A Survey of Propulsion Options for Cargo and Piloted
Missions to Mars
â
K. Sankaran, L. Cassady, A. D. Kodys and E. Y. Choueiri
Electric Propulsion & Plasma Dynamics Laboratory
Mechanical & Aerospace Engineering Department
Princeton University, Princeton, NJ 08544
â
January 22, 2003
In this paper, high-power electric propulsion options are surveyed in the context of cargo and piloted
missions to Mars. A low-thrust trajectory optimization program (RAPTOR) is utilized to analyze this
mission. Candidate thrusters are chosen based upon demonstrated performance in the laboratory. Hall,
self-field magnetoplasmadynamic (MPDT), self-field lithium Lorentz force accelerator (LiLFA), arcjet,
and applied-field LiLFA systems are considered for this mission. In this first phase of the study, all
thrusters are assumed to operate at a single power level (regardless of the efficiency-power curve), and the
thruster specific mass and powerplant specific mass are taken to be the same for all systems. Under these
assumptions, for a 7.5 MW, 60 mT payload, piloted mission, the self-field LiLFA results in the shortest
trip time (340 days) with a reasonable propellant mass fraction of 57% (129 mT). For a 150 kW, 9 mT
payload, cargo mission, both the applied-field LiLFA and the Hall thruster seem reasonable choices with
propellant mass fractions of 42 to 45 % (7 to 8 mT) . The Hall thrusters provide better trip times (530-570
days) compared to the applied-field LiLFA (710 days) for the relatively less demanding mission.
Nomenclature
α
p
Specific mass of power supply (kg/kW)
α
t
Specific mass of thruster (kg/kW)
â
V
Spacecraft velocity change (km/s)
η
th
Thrust efficiency (%)
B
Magnetic induction (T)
E
Electric field strength (V/m)
g
o
Acceleration due to gravity on earth (m/s
2
)
I
sp
Specific impulse (s)
=
u
e
/g
o
J
Total current (A)
j
Current per unit area (A/
m
2
)
P
Input power to the thruster (kW or MW)
P
th
Thrust power (kW or MW)
T
Thrust (N)
Ë
m
Propellant mass flow rate (kg/s)
u
e
Propellant exhaust velocity (km/s)
1
Introduction
F
or the first time in over a decade NASA has been given
the green light to pursue nuclear options for spacecraft
propulsion. The Nuclear Space Initiative (NSI), approved
under NASAâs FY2003 budget, is a multi-year program
expected to total $2 billion with one goal being the devel-
opment of space nuclear systems capable of 10-100 kW of
power in space over the next ten years[1]. This initiative
promises to open up the outer solar system to exploration
by reducing spacecraft weight (propellant mass savings)
and transit times (5 years versus 10 years to Pluto, with
respect to chemical thrusters, at high power levels ), and
by providing a power supply to do science once the space-
craft arrives at the destination. As a result, the surface of
Mars may now be accessible for long-term robotic and
human exploration. This paper describes the first-phase
of our study comparing near-term propulsion options for
a two-stage (cargo & piloted) mission to Mars.
Because of their high exhaust velocities, electric
â
The authors thank the NASA Johnson Space Center for providing the RAPTOR code.
â
Presented at the International Conference on New Trends in Astrodynamics, Jan.20-22, 2003. Copyright by authors.
1
Sankaran, Cassady, Kodys, and Choueiri: PROPULSION OPTIONS FOR MISSIONS TO MARS
2
propulsion (EP) systems can provide significant pro-
pellant savings over chemical thrusters for high
â
V
missions[2] like this one, and have been popular in such
mission studies.
With the success of the ion thruster
as the primary propulsion system on the Deep Space 1
mission[3], and other recent EP enabled missions, the
field of electric propulsion has now come of age as a re-
liable and efficient way of accomplishing relatively low
energy missions. Currently, various types of EP devices,
resistojets, arcjets, ion thrusters, Hall thrusters, and to a
lesser extent, pulsed plasma thrusters, are routinely used
for station keeping and maneuvering satellites. However,
research on high-power electric propulsion has stagnated
over the last two decades. As a consequence of the revival
of interest in nuclear space systems, high-power propul-
sion options, first investigated in the 1960s to early 1980s
(cf. ref.[4]) and then abandoned or continued at lower
power levels due to lack of power in space, are receiving
renewed attention.
1.1
Review of Previous Studies
Since the dawn of the space age, many mission studies
have been performed on expeditions to Mars. Stuhlinger
et al.[5] were among the first to propose the use of elec-
tric propulsion for missions of this kind. In the last two
decades, several noteworthy studies have examined the
advantages and disadvantages of various propulsion sys-
tems for a mission to Mars. We will briefly survey some
piloted and cargo mission studies.
Coomes et al.[6] propose the use of a magnetoplas-
madynamic thruster (MPDT - to be described in
§
3.2.1)
operating at a power level of 6 MW for a piloted mission,
and calculate a trip time of 600 days for Earth to Mars at
this power level. King et al.[7] also examine the use of a
MPDT for a similar mission and propose systems with in-
put power up to 200 MW that can accomplish Earth-Mars
round trip in less than a year. It is also suggested that
MPDTs can offer trip time savings over chemical thrusters
at power levels of 10 MW or higher. Gilland et al.[8] com-
pare the use of MPDTs versus an array of ion thrusters for
a similar mission, using a curve-fit for
η
th
vs.
I
sp
for the
performance of these thrusters. Clark et al.[9] examine
a 8 MW piloted (35 mT) fast trajectory mission for trip
time, safety and reliability, abort options, and other costs.
Pelaccio et al.[10] provide a technology readiness assess-
ment of various thrusters for such a mission.
