=Γ(nch , Tch)
mp
(AinΘint.2+Aout Θout)(7)
Therefore Γcan be extracted and, thus, the
density nch inside the chamber from Eq. 3 results
in:
nch = Γ(nch, Tch )s2π
mpkBTch
(8)
The pressure can be calculated by applying the
ideal gas condition as:
pch =nchkBTch (9)
Collection Efficiency ηcin Eq. 10, pressure ratio
in Eq. 11 and number density ratio in Eq. 12 are
important values for the evaluation of an intake.
ηc=˙
Nout
˙
Nin
=Θint.1
Ain
Aout Θint.2+ 1 (10)
pch
pin
=mp˙
NinΘint.1
AinΘint.2+Aout Θout s2π
mpkBTch
Tch
Tinnin
(11)
nch
nin
=mp˙
NinΘint.1
AinΘint.2+Aout Θout s2π
mpkBTch
1
nin
(12)
This model has been verified through
particle-based Monte Carlo simulations [2]. Intake
7 OPTIMIZATION FOR IPG6-S
efficiencies, pressures and mass flows have been
evaluated for different geometries, in particular for
the use of IPG6-S as thruster candidate for the ABEP
system.
6 INTAKE EVALUATION
The balancing model has been used to extend the
performance evaluation of the JAXA and BUSEK
intake concepts as well as with adapted geometry
to IPG6-S. As explained before, the JAXA concept
has a ring intake region and the BUSEK concept is a
long tube that terminates with a converging cone to
the thruster. Both of them incorporate a honeycomb
structure of ducts in the front to limit the backflow.
In this study is also considered that the front area of
the S/C is dominated by the intake, therefore its size
determines the major contribution to the drag. The
evaluation has been performed for both Mars and
Earth orbits into respective ABEP altitude ranges.
6.1 Inlet Structure of Ducts
For both concepts, a honeycomb inlet structure of
ducts is foreseen. From our precedent study [2]
it has been found that, even though the incoming
flow is very collimated due to very high velocity and
low density, not all the particles will flow through
a single duct without impacting its walls, therefore,
losing energy, thus, velocity. It has been shown that
the transmittance can be divided into a directly and
indirectly passing part and that both parts depend,
in the applicable velocity regime, only on the thermal
to macroscopic velocity ratio multiplied by L/R.
Together with the known Clausing factors [17] for the
backflow, a direct geometry with optimal Θcan be
determined. Sensibility analysis has been performed
for the given intake design.
6.2 Intake Outlet
The outlet area of the intake, has been set, in both
cases, to that of IPG6-S. The discharge channel of
IPG6-S has an outer diameter of dout =40 mm with
a wall thickness t=1.5 mm [5]. Therefore, Aout =
π
4d2
int =1.075 ×10−3m2, where dint =dout −2×t=
37 mm. The transmittance is at first approximation
Θ = 1. This has been set as the fixed geometry for
all the intakes.
7 OPTIMIZATION FOR IPG6-S
The first analysis consists in the BM applied to the
baseline intakes of BUSEK and JAXA, respectively
at 110 km Mars orbit and 140 km Earth orbit, with the
outlet area of IPG6-S while the inflow areas are kept
the same. MCD v5.2 and NRLMSISE-00 have been
used to determine the inflow conditions.
Ain Af˙mthr pch ηc
m2m2mg/s Pa %
JAXA 0.271 1.057 0.151 0.112 2.28
BUSEK 0.283 0.283 1.001 0.562 2.88
Table 1: JAXA, BUSEK applied to IPG6-S
Results in Tab. 1 show that ηcis very low for
both cases, ˙mthr is also small for the operation of
IPG6-S, especially in the JAXA design. This might
not be enough to produce enough thrust for full drag
compensation but also for ignition and sustainment
of the discharge itself. Therefore, an optimization of
the intake is necessary to increase ηcand ˙mthr for
the application to IPG6-S as candidate thruster.
7.1 Areas, Intake Efficiency and Mass
Flow
Commencing with analytical evaluations, the area
ratio between inlet and outlet (Ain/Aout), as can be
also seen in Eq. 10, directly influences the intake
efficiency ηc. Considering constant transmittances
Θ, higher values of ηcare obtained for very small
Ain
Aout . If Aout is kept constant, Ain has to be reduced
to increase ηc. This can be explained by the fact
that the thermal flux inside the chamber section is
uniform in all directions. A smaller Ain will therefore
favor ˙
Nout over the backflow. However, a smaller Ain
will decrease ˙
Nin and in total also ˙mthr. If a higher
˙mthr is required, ηcwill decrease, pointing the fact
that a maximum value for both ηcand ˙mthr, based
on the BM, cannot be achieved at the same time.
