-
Logistics and Operations versus Heavy Lift:
Examining Approaches to Human Exploration in a
Cost-Constrained Era
David L. Akin,�
University Of Maryland, College Park, MD, 20742, USA
Although recent review panels have called into question the
economic viability of ad-vanced heavy-lift vehicles, the
conventional wisdom still demands some form of shuttle-derived
heavy-lift launch vehicle prior to initiating human exploration
beyond low Earthorbit. Recent publications by the author have
demonstrated that existing evolved expend-able launch vehicles,
speci�cally the current version of the Delta IV Heavy, along with
asmaller human spacecraft and in-orbit modular propulsion stages,
are capable of supportinga robust and extensible program of human
lunar exploration, starting from single-vehiclelunar orbital
missions to �ve-launch scenarios for lunar landing and return. This
systemprovides routine lunar surface access for both humans and
cargo, based on a architectureutilizing a low lunar orbit logistics
site for stockpiling propulsion stages and supporting theassembly
of lunar landing vehicles via autonomous rendezvous and docking.
These priorpublications have also documented probabilistic risk
analyses which demonstrate that themodular approach is capable of
equal or higher reliability than a monolithic heavy-lift mis-sion,
due to redundancy in propulsive options from active spare
propulsion vehicles basedin low lunar orbit.
This paper continues and extends the analysis of a
cost-constrained modular approachto exploration by examining the
potential of such a system to provide access to otherFlexible Path
sites, including human missions to near-Earth objects and Mars
orbit. Insome cases, in-space technology additions such as inatable
habitats for longer-durationhuman missions will su�ce to support
these extended-range objectives. For the di�cultgoal of human Mars
missions, the analysis will examine the feasibility of human
missionssupported solely by current EELV launch vehicles, and will
perform trade studies againstmissions with smaller numbers of
launches by investigating the impact of larger modularpropulsion
stages and larger vehicle currently under private development. Even
with thesefar more ambitious mission objectives, the analyses
documented in this paper still supportsthe basic concept of \spend
the money ying", rather than postponing human explorationto await
the more elegant solution of heavy-lift launch vehicles.
Acronyms
CPM Cryogenic Propulsion ModuleDIVH Delta IV HeavyEELV Evolved
Expendable Launch VehicleEVA ExtraVehicular ActivityFH Falcon
HeavyGEO Geostationary OrbitLEO Low Earth OrbitLLM Lunar Landing
ModuleLLO Low Lunar OrbitLMO Low Martian OrbitLOI Lunar Orbit
Insertion
�Director, Space Systems Laboratory. Associate Professor,
Department of Aerospace Engineering. Senior Member, AIAA
1 of 15
American Institute of Aeronautics and Astronautics
AIAA SPACE 2011 Conference & Exposition27 - 29 September
2011, Long Beach, California
AIAA 2011-7221
Copyright © 2011 by University of Maryland . Published by the
American Institute of Aeronautics and Astronautics, Inc., with
permission.
-
MMH Monomethyl HydrazineN2O4 Nitrogen TetroxideNASA National
Aeronautics and Space AdministrationOPM Orbit Propulsion ModuleSSL
Space Systems LaboratoryTEI Trans-Earth InsertionTLI Trans-Lunar
InsertionUMd University of Maryland
I. Introduction
There would be little argument to the assertion that, as of the
time of writing this paper, the U.S. humanspace ight program is in
a state of limbo. With the 135th and �nal ight of the space shuttle
program,it is unclear whether future U.S. human access to orbit
will depend on NASA-developed or commercialspacecraft, own on
existing or future launch vehicles. There is a directive that, over
and above continuingthe International Space Station program, human
space ight will be focused on visiting a near-Earth
asteroid,although many in and out of NASA treat this more as a
temporary aberration than an meaningful long-termchange in
destination. Congressional support of NASA seems to be focused on
its value as a high-tech jobsprogram, with the Space Launch System
(or, perhaps more accurately, \Senate Launch System") mandatinga
vehicle which will consume most of the available operations funds
for more than a decade, and resultingin a vehicle which may well be
impossible to y economically.1
In counterpoint to the \conventional wisdom" emphasizing the
\Apollo on steroids" approach to humanspace exploration, the
University of Maryland Space Systems Laboratory has been
independently pursuingpossible alternate approaches. The core axiom
of this e�ort can be simply stated: spend the money ying.Rather
than pursue architectures which will require a decade or more of
full NASA funding to create a heavy-lift launch vehicle, this
research e�ort is focused on using existing vehicles and
technologies to facilitate earlyand continuous human space ights in
support of an ambitious program of space exploration.