Clark et al.[9] also consider 4 MW minimum energy
trajectory for a Mars cargo mission. They estimate that
an array of ion thrusters offer significant mass savings
over nuclear thermal systems, while maintaining compa-
rable trip times. Frisbee et al.[11, 12] assess the technol-
ogy readiness and development requirements for dynamic
power conversion, power processing, and thrusters for
Mars cargo mission. Polk et al.[13] examine the lithium
Lorentz force accelerator technology (LiLFA - to be de-
scribed in
§
3.2.2) for reusable orbit transfer vehicle with
a parametric study of required power level, specific mass
of power plant and performance to focus technology de-
velopment. Noca et al.[14] consider robotic missions to
outer planets with power levels ranging from 100 kW to
1 MW, using ion engines. Woodcock et al.[15] consider
three outer planetary missions with small payloads, and
consider the use of various propulsion systems.
As described above, a lot of work has been done
on investigating propulsion options for missions to Mars.
However, the abovementioned studies either perform the
analysis with extrapolated data, and/or look at the prob-
lem from the perspective of research guidelines for a spe-
cific thruster. Therefore, there is a need for a comparison
of multiple propulsion options for this mission using mea-
sured performance data only, and that is the goal of this
paper.
1.2
Outline
The two-stage human Mars mission, chosen for this study,
is described in detail in
§
2. This mission will benefit from
nuclear power by yielding mass savings and reduced trip
time, due to the ability to operate at higher power than
would be otherwise available. In
§
3, we will examine
some of the existing EP devices capable of taking advan-
tage of near-future nuclear power systems, and we will
discuss the current status of the technology and current
trends in research. As will be described in
§
3.1, we will
limit this study to thrusters that have been successfully op-
erated (thrust measured) in the laboratory to keep with the
near-term (10-20 years) spirit of the study and to perhaps
provide some insight into technology drivers. The results
of the mission analysis will be presented in
§
4. Following
that, in
§
5, we will briefly discuss the propulsion options
that were not considered in this analysis.
2
Mission Description
In this study we consider a two-stage mission to Mars.
The power available for each stage of the mission was
chosen in accordance with the near-term technology as-
sessment of the thrusters, though many of the studies men-
tioned in
§
1.1 consider missions with larger power sup-
plies.
In the first stage, 90 metric tons of cargo would be
transported from Earth orbit to Mars orbit. The propulsion
Sankaran, Cassady, Kodys, and Choueiri: PROPULSION OPTIONS FOR MISSIONS TO MARS
3
systems for this mission would have a total power sup-
ply of
O
(100 kW) available, and the trip time is expected
to be approximately two years. Because of the relatively
low power available in this mission, the total payload of
90 mT would be transported in ten vehicles (each with a
payload of 9 mT) to accomplish the trip within two years.
The propulsion system for this stage of the mission should
accomplish this with minimum propellant mass.
The second stage would carry the crew and supplies,
totaling 60 metric tons of payload, and would be launched
approximately two years after the launch of the cargo
stage. The propulsion systems for this mission would
have a total power supply of
O
(1 MW) available, and the
trip time must be less than one year, due to human health
factors.
2.1
Trajectory Calculations
The mission simulation is accomplished with RAPTOR
(RAPid Trajectory Optimization Resource), an optimiza-
tion program developed at NASA-Johnson Space Center
for low-thrust, interplanetary missions
The position and velocity of the departure and arrival
planets are the boundary conditions. Given these, the code
minimizes the total acceleration of the interplanetary tra-
jectory. RAPTOR contains a genetic algorithm to con-
verge on the proper Lagrange multipliers, trip length and
departure date for the heliocentric code.
More details of how the code was used for this mission
analysis will be given in
§
4.2.
3
Propulsion Options
In this section we will briefly describe the propulsion sys-
tems that may be suitable for this mission. Because of
their high
I
sp
, EP systems are naturally attractive candi-
dates for this type of mission. Within the family of elec-
tric propulsion devices, several types of thrusters, con-
ceptually, have the ability to process 100s of kilowatts to
megawatts of power at reasonably high efficiencies. We
list them in table (3), and in
§
3.1, we narrow down the
field to a few thrusters which will be considered in the
mission analysis. Further information on these devices
can be obtained from recent surveys, such as refs.[16, 17].
First, we will describe the criteria we used to select the
thrusters.
3.1
Selection Criteria
While a variety of propulsion systems have been proposed
for interplanetary missions, we restrict our analysis only
to those that have:
1. been successfully characterized in a laboratory as a
thruster (i.e., thrust and efficiency have been mea-
sured directly),
2. demonstrated a potential for attaining a significant
lifetime (
O
(100 hours)),
3. the ability process at least:
âą
25 kW per thruster for the cargo mission,
âą
500 kW per thruster for the piloted mission,
so that the number of thrusters per spacecraft is rea-
sonable.
Since, at present, there do not exist conclusive lifetime
assessment tests of any of the devices, we consider those
that have operated with tolerable erosion for 100 hours.
At high power levels, measured thrust and efficiency
data is available for only three main classes of thrusters,
Hall thrusters, thermal arcjets, and magnetoplasmady-
namic thrusters (MPDT). Hall thrusters and thermal ar-
cjets have been operated at power levels up to 150 kW,
and are therefore well suited for the cargo phase of the
mission. Some Hall thrusters examined in this study have
operated only at lower power levels (25-75 kW) and there-
fore a cluster of 2-6 thrusters will be required to accom-
plish the mission. The only EP device to date to have
demonstrated the ability to operate at megawatt power lev-
els with a single (or reasonably small number of) thruster
is the MPDT. For the MPDT two distinct variations exists
differing in propellant and electrode design, both of which
have been operated in the laboratory and will be discussed
here.
Other promising thruster concepts, such as the ion
thruster, Pulsed Inductive Thruster (PIT), and the VAri-
able Specific Impulse Magnetoplasma Rocket (VASIMR),
that did not meet our selection criteria are discussed in
§
5.