Pressure in the chamber pch has the same behavior
of ˙mthr . These considerations can be seen in Fig. 10
and Fig. 11.
7.2 JAXA-Design and IPG6-S
The design of JAXA has been modified to fit with
the Aout given by IPG6-S discharge channel. In this
analysis, the altitude has been set to 140 km in Earth
7.3 BUSEK-Design and IPG6-S 7 OPTIMIZATION FOR IPG6-S
orbit that corresponds to nin =0.7737 ×1017 m−3
according to NRLMSISE-00, and the ratio between
inlet and core area has been kept constant
to maintain the ring transmittance of the same
value. Ain has been varied for sensitivity analysis
purposes, in the range 1.0×10−4−0.271 m2, the
latter is the JAXA value, and the L/R of the ducts
and their corresponding Θiterated for each Ain to
obtain ηc,max. In the plots, results from Ain =
0.04 m2on, are estimation of the curve, by not
further optimizing L/R ducts ratio because the upper
limit of available data is reached. This is still
a good approximation since the most values are
not changing anymore. Moreover, the required ce
for the electric thruster, considering continuous full
drag compensation, exceeds for those high Ain an
unrealistic value of 100 km/sanyway. Results are
shown in Fig. 10.
0.04 0.1 0.2 0.2714
Ain ,m2
0.01
0.1
0.5
1
2
4
˙mthr,ηc,pch ,D,ce,
mg/s, -, P a,N,×105m/s
0.151
0.023
0.11
0.22
JAXA @ Earth, h= 140 km,
IPG6-S
˙mthr ηcpch D ce, T/D=1
Figure 10: JAXA Design Performance
Fig. 10 shows that for an increasing Ain, and
a constant transmittance of the ring section, ˙mthr
and pch have an asymptotic-like behavior, while ηc
decreases. The drag Dis also plotted and it is
calculated based on Afof the intake and a CD= 2.2
for Earth [5] and CD= 3 for Mars [11], by Eq. 13
D=1
2ρ(h)Afv(h)2CD(13)
In Tab. 2 the values for at intersection of ˙mthr and pch
curves with ηcare shown, as well as the points for
which ˙mthr is about 90% and 95% of the asymptotic
value. ∆asymp represents the difference of ˙mthr in
percentage from the asymptotic value.
Ain ˙mthr pch ηc∆asy mp.
m2mg/s Pa −%
0.041 0.120 0.088 0.120 20.5
0.055 0.126 0.094 0.094 16.6
0.083 0.136 0.101 0.067 10.1
0.132 0.144 0.107 0.045 5.0
0.271 0.151 0.110 0.023 −
Table 2: JAXA Optimization
7.3 BUSEK-Design and IPG6-S
Analogous to the JAXA design, the design of
BUSEK has been modified to fit with the Aout given
by IPG6-S discharge channel. In this analysis,
the altitude has been set to 110 km in Mars orbit
considering only CO2, corresponding to nin =
5.028 ×1017 m−3, according to MCD v5.2. Ain has
been varied, for sensitivity analysis purposes, in the
range 1.0×10−4−0.2827 m2, the latter is the BUSEK
value, and, again, the L/R of the ducts optimized
for each Ain to obtain ηc,max. Results from Ain =
0.17 m2on, are the same as explained for the JAXA
case. The L-to-Rof the section after the ducts part
has been kept constant to maintain its transmittance
at the same value. Results are shown in Fig. 11.
0.03 0.1 0.17 0.2 0.2827
Ain,m2
0.01
0.1
0.5
1
2
4
˙mthr,ηc,pch ,D,ce,
mg/s, -, P a,N,×105m/s
1.02
BUSEK @ Mars, h= 110 km,
IPG6-S
0.029
0.57
0.18
˙mthr ηcpch D ce,T/D = 1
Figure 11: BUSEK Design Performance
In Tab. 3 the values for at intersection of ˙mthr and
pch curves with ηcare shown, as well as the points for
which ˙mthr is about 90% and 95% of the asymptotic
value. ∆asymp represents the difference of ˙mthr in
percentage from the asymptotic value.