This study is not limited to avoiding new launch vehicle
development, but indeed to systems which canreduce the size and
cost of architecture elements, even if capabilities are
proportionately reduced from theambitious visions for the
Constellation program. Is support of this e�ort, the following
assumptions havebeen adopted a priori :
� Only existing (or currently funded) launch vehicles will be
considered.
� New systems will be minimized in number, size, and complexity
to minimize nonrecurring costs anddevelopment time.
� Crew and spacecraft sizes will minimized to save mass and
cost.
� New technologies will be restricted to only those which are
enabling, and which are demonstrated toproduce clear and immediate
improvement in system e�ectiveness.
� Cost mitigation will focus on learning curve e�ects from
economy of scale in larger production runs,rather than
reusability.
� Cost analysis will include opportunity costs of money, which
favors higher operational costs if so doingminimizes initial
nonrecurring costs.
These assumptions were used to de�ne a series of speci�c design
choices which have been examinedthrough trade studies and
propagated throughout the study. These speci�c architecture
decisions include
� 3 person nominal crew size
� Propellants limited (at least at �rst) to standard storable
hypergolic systems
� No propellant transfer or depots
� Multi-launch vehicle on-orbit integration limited to
conventional docking technologies
2 of 15
American Institute of Aeronautics and Astronautics
-
In the �rst phase of this study, analysis2 showed that small
spacecraft supporting 2-3 person crewswere suitable for Delta IV
Heavy (DIVH) missions, up to and including single-launch human
lunar orbitmissions. This required the development of a crew launch
and entry vehicle below 6500 kg (including theaccommodation of a
launch escape system for human launch), and a standard orbital
propulsion module(OPM) for lunar orbit insertion (LOI) and
trans-Earth insertion (TEI). Trade studies showed that
on-orbitlogistics and operations for lunar missions was best
performed in lunar orbit, to take maximum advantageof the high
speci�c impulse of cryogenic upper stages of the launch vehicles.
This resulted in a 9980 kg masslimit for payloads delivered to
trans-lunar insertion (TLI). Using standard vehicle performance
parametersbased on regression analysis of past vehicles, it was
shown that the favored method for payload transportto LLO is to
create a standard orbital propulsion module which matches the
maximum payload delivered.Rather than design a smaller module to
provide the 837 m/sec lunar orbit insertion �V, the
architectureadopted used two OPMs per launch: one fully loaded with
propellant to be delivered to LLO, and one witha partial propellant
load su�cient to deliver the fully loaded one into lunar orbit.
This stage of the analysisidenti�ed a program plan which consisted
of four three-crew ISS rotation missions per year, two human
lunarmissions per year each with a pre-emplaced cargo module of
1870 kg at the landing site, and a Flexible Pathmission in
alternate years. By using existing launch vehicles and limiting new
vehicle development to thepropulsion module, a moderate OPM variant
incorporating landing gear and avionics for terminal landingon the
moon, and a small (4700 kg) human spacecraft, this scenario allowed
reaching operational missionstatus in less than a decade and
accommodated steady-state ight operations at this scale for less
than $3Bper year total.
One of the obvious concerns about a system with multiple
launches and in-orbit docking operations is theoverall reliability
of a mission with so many critical elements. In the second phase of
this ongoing research,analysis focused on probabilistic risk
analysis and mitigating strategies. Since most of the Earth
launchescarry standard interchangeable OPMs, it was shown3 that the
ability to store additional mission elements atthe LLO staging
location produced the ability to increase mission reliability to
levels equalling or exceedingmore monolithic architectures. This
phase also examined the use of human-rated Atlas 402 launch
vehiclesfor the ISS crew rotation missions, as the more expensive
DIVH has substantially more payload performancethan required for a
simple LEO mission.