3.2
Piloted Mission
Few thrusters have demonstrated performance at power
levels of
O
(MW), and have survived many hours of lab-
oratory testing. Consequently, the field narrowed down
to gas-fed magnetoplasmadynamic thrusters (MPDT) and
lithium Lorentz force accelerators (LiLFA). They will be
briefly described in the subsequent sections.
Sankaran, Cassady, Kodys, and Choueiri: PROPULSION OPTIONS FOR MISSIONS TO MARS
4
3.2.1
Magnetoplasmadynamic Thrusters
In the magnetoplasmadynamic thruster (MPDT), a volt-
age applied between concentric electrodes breaks down a
propellant gas, creating a quasi-neutral plasma within the
thruster chamber. A high current (
O
(10
2
â
10
4
)
A) car-
ried by the plasma to the electrodes induces an azimuthal
magnetic field, causing a Lorentz force (j
Ă
B per unit
volume) to accelerate the plasma out of the thruster at ve-
locities of
O
(10 km/s)[4]. As shown in fig.(1), this body
force accelerates the fluid in the direction perpendicular
to both the electric and the magnetic fields.
The MPDT has a unique place among electric
thrusters in its ability to process megawatts of electri-
cal power in a small, simple, compact device and pro-
duce thrust densities (thrust per unit exhaust area) of
O
(10
5
)
N/m
2
[17]. However, this major advantage of the
MPDT has also been the disadvantage to its development.
Since high efficiencies (
>
30
%) are only reached at high
power levels (
>
200 kW), MPDTs require power levels
that are an order of magnitude higher than what is cur-
rently available on spacecraft in order to be competitive
with other propulsion options. Therefore, research on
MPDTs was largely sidelined, in favor of thrusters that
have higher efficiencies at lower power levels. Renewed
interest in high-power MPDTs led to a flight-test of a
quasi-steady MPDT in 1996 aboard the Japanese Space
Flyer Unit, which was operated successfully at 1 kW [18].
Still, steady-state testing at the megawatt level is dif-
ficult, and to date all data in the 1 - 6 MW range has been
taken in quasi-steady mode. In this mode, the thruster is
operated for current pulse length of
O
(1 ms), and data
from this mode is expected to be a good indication of
its steady-state performance[19]. Databases of measured
quasi-steady thruster performance have been compiled in
Japan [20] and at Princeton University [21]. A MW-class
pulsed facility at the NASA-Glenn Research Center began
operation in 2001, with plans to develop it to a steady-
state facility[22]. So far, steady-state data is limited to less
than 1 MW, and has been obtained mostly at the NASA
Glenn (formerly Lewis) Research Center[23], and at the
University of Stuttgart[24]. The NASA-Lewis test facility
had the capability to operate at steady-state power level of
up to 600 kW, but research was discontinued by the early
1990s. The focus of the research at University of Stuttgart
is on investigating arc and plasma instabilities at power
levels from 0.1 to 1 MW. At 0.5 MW, efficiencies of 28%
at 1099 s were achieved during steady-state operation.
To date, the best gas-fed MPDT data has been
achieved with hydrogen in quasi-steady mode, 43% ef-
ficiency at 5000s has been reported [25]. In general, how-
ever, results have been in the 10 - 35% in range at 1000 -
4500 s specific impulse using Ar, NH
3
, and N
2
as propel-
lants [26].
A major concern about the MPDT technology is the
erosion of its cathode (cf. ref.([27]), which has often lim-
ited the lifetime of laboratory studies. However, there are
some indications from recent research[24, 28] that this
problem may be manageable.
To summarize, the three major technological issues
exist in the development of the MPDT are: accurate per-
formance measurements at steady state, characterization
and optimization of the thrusterâs stable operating range,
and demonstration of lifetimes of the order of mission re-
quirements (8,000-10,000 hours).
3.2.2
Lithium Lorentz Force Accelerators
The lithium Lorentz Force Accelerator (LiLFA) can be
considered the next-generation MPDT [17]. Its operat-
ing principle is essentially identical to that of the MPDT.
The name is a result of largely historical reasons. How-
ever, two major differences between the LiLFA and the
MPDT are to be noted. First is the choice of propellant.
Whereas the MPDT traditionally uses inert gas propel-
lants, such as argon, helium, and hydrogen, the LiLFA,
as its name indicates, uses lithium vapor. Furthermore,
the central electrode of the LiLFA differs from the sin-
gle rod design common to most gas-fed MPDTs. Instead,
the LiLFA employs multiple rods, tightly packed within
a hollow tube. Propellant flow (see fig.(2)) is through the
channels in the cathode that are created in between these
smaller rods, rather than from the electrode base, as in the
MPDT (fig.(1)).
These two major differences address some of the fun-
damental limitations of the MPDT uncovered during ex-
tensive testing in the 1960s and 70s. First, the choice
of a low-ionization energy propellant (lithium) reduces
the non-recoverable energy lost in ionizing the propellant,
thus improving efficiency, especially at low power levels
(
<
200 kW) where the ionization sink can approach 50%
of the total input power. The use of lithium also precludes
the need for high-voltage ignition capacitors and hard-
ware required to achieve breakdown in inert gas systems.
Moreover, lithium can be stored in solid form onboard
the spacecraft, leading to potential mass savings. How-
ever, no space-qualified system exists for feeding lithium
propellant. The multi-channel design for the central elec-
trode, combined with the lithium propellant, has been
shown to improve efficiency and increase thruster life-
time by reducing electrode erosion[29]. The role of the
electrode design in increasing efficiency and lifetime is,
however, poorly understood.
The possibility of the LiLFA as a high-power propul-
sion option was shown nearly a decade ago with the
demonstration of 500 hours of erosion-free operation, at
Sankaran, Cassady, Kodys, and Choueiri: PROPULSION OPTIONS FOR MISSIONS TO MARS
5
Table 1: Summary of Propulsion Options for Two-Phase Mars Mission.