7.4 EFD Design (IPG6-S) 7 OPTIMIZATION FOR IPG6-S
Ain ˙mthr pch ηc∆asy mp.
m2mg/s Pa −%
0.009 0.393 0.221 0.393 61.5
0.014 0.500 0.281 0.281 50.1
0.122 0.919 0.516 0.061 9.9
0.174 0.969 0.544 0.045 5.0
0.283 1.020 0.570 0.029 −
Table 3: BUSEK Optimization
7.4 EFD Design (IPG6-S)
The BUSEK-like design, compared to the JAXA’s
one, shows generally a slightly higher collection
efficiency combined with a smaller total front area
that leads to less drag produced by the intake. The
difference is that Ain of the JAXA design is the
area of the inlet ring, to which the S/C core in the
middle has to be added. Therefore, further analysis
has been done with such a design, that from now
on, will be named the Enhanced Funnel Design
(EFD). The EFD design is thus applied to IPG6-S
for both Earth’s and Mars’ orbits and investigation
regarding its optimization as function of altitude, for
different Ain/Aout, see Tab. 4, and L/R for the
inlet structure and for the main duct is conducted.
For Earth’s orbit, a variation of the altitude from
120 −250 km, see [5], has been done. This implies
a variation of components and their concentration,
see Fig. 2, together with nin, Tin , vin, combined with
10 different Ain and their corresponding L/R duct
ratios for ηc,max. Results show ˙mthr, Fig. 12 and pch ,
Fig. 13, as function of hand Ain /Aout. For Mars’
orbit, a variation of the altitude from 88 −180 km has
been done. This implies a variation of components
and their concentration, see Fig. 3, together with
nin, Tin , vin, combined with 10 different Ain and their
corresponding L/R duct ratios for ηc,max.Results
show ˙mthr , Fig. 14 and pch , Fig. 15, as function of
hand Ain/Aout. The dashed line are, again, those
corresponding to not completely optimized ducts
Ain Ain/Aout
m2−
1.075 ×10−40.1
5.376 ×10−40.5
1.075 ×10−31
2.150 ×10−32
5.376 ×10−35
1.075 ×10−210
2.150 ×10−220
5.376 ×10−250
1.075 ×10−1100
2.796 ×10−1260
Table 4: Variation of Ain
120 160 200 250
h,km
1.0E-4
1.0E-3
1.0E-2
1.0E-1
1.0E0
˙mthr ,mg/s
Aout = 0.001075 m2
EFD IPG6-S @ Earth BM
Ain/A out
Figure 12: Mass flow, EFD @ Earth
120 160 200 250
h,km
1.0E-4
1.0E-3
1.0E-2
1.0E-1
1.0E0
pch,P a
Aout = 0.001075m2
EFD IPG6-S @ Earth BM
Ain /Aout
Figure 13: Chamber Pressure, EFD @ Earth
88 120 160 183
h,km
1.0E-4
1.0E-3
1.0E-2
1.0E-1
1
5
20
˙mthr ,mg/s
Aout = 0.001075 m2
EFD IPG6-S @ Mars BM
Ain/A out
Figure 14: Mass flow, EFD @ Mars
8CONCLUSION
88 120 160 183
h,km
1.0E-4
1.0E-3
1.0E-2
1.0E-1
1
10
pch,P a
Aout = 0.001075 m2
EFD IPG6-S @ Mars BM
Ain/A out
Figure 15: Chamber Pressure, EFD @ Mars
In Fig. 16 and 17 the required exhaust velocity ce
for the thruster, in case of full drag compensation, is
plotted as a function of hand Ain/Aout (now in the
range 1to 260) for both Earth and Mars orbit. ceis
calculated as a function of the collectible ˙mthr and
the respective D, Eq. 13, given by the variation on
Ain (Af=Ain), see Eq. 14:
ce=D(Ain, h)
˙mthr (Ain , h, ηc)(14)
Dashed line in the plots are estimation performed by
not further optimizing L/R ducts ratios, because the
upper limit of available data is reached. This is still
a good approximation since the most values are not
changing anymore.
120 160 200 250
h,km
5x10 3
1x10 4
1x10 5
5x10 5
ce,m/s
EFD IPG6-S @ Earth BM
Aout = 0.001075m2
Ain/A out
Figure 16: Required cefor T /D = 1, EFD @ Earth
88 120 160 183
h,km
5x10 3
1x10 4
1x10 5
5x10 5
ce, m/s
EFD IPG6-S @ Mars BM
Aout = 0.001075m2
Ain/A out
Figure 17: Required cefor T /D = 1, EFD @ Mars
The increase and sharp jump in the required ce
in Mars orbit, Fig. 17, is due to the temperature
profile of the atmosphere that, around h=120 km,
suffers from a sudden increase due to UV radiation
heating [9].