Thruster
Laboratory Models
Measured Thrust
Lifetime
>
100
hrs
at P
>
25 kW
Arcjet
â
â
â
Hall
â
â
â
Ion
â
â
LFA
â
â
â
MPDT
â
â
â
PIT
â
â
VASIMR
â
a power level of 500 kW. At this power level, an exhaust
velocity of 40 km/s, thrust of 12 N, and thrust efficiency
of 60 % were reported [29]. Due to its high efficiency
and high thrust-to-power ratio, the LiLFA has emerged as
a promising candidate for the piloted mission. However,
since very little data is available on this thruster aside from
what has been cited in ref.[29], there is a need for further
verification of this performance. To this end, NASA-JPL
is in the process of developing a high-power steady state
test facility for LiLFA at MW power level.
Figure 1: Schematic of the operation of the MPDT.
Figure 2: Schematic of the operation of the self-field
LiLFA.
3.3
Cargo Mission
As mentioned earlier, there has been more research done
on relatively low power propulsion than on MW-class
propulsion required for the piloted mission. Hence, for
the cargo mission, more thrusters are available for consid-
eration. The thrusters that have met our selection criteria
will be briefly described in the subsequent sections.
Figure 3: Schematic of the operation of the arcjet.
3.3.1
Arcjets
In this class of thrusters, an electric arc is used to add
enthalpy to the propellant. Akin to a chemical thruster,
part of the enthalpy in the flow is converted to directed
kinetic energy using a nozzle.
Sankaran, Cassady, Kodys, and Choueiri: PROPULSION OPTIONS FOR MISSIONS TO MARS
6
As shown in fig.(3), a tightly constricted electric arc,
carrying currents up to
O
(100) A, heats the core of the
propellant stream to temperatures up to 10000 K, while
the walls of the thruster are maintained at much lower
temperatures (
<
3000 K) to prevent melting. Because
of the higher temperatures in the core, and consequently,
higher specific enthalpy, the exhaust velocity of an arcjet
can reach, or even exceed, 10 km/s, as opposed to only
4 km/s for a chemical thruster. Its simple design and its
high thrust density are some of the attractive features of
the arcjet.
We will see in
§
4.3, however, that arcjets tend not to
be competitive as our other option at the power levels con-
sidered.
3.3.2
Hall Thrusters
Hall thrusters derive their thrust by accelerating heavy
ions, using an electric field, to high exhaust velocities.
Under high magnetic fields, and low densities, the
current in the direction perpendicular (specifically, in the
E
Ă
B
direction) to the electric field (commonly called
the Hall current) can exceed the current along the elec-
tric field. As implied by their name, Hall thrusters uti-
lize this Hall current to lock electrons into a nearly col-
lisionless cross-stream drift, leaving the positive ions
free to be accelerated by the applied electric field (cf.
fig.(4)). In a sense, these devices are hybrid electrostatic-
electromagnetic accelerators with space-charge neutral-
ization automatically provided by the background of drift-
ing electrons.
Because of this, the Hall thrusters are
not affected by space-charge limitations.
Therefore,
Hall thrusters produce higher thrust densities than space-
charge limited devices, such as ion thrusters.
Because the magnetic fields in these devices are ex-
ternally supplied, and because the mass flow densities
are intrinsically low, these thrusters optimize their per-
formance at considerably lower powers than those of the
self-field MPDTs. In fact, the requirement of low mass
density (to maintain significant Hall effect) precludes the
Hall thruster from producing thrust densities comparable
to the MPDT and the LiLFA at a given power level.
The first Hall thrusters were developed in the US in
the 1960s, but the research was discontinued in favor of
ion engines[17]. However, research in the former Soviet
Union picked up in the late 1960s, and by 1990s they had
demonstrated efficiencies greater than 50%.
Hall accelerators of the closed-drift type are at present
the most commonly used plasma thrusters. Since 1972,
more than 100 Hall thrusters have been flown on Russian
spacecraft. They are developed and used by the commer-
cial sector for orbit insertion, attitude control, and drag
compensation of satellites.
The Hall thrusters chosen for this study were the
NASA-457M high power model[30] using xenon propel-
lant, and the thruster with anode layer (Bi-TAL) from
TsNIIMASH[31] using bismuth propellant.
The NASA-457M has been tested for power levels up
to 75 kW. At this power level, its measured efficiency was
58%, specific impulse 2900 s, and thrust 2.95 N. A vari-
ant of this thruster[32] had an efficiency of 62%, specific
impulse of 3250 s, and thrust 0.95 N at 25 kW of input
power.
The Bi-TAL had efficiency greater than 70%, and a
specific impulse of 8000 s at 140 kW input power. A
variant of this thruster (designated TAL-200) had an ef-
ficiency of 67%, specific impulse of 3000 s, thrust of 1.13
N, at an input power level of 25 kW[31]. As a result of
its high
I
sp
and high efficiency, this thruster has emerged
as a promising candidate for the cargo mission. However,
very little data is available on this device, and therefore,
there is a need for further verification of its performance
and lifetime.
Figure 4: Schematic of the operation of the Hall thruster.
3.3.3
Applied Field LiLFA
Stated simply, the applied field LiLFA (AF-LFA) is a de-
sign to increase MPDT/LiLFA efficiency at power levels
less than 200 kW, by adding an additional source of mag-
netic field. By using an external solenoid to enhance the
magnetic field, efficient electromagnetic acceleration can
occur at power levels where the current is too low to in-
duce a substantial magnetic field. The AF-LFA offers the
advantage higher efficiencies (
â„
40 %) at lower power
(
<
200 kW) compared to MPDT and LiLFA, while main-
taining exhaust velocities (10-35 km/s) that are compara-
ble. Apart from high-energy planetary missions, potential
applications of the AF-LiLFA include missions requiring
high thrust-to-power ratios, such as orbit transfer, N-S sta-
tionkeeping, and drag compensation.
Sankaran, Cassady, Kodys, and Choueiri: PROPULSION OPTIONS FOR MISSIONS TO MARS
7
Table 2: Summary of performance of thrusters for cargo phase of mission
Thruster
Power (kW)
Thrust (N)
I
sp
Efficiency (%)
Reference
H
2
Arcjet
75
2.5
1715
29
[34]
Bi Hall
25
1.13
3000
67
[31]
150
2.5
8000
70
[31]
Xe Hall-1
25
0.95
3250
60
[32]
Xe Hall-2
75
2.9
2900
56
[30]
AF-LiLFA
150
4.0
3290
40
[33]
Argon MPDT
150
6.7
849
19
[24]
Table 3: Summary of performance of thrusters for piloted phase of mission (* denotes quasi-steady data.)
Thruster
Power (MW)
Thrust (N)
I
sp
Efficiency (%)
Reference
LiLFA
0.5
12
4077
60
[29]
H-MPDT-1*
1.5
26.3
4900
43
[25]
H-MPDT-2*
3.75
88.5
3500
43
[25]
H-MPDT-3*
7.5
60.0
6000
25
[21]
Argon MPDT
0.5
25.5
1099
28
[24]
At the power levels considered for cargo missions,
the applied field lithium Lorentz force accelerator, jointly
developed by the Moscow Aviation Institute (MAI) and
NASA-JPL was a suitable candidate for consideration.
The conclusion of a 5-year AF-LiLFA research program
at the MAI [33] was the design of a 48% efficient thruster
operating at 185 kW with 4200s and 4.5 N. For 130 kW
of input power, this thruster has demonstrated 40% effi-
ciency, 3290 s specific impulse, and 4.0 N of thrust in
their study[33]. However, like in the case of its self-field
variant, very little data is available on this thruster, and
hence there is a need for further verification of its perfor-
mance. Therefore, the collaborative effort by NASA-JPL
and Princeton University is aimed at testing the applied-
field LFA at power levels
O
(10 - 100 kW).
4
Mission Analysis
4.1
Assumptions
In order to simplify our analysis, we have made the fol-
lowing assumptions for the present study, some of which
will be relaxed in future studies.
1. The specific mass of the power supply,
α
p
, is as-
sumed to be 4.0 kg/kW for all the cases. This is
within the range of previous studies such as ref.[12]
and ref.[8].
2. The specific mass of the thrusters for the piloted
mission were all assumed to be 0.35 kg/kW [13].
Since many of the thruster considered in this study
are still laboratory models, it is not easy to arrive at
an accurate estimate for
α
t
. The influence of this
assumption on the result is yet to be determined,
since the mass of the thruster is expected to be only
a small fraction of the total mass.
3. For the cargo mission, the mass of the thruster sys-
tem was assumed to be negligible compared to the
total mass, and was hence neglected.
4. The arrival date on Mars orbit was fixed to be the
same for all cases of each of the two (cargo and
piloted) missions.
5. The cargo mission was analyzed at a power level of
150 kW, irrespective of the optimum power level of
the thruster. The piloted mission was analyzed at a
power level of 7.5 MW, irrespective of the optimum
power level of the thruster.
6. For the cargo mission, the total payload of 90 mT
was evenly distributed among ten spacecraft. This
Sankaran, Cassady, Kodys, and Choueiri: PROPULSION OPTIONS FOR MISSIONS TO MARS
8
was done in order to ensure a reasonable trip time,
given the limited (150 kW) power supply.
4.2
Calculations
For the thrusters that met our selection criteria (
§
3.1), we
selected the highest measured performance data that was
available and used it as input into the RAPTOR code. A
summary of the thruster data is presented in tables (2 &
3). As noted in table (3), the data for the three types of
hydrogen MPDT were obtained in a quasi-steady mode of
operation, and it is expected to be a good indication of the
steady-state performance as well[4].
For this study, we did not use the genetic algorithm,
described in
§
2.1 to optimize the departure date due to the
large amount time required to calculate the shortest trip.
The dates of 12/1/2016 for the cargo, and 12/1/2018 for
the piloted missions, were chosen as the arrival date at
Mars (before the spiral) because those dates are expected
to be near the minimum for those missions. The genetic
algorithm was used to find the Lagrange multipliers and
trip length that best satisfied the mission.
Given an initial mass in Earth orbit, the RAPTOR
code can determine the final mass, or given a payload to
Mars, the code can find the initial mass required. Since
we have chosen a payload, RAPTOR will be run in latter
mode. First, RAPTOR executes the spiral in to Mars or-
bit to determine that portionâs duration and the propellant
the maneuver requires. The heliocentric code then uses
the mass of the payload and propellant to begin its opti-
mizations. In this mode the date of arrival in Marsâ sphere
of influence is the only controllable date, all else is refer-
enced to that date. The heliocentric code optimizes âback-
wardâ in time to find the minimum acceleration (i.e., min-
imum propellant) trajectory. Finally, the spiral to escape
from Earth is executed and the mass at escape is matched
with that at the beginning of the heliocentric transfer. The
genetic algorithm is used only with the heliocentric por-
tion of the code.
4.3
Results
The results of the RAPTOR code under the assumptions
of this study are given in figures (5 & 6). The results for
the trip time can be considered accurate to within
±
10
days, and for the propellant mass within
±
1 mT for both
stages of the mission. The accuracy of the trip time is
based on the sum of the round offs in the convergence cal-
culation of various phases of the trip, and the accuracy
of the mass estimate is based on the sum of the uncer-
tainty associated with estimating the mass of the compo-
nents such as the tank mass, and other structural mass.
The RAPTOR code does not explicitly optimize for
trip time, rather it finds the trajectory that minimizes ac-
celeration. This amounts to minimizing the required ini-
tial mass in nuclear-safe earth orbit. Since the thruster
mass, payload mass, and power supply masses were as-
sumed to be constant for each stage of the mission, the
initial mass is a function of propellant mass alone. The
minimum acceleration trajectory will result in the mini-
mum propellant used and hence the minimum initial mass.