8 CONCLUSION
The results shown in Fig. 12, 13, 14, and 15 show
that increasing Ain/Aout increases the theoretically
collectible mass flow and the achievable pch.
Increasing Ain/Aout, in this case only Ain , as Aout is
kept constant as that of IPG6-S, makes sense only
until a certain ratio, as the increase of Ain directly
enlarges the drag that has to be counteracted. By
calculating the increase of ˙mthr over Ain/Aout , it
can be also seen that, for the same enlargement
of Ain/Aout, the ˙mthr gain reduces, Tab. 5. With
an increasing Ain/Aout,ηcdecreases continuously,
Dincreases linearly, and less than 10% increase
of ∆ ˙mthr is achieved for a ∆(Ain/Aout) = 100.
Therefore its maximum value should not exceed
Ain/Aout <100. For space reason, in Tab. 5 the
following renaming has been done: Ain,i /Aout,i =
ARiand Ain,i+1/Aout,i+1 =ARi+1
Ain/Aout ∆(Ain/Aout) ∆ ˙mthr
ARi→ARi+1
ARi+1−ARi
ARi
˙mthr,i+1 −˙mthr,i
˙mthr,i
-% %
1→2 100 74 −75
2→5 150 84 −88
5→10 200 41 −47
10 →20 200 28 −35
20 →50 150 23 −34
50 →100 100 10 −16
100 →260 160 7.5−13
Table 5: Ain
Aout vs ∆ ˙mthr Earth and Mars averaged
The difference of the curve shape for Mars
and Earth case is due to the different density
vs altitude profile in the respective ABEP altitude
ranges. Moreover, the mass of CO2is higher
than that of the components of Earth’s atmosphere.
The temperature of Mars’ atmosphere is lower than
that of Earth’s in the respective altitude ranges.
Both these factors increases the intake ηcaccording
to [2]. It has to be noted that ηcremains almost
constant in the ABEP altitude ranges for both Earth
REFERENCES
and Mars. Atmosphere’s density is higher in Mars
orbit at 88 km than in Earth’s at 120 km, and this
leads to a higher mass flow for the Mars case. In
the optimization for an IPT thruster, a compromise
between the required ˙mthr ,pch and the amount of
drag to be compensated has to be found. In Fig. 16
and 17 can be seen that the required cefor very high
Ain/Aout exceeds 100 km/s, when conventional EP
estimates a maximum of 50 km/s. The required ce
is lower in Mars orbit and, again, this would place
Mars orbit as a more suitable candidate for a full
drag compensation ABEP application. The tendency
of the curve suggest that an altitude slightly below
120 km would be favorable for ABEP application due
to a lower required ceand a better atmosphere
collection.
9 OUTLOOK
Further analysis is required to estimate the heat
load on the S/C in low orbits around Earth and
Mars to better evaluate an eventual limitation due
to overheating of the S/C. Additionally, further
analysis regarding gas-surface interactions should
be done including accommodation and molecule
recombination. The major next step is to operate
IPG6-S with the parameters found in this research
to evaluate its performance. In particular a minimum
pressure of ignition and minimum mass flow have
to be experimentally and theoretically investigated
to determine the minimum conditions for ignition
and operation. The application of a nozzle to
IPG6-S is in progress and this will aid the estimation
of the exhaust velocity to be compared to the
required ceevaluated in this research. Due
to the limited pch, if this is not enough for an
IPT, the use of a Knudsen compressor might be
included. A Knudsen compressor is a passive
device that uses a difference of temperature to
increase the pressure of a very rarefied flow and
it is based on the phenomena of thermal creep
flow. Experimental studies operated also in the
low pressure regime [18], 1< p < 100 Pa, and
an increase of pressure of a factor of 10 has been
achieved with a 33 stages Knudsen compressor and
the total required power, to sustain a ∆T=100 K, is
of P=1.1 W [19]. A concept applied to IPT is shown
in Fig. 18.
Inflow
from
Intake
Knudsen Compressor
Chamber Injector
PLASMA
𝑝"# > 𝑝%&'()*
𝑝"#
𝑝%&'()*
IPT
Figure 18: IPT with Knudsen Compressor
This concept includes a chamber in which the
propellant is transferred and compressed by the
Knudsen compressor. A passive injector releases
the particles only after a certain pressure threshold
is reached. The propellant reaches the IPT and it
is ionized and expelled at high velocity for thrust
generation, therefore, the operation of IPT will be
intermittent and transient higher ˙mthr might be
achieved.