4.4
Cargo Mission
The cargo mission was restricted to a power level of 150
kW in order to utilize measured thruster data in this study
rather than projections of performance. Our power restric-
tion was found to place two major constraints on the anal-
ysis. First, as was discussed in the assumptions (
§
4.1), we
found it necessary to divide the 90 mT of cargo into 10
smaller payload missions to reduce the trip times to some-
thing reasonable (in the neighborhood of 2 years). Sec-
ond, even with the lower payload mass, the heliocentric
portion of the mission was being optimized for trip times
much longer than two years. To limit the mission length,
300 days was chosen for the heliocentric portion. Essen-
tially, we picked the trip time and power available and
from that allowed the code to find which thruster would
use the least propellant to accomplish the mission. The
differences in trip time come largely from the different
times requires to spiral out of earth orbit. The assumption
of restricting the heliocentric trip time did have one signif-
icant drawback. The code did not converge for the high-
power bismuth Hall thruster in table(2), since its thrust-to-
power ratio was too low to complete the heliocentric trip
in 300 days.
Within the accuracy limits of our calculation, the
AF-LiLFA and the Xe-Hall thrusters required the lowest
launch mass fraction (42 to 45%, i.e., 7 to 8 mT). How-
ever, the time to spiral out of earth orbit for the AF-LiLFA
is 140 to 178 days longer for a savings of only 1 to 2
mT over the Hall thrusters. Since this is close to our es-
timated error (
±
1
mT) it is not clear that the AF-LiLFA
would provide significant mass savings over any of the
Hall thrusters. From this analysis, the arcjet would not
be the preferred choice for this mission. With a propel-
lant mass fraction of 65% (18 mT), it required signifi-
cantly more propellant mass and took longer time to ar-
rive at Mars (712 days) than the AF-LiLFA or the Xe-
Hall thrusters. The argon MPDT required the highest pro-
pellant fraction, 85% (55 mT), and took the longest(874
days) of the options considered.
Sankaran, Cassady, Kodys, and Choueiri: PROPULSION OPTIONS FOR MISSIONS TO MARS
9
4.5
Piloted Mission
For the piloted portion of the mission, the desired propul-
sion option would be the one which accomplishes the mis-
sion in the least amount of time. Trip times ranged from
490 days for the H-MPDT-3 ([21]) to just under 340 days
for the LiLFA. The 3.75 MW H-MPDT-2 ([25]) had a
trip time of only about one month longer (380 days) than
the LiLFA. However, the initial mass required was also
higher than the LiLFA. Due to its high specific impulse,
the MPDT-3 required the least propellant mass fraction,
33% (45 mT), for the mission. However, its low efficiency
(25%), and its lower thrust-to-power (compared to other
choices at the same power level), prevented it from being
competitive because of long trip time.
The range of trip times (340-490 days) depends upon
the power level chosen (7.5 MW) for this stage of the mis-
sion. At that power level, the LiLFA is the best option,
with the minimum trip time and a moderate propellant
requirement (57% = 129 mT lithium) compared to other
choices.
5
Other Viable Candidates
As mentioned in
§
3.1, we restricted our analysis to
thrusters that have measured performance data, and have
demonstrated significant lifetime. This eliminated many
thruster concepts that may be promising for this mission.
Ion propulsion has demonstrated the in-space perfor-
mance and lifetime necessary to be incorporated into fu-
ture mission design and planning [35]. Good power throt-
tling over a rather broad power range make ion propul-
sion ideal for solar electric propulsion missions where the
electric power available varies with distance from the sun.
In fact, the next-generation ion thrusters are being devel-
oped for such missions. NEXT, NASAâs Evolutionary
Xenon Ion Thruster, will provide higher power capabil-
ities and lower specific mass with slightly increased ex-
haust velocities over Deep Space 1 technology[36]. These
advances will meet the requirements of several near-term
planetary missions including a Neptune orbiter and a Ti-
tan explorer[36].
Due to the electrostatic nature of ion propulsion,
increased power (exhaust velocities) and propellant
throughput (thrust) require corresponding increases in
thruster size. The 30-cm DS1 thruster was capable of
operation at up to 2.5 kW. The NEXT thruster will in-
crease the effective area by 2, by moving to 40-cm diam-
eter optics, and power capabilities near 10 kW. NASAâs
long-range goal for the development of ion engine tech-
nology is the demonstration of operation at 30 kW and
above [35]. Work in 1968 investigated the feasibility of
much higher power (
>
100 kW) ion thrusters. Preliminary
tests on a 150-cm engineering model showed that opera-
tion at 177 kW was possible with exhaust velocities in ex-
cess of 7000s and calculated efficiencies of 76%. Thruster
conditioning and grid stability issues arose at this size and
power, as well as a need for higher power electron sources
[37]. Lack of potential missions at that time caused the re-
search program to end before these issues were solved or
thrust measurements could be obtained. However, there
appears to be no fundamental limit on thrusters of this size
and power[35].
Another thruster concept that could be promising is
the pulsed inductive thruster (PIT) [38]. Using ammonia
as propellant, this thruster demonstrated 48% efficiency,
with Isp of 4000 s at discharge energy of 2 kJ per pulse.
If this thruster can be operated at a pulsing frequency
of (
O
(100-1000 Hz)), it would be competitive with the
LiLFA for the piloted mission. However, the PIT has yet
to show potential for lifetime of the order of the mission
duration for it to be a serious candidate.
In addition, there are other thruster concepts, such
as the VAriable Specific Impulse Magnetoplasma Rocket
(VASIMR)[39], that may be suited for this mission. The
VASIMR is a two-stage plasma propulsion device: the
production of the plasma is accomplished in the first stage,
and the heating and acceleration in the second. It is hoped
that the separation of these two processes would allow for
better control of the exhaust velocity, while utilizing max-
imum available power. This device is intended to operate
at power levels ranging from 10 kW to 100 MW. If proven,
its ability to vary specific impulse independent of power
(which will likely require varying the propellant), can re-
duce both trip time and propellant utilization. However,
this device has not yet been successfully operated in the
laboratory as a thruster, and propulsive characteristics and
performance have not been directly measured.
6
Concluding Remarks
The goal of this study was to examine electric propulsion
options for near-term (10-20 years) cargo and piloted mis-
sions to Mars. Thrusters for the study were chosen from
the highest performance data available, subject to the fol-
lowing constraints: that they had demonstrated operation
at power levels of 25 kW (cargo) or 500 kW (piloted) in a
single laboratory thruster, that thrust measurements at this
power level had been published, and has demonstrated a
potential for lifetimes on the order of at least 100 hours.
Power levels chosen for this study were 150 kW for the
cargo mission and 7.5 MW for the piloted mission. Tra-
jectory analysis was performed by the NASA-JSC RAP-
Sankaran, Cassady, Kodys, and Choueiri: PROPULSION OPTIONS FOR MISSIONS TO MARS
10
TOR code which optimized acceleration for the heliocen-
tric portion of the mission.
The cargo mission results showed that several of the
thrusters we considered are promising candidates. For a
chosen power level of 150 kW, the AF-LiLFA and all three
of Hall thrusters considered could deliver the 9 mT pay-
load with nearly the same mass in earth orbit.
For the piloted mission at 7.5 MW, the lithium Lorentz
force accelerator (LiLFA) provided trip times savings of
at least one month over any of the MPDTs in the study.
The initial mass required to accomplish this was in the
middle of the range of the thrusters considered. Overall,
the LiLFA seems to be a promising technology for high-
power, high
â
V
missions of this type. Because the power
available for this mission is fixed at 7.5 MW, the range of
trip times (340-490 days) is longer than the estimates in
other studies that consider much higher power levels.
This study provides a survey of electric propulsion
options for cargo and piloted Mars missions. In order
to more completely determine the relative strengths and
weaknesses of each system considered, several of our as-
sumptions need to be addressed in future work. So far,
we have completed only the first phase of this study, with
strong assumptions on specific mass of components, and
thruster operation at a single power level only. The next
step would be to perform trajectory analysis for each mis-
sion stage at a range of power levels. This would al-
low for the determination on of the optimum thruster and
power level for a given mission. In addition, the assump-
tions of constant thruster and power supply specific mass
for all thruster options might have influenced our results,
given the large variations in power requirements of each
thruster. Obtaining better estimates of these values would
increase the relevance of our results. Finally, in the next
phase of this work, a parametric study of thruster effi-
ciency and specific impulse will be undertaken, which
could provide guidelines for future research in thruster de-
sign and optimization.
Figure 5: Results of the cargo mission analysis (all set to arrive on 12/1/2016), with increasing propellant consumption
from left to right.
Sankaran, Cassady, Kodys, and Choueiri: PROPULSION OPTIONS FOR MISSIONS TO MARS
11
Figure 6: Results of the piloted mission analysis (all set to arrive on 12/1/2018), with increasing trip time from left to
right.
References
[1] The Planetary Society. Nuclear Space Initiative: A Planetary Society White Paper. The Planetary Report, 2002.
[2] K.E. Tsiolkovsky. Exploration of the Universe with Reaction Machines. The Science Review, 5, 1903. English
translation from the website of Tsiolkovsky State Museum of the History of Cosmonautics.
[3] J. E. Polk, R. Y. Kakuda, J. R. Anderson, J. R. Brophy, V. K. Rawlin, M. J. Patterson, J. Sovey, and J. Hamley.
Validation of the NSTAR Ion Propulsion System on the Deep Space 1 Mission. AIAA-99-2274, 1999.
[4] R.G. Jahn. Physics of Electric Propulsion. McGraw-Hill, 1968.
[5] E. Stuhlinger and J.C. King. Concept for a Manned Mars Expediction with Electrically Propelled Vehicles.
Electric Propulsion Development: Progress in Astronautics and Aeronautics, 9:647â664, 1963.
[6] E.P. Coomes, D.Q. King, and M.J. Patterson. PEGASUS: A Multi-Megawatt Nuclear Electric Propulsion System.
Third Annual Symposium on Space Nuclear Power, 1986.
[7] D.Q. King and J.C. Sercel. A Review of the Multi-Megawatt MPD Thruster and Current Mission Applications.
AIAA-86-1437, 1986.
[8] J.H. Gilland and R.M. Myers. Multimegawatt Electric Propulsion System Design Considerations. AIAA-90-2552,
1990.
[9] John S. Clark et. al. Nuclear electric propulsion: A âbetter, safer, cheaperâ transportation system for human
exploration of mars. In Space Technology and Application International Forum, Albequrque, NM, USA, 1994.
January 9-13.
Sankaran, Cassady, Kodys, and Choueiri: PROPULSION OPTIONS FOR MISSIONS TO MARS
12
[10] D.G. Pelaccio, G.A. Rauwolf, G. Maggio, S. Patel, and K. Sorenson. An Examination of Emerging In-Space
Propulsion Concepts for One-Year Crewed Mars Misisons. Annual Symposium on Space Nuclear Power, 2002.
[11] Robert H. Frisbee, Nathan J. Hoffman, and Kathy H. Murray. Sp-100 dynamic power and lithium propellant
MPD nuclear electric propulsion technology requirements. In Space Technology and Application International
Forum, Albequrque, NM, USA, 1994. January 9-13.
[12] R.H. Frisbee and N.J. Hoffman. Electric Propulsion Options for Mars Cargo Missions. AIAA-96-3173, 1996.