10 ACKNOWLEDGEMENTS
F. Romano gratefully thanks for the financial support
the Landesgraduiertenf ¨
orderung of the University of
Stuttgart .
The authors thank Mr. Pietro Carlo Boldini for his
very valuable contribution in this research.
References
[1] Sch¨
onherr, T., Komurasaki, K., Romano,
F., Massuti-Ballester, B., and Herdrich,
G., “Analysis of Atmosphere-Breathing
Electric Propulsion,” Plasma Science, IEEE
Transactions on, Vol. 43, No. 1, Jan 2015,
pp. 287–294.
[2] Romano, F., Binder, T., Herdrich, G., S,
F., and Sch ¨
onherr, T., “Air-Intake Design
Investigation for an Air-Breathing Electric
Propulsion System,” 34th International Electric
Propulsion Conference, Kobe, Japan., 2015.
[3] ESA, “GOCE System Critical Design Review,”
2005.
[4] Di Cara, D., Santovincenzo, A., Dominguez,
B. C., Arcioni, M., Caldwell, A., and Roma,
I., “RAM electric propulsion for low earth orbit
operation: an ESA study,” 2007.
REFERENCES REFERENCES
[5] Romano, F., System Analysis and Test-Bed for
an Air-Breathing Electric Propulsion System,
Master’s thesis, Universit ¨
at Stuttgart, University
of Padova, 2014.
[6] Young, M., Muntz, E., and Wang, J.,
“Maintaining Continuous Low Orbit Flight
by Using In-Situ Atmospheric Gases for
Propellant,” Rarefied Gas Dynamics: 22nd
international symposium, Vol. ADA409086,
2001.
[7] Picone, J., Hedin, A., Drob, D., and Aikin,
A., “NRLMSISE-00 Empirical Model of the
Atmosphere: Statistical Comparisons and
Scientific Issues,” Journal of Geophysical
Research, Vol. 107, No. A12, 2002.
[8] Alexander, M., “Mars transportation
environment definition document,” 2001.
[9] Madeleine, J.-B., Forget, F., Millour, E.,
Montabone, L., and Wolff, M. J., “Revisiting
the radiative impact of dust on Mars using
the LMD Global Climate Model,” Journal of
Geophysical Research (Planets), Vol. 116, Nov.
2011, pp. E11010.
[10] Andersson, L., Weber, T., Malaspina, D., Crary,
F., Ergun, R., Delory, G., Fowler, C., Morooka,
M., McEnulty, T., Eriksson, A., et al., “Dust
observations at orbital altitudes surrounding
Mars,” Science, Vol. 350, No. 6261, 2015,
pp. aad0398.
[11] Hohman, K., “Atmospheric Breathing Electric
Thruster for Planetary Exploration,” Final
Report, 2012.
[12] Hohman, K., “Atmospheric Breathing Electric
Thruster for Planetary Exploration,” Presented
as the NIAC Spring Symposium, Presentation,
2012.
[13] Fujita, K., “Air-intake Performance Estimation
of Air-breathing Ion Engines,” Transactions of
the Japan Society of Mechanical Engineers. B,
Vol. 70, No. 700, dec 2004, pp. 3038–3044.
[14] Hisamoto, Y., Nishiyama, K., and Kuninaka,
H., “Design of Air Intake for Air Breathing
Ion Engine,” 63rd International Astronautical
Congress, Naples, Italy., 2012.
[15] Herdrich, G., “Aufbau, Qualifikation
und Charakterisierung einer induktiv
beheizten Plasmawindkanalanlage zur
Simulation atmosph¨
arischer Eintrittsman ¨
over,”
Dissertation, 2004.
[16] Bird, G. A., Molecular Gas Dynamics and the
Direct Simulation of Gas Flows, Clarendon,
Oxford, 1994.
[17] Clausing, P., “ ¨
Uber die Str¨
omung sehr
verd¨
unnter Gase durch R¨
ohren von beliebiger
L¨
ange,” Annalen der Physik, Vol. 404, No. 8,
1932, pp. 961–989.
[18] Han, Y., Young, M., Muntz, E., and Shiflett,
G., “Knudsen Compressor Performance at Low
Pressures,” Rarefied Gas Dynamics: 24th
International Symposium on Rarefied Gas
Dynamics, Vol. 762, 2005, pp. 162–167.
[19] Young, M., Shiflett, G., Muntz, E., Vargo, S.,
and Green, A., “The Knudsen Compressor as
an Energy Efficient Micro-Scale Vacuum Pump,”
Proceedings of IMECE2001, New York , 2001.