[13] James E. Polk et. al. Technology requirements for high-power lithium lorentz force accelerators. In Space
Technology and Application International Forum, Albequrque, NM, USA, 2001.
[14] M. Noca, J.E. Polk, and R. Lenard. An Evolutionary Strategy for the Use of Nuclear Electric Propulsion. Annual
Symposium on Space Nuclear Power, 2000.
[15] G. Woodcock et al. Benefits of nuclear electric propulsion for outer planet exploration. AIAA-02-3548, 2002.
[16] M. Martnez-Sanchez and J.E. Pollard. Spacecraft electric propulsion- an overview. Journal of Propulsion and
Power, 14:688â699, 1998. September-October.
[17] R.G. Jahn and E.Y. Choueiri. Electric Propulsion. Academic Press Encyclopedia of Science & Technology, 3rd
edition(5), 2002.
[18] K. Toki, Y. Shimuzu, and K. Kuriki. Electric Propulsion Experiment (EPEX) of a Repetitively Pulsed MPD
Thruster System Onboard Space Flyer Unit (SFU). In International Electric Propulsion Conference, 1997. IEPC
97-120.
[19] K.E. Clark and R.G. Jahn. Quasi-Steady Plasma Acceleration. AIAA Journal, 8:216â220, 1970.
[20] Akihiro Sasoh and Y. Arakawa. Electromagnetic Effects in an Applied-Field Magnetoplasmadynamic Thruster.
Journal of Propulsion and Power, 8:98â102, 1992. January-February.
[21] E.Y. Choueiri and J.K. Ziemer. Quasi-Steady Magnetoplasmadynamic Thruster Performance Database. Journal
of Propulsion and Power, 17:967â976, 2001. September-October.
[22] LaPointe Michael R. and Pavlos G. Mikellides. Design and operation of mw-class mpd thrusters at the nasa
glenn research center. In Joint Propulsion Conference, Indianapolis, IN, USA, July, 2002. AIAA 2002-4113.
[23] R. Myers, M. Mantentiek, and M. LaPointe. MPD Thruster Technology. Conference on Advanced SEI Technolo-
gies, AIAA-91-3568, September, 1991.
[24] T. Wegmann et. al. Experimental comparison of steady-state nozzle type and cylindrical MPD thrusters at high
current levels. In International Electric Propulsion Conference, Seattle, WA, USA, 1993. IEPC 93-122.
[25] H. Tahara, H. Yasui, Y. Kagaya, and T. Yoshikawa. Development of a Quasi-Steady MPD Arcjet Thruster for
Near-Earth Missions. AIAA-87-1001, 1987.
[26] James S. Sovey and Maris A. Mantenieks. Performance and Lifetime Assessment of Magnetoplasmadynamic
Arc Thruster Technology. Journal of Propulsion and Power, 7:71â83, 1991. January-February.
[27] J. E. Polk. Mechanisms of Cathode Erosion in Plasma Thrusters. PhD thesis, Princeton U., 1995.
[28] H. L. Kurtz, M. Auweter-Kurtz, W. Merke, and H.O. Schrade. Experimental MPD Thruster Investigations.
Journal of the British Interplanetary Society, 41:223â232, 1998.
[29] V.P. Ageyev and V.G. Ostrovsky. High-current stationary plasma accelerator of high power. In Proceedings of
the 23rd International Electric Propulsion Conference, Seattle, WA, USA, 1993. IEPC-93-117.
[30] D.H. Manzella, R.S. Jankovsky, and R.R. Hofer. Laboratory Model 50 kW Hall Thruster. AIAA-2002-3676,
2002.
Sankaran, Cassady, Kodys, and Choueiri: PROPULSION OPTIONS FOR MISSIONS TO MARS
13
[31] S. Tverdokhlebov, A. Semenkin, and J. Polk. Bismuth Propellant Option for Very High Power TAL Thruster.
AIAA-2002-0348, 2002.
[32] D.T. Jacobson and R.S. Jankovsky. Performance Evaluation of a 50 kW Hall Thruster. NASA-TM-1999-209447,
1999.
[33] G. Popov, V. Kim, V. Tikhonov, S. Semenikhin, and M. Tibrina. The Fourth (Final) Quarterly Report on the
Milestones (a)(4) and (a)(5)(D) of SoW of Contract No 960938 Between RIAME-MAI and JPL to NASA-JPL
(Items 8 and 9 of Delivery Schedule). RIAME-MAI Technical Report, Moscow, Russia, 1998.
[34] M. Auweter-Kurtz, T. Golz, H. Habiger, F. Hammer, H. Kurtz, M. Riehle, and C. Sleziona. High-Power Hydrogen
Arcjet Thrusters. Journal of Propulsion and Power, 14:764â773, 1998.
[35] Matthew T. Domonkos, Michael J. Patterson, and Robert S. Jankovsky. Ion engine and hall thruster development
at the NASA Glenn Research Center. In International Mechanical Engineering Congress, New Orleans, LA,
USA, November 17-22 2002. IMECE 2002-34444.
[36] Steven Oleson et al. Mission Advantages of NEXT: NASAâs Evolutionary Xenon Thruster. In AIAA Joint
Propulsion Conference, Indianapolis, IN, July 7-10 2002. AIAA 2002-3969.
[37] S. Nakanishi and E.V. Pawlik. Experimental Investigation of a 1.5-m-diam Kaufman Thruster. Journal of
Spacecraft and Rockets, 5:801â807, 1968. July.
[38] C.L. Dailey and R.H. Lovberg. Pulsed Inductive Thrusters Performance Database for Megawatt Class Engine
Applications. AIAA-93-0103, 1993.
[39] F. R. Chang Diaz, J. P. Squire, R. D. Bengston, B. N. Breizman, F. W. Baity, and M. D. Carter. The Physics and
Engineering of the VASIMR Engine. AIAA-00-3756, 2000.