The current (third) phase of this research project is focused on
two parallel e�orts: increasing the �delityof the vehicle design
parameters through more sophisticated models and sensitivity
analyses, and lookingin detail at evolutionary missions including
human missions to near-Earth objects (NEOs) and to Phobosand low
Martian orbit (LMO). Assumptions made in the earlier phases as to
vehicle design parameters aretested against more detailed models,
and the revised baseline architecture components will be analyzed
forapplication to lunar, NEO, and Mars missions.
II. Revisiting Vehicle Design Estimates
Before going further into analysis on more ambitious missions,
it would make sense to revisit past as-sumptions to verify that
best practices are being followed, and that data to date �ts known
trends wherepossible. To this end, several of the more speculative
assumptions from past publications will be analyzedhere for
correspondence to known data, and used to reassess past studies
based on more accurate (andinvolved) predictive models.
A. Mass-Dependent Inert Mass Fraction
As summarized above. past analyses demonstrated the bene�ts of
adopting a single propulsion vehicle design,which is sized to be
delivered fully fueled to LLO by a second propulsion module
launched partially fueled.Based on a regression analysis of similar
vehicles, a stage inert mass fraction
� =minert
minert +mprop= 0:10 (1)
was assumed.4 This value was further assumed to be invariant
based on total propulsion module size. For the9980 kg TLI injected
payload of the Delta IV Heavy, the optimum worked out to be a fully
fueled propulsionmodule mass of 6950 kg, with 695 kg of inert mass
and 6255 kg of propellants. Since one of these propulsionmodules is
used for crew module ascent from the moon, the ascent �V of 2334
m/sec from the surface of
3 of 15
American Institute of Aeronautics and Astronautics
-
the moon into low lunar orbit produces a maximum lunar crew
module size of 4966 kg. A survey of pasthuman spacecraft in
reference2 showed that this was entirely feasible, falling slightly
behind Apollo in entryvehicle mass but considerably larger than
Gemini.
However, there has long been a item of \conventional wisdom"
that there is a physical economy of scalein larger vehicle systems.
From a survey of published sources of mass estimating relations,
the followingequations for stage inert mass fractions were derived
by regression from past vehicle data:
�storables = 1:6062 (Mstagehkgi)�0:275 (2)
�LOX=LH2 = 0:987 (Mstagehkgi)�0:183
(3)
The trend of these equations as a function of stage gross mass
is seen in Figure 1. Smaller stages are ata clear disadvantage due
to higher stage inert mass fractions. It is also clear that the
prior assumption of�=0.10 was particularly optimistic, given the
focus on small stage sizes in this architecture.
Figure 1. Trend of stage inert mass fraction with stage gross
mass for high-density (e.g., storable) and low-density (LOX/LH2)
propellants
B. Lunar Orbit Delivery Architecture
In the �rst phase of this study, the conclusion was reached that
delivering loaded OPMs to lunar orbit wasbest done by launching two
OPM systems at a time: one fully loaded with propellant to perform
requiredmissions to and from the moon, and one partially fueled to
perform the lunar orbit insertion (LOI) maneuver.Given the
conversion to mass-dependent inert mass fraction estimation, it
seems reasonable to revisit thisidea as compared to the creation
and use of a ideally-sized LOI stage.
Figure 2 shows the e�ect of launch vehicle size on the relative
sizes of payload delivered. Although thereis a penalty associated
with the use of a partially-loaded OPM stage for LOI, the relative
size of the masspenalty is small. Figure 3 shows a much more
critical variation in these approaches. With two separatevehicles
to be designed and developed, the nonrecurring costs for the
optimized LOI stage approach is almosttwice as large as for the use
of a single module design for both LOI and OPM functions. Based on
theseresults, the original decision for a single OPM design
stands.
It should be noted that Figures 2 and 3 illustrate the rationale
behind the basic assumption of this studythat focus should be
placed on existing upper-end launch vehicles, rather than waiting
to develop super-heavy launch vehicles. As the size of the launch
vehicle grows, the resultant size of the in-orbit spacecraftgrows,
along with the cost. To minimize \up-front" costs in a program,
keeping all of the vehicles (andtheir non-recurring costs) as small
as practical is key to cost containment. As will be discussed
later, theexception to this paradigm is the size of the human
spacecraft, which tends to be more invariant than thepropulsion
stages.
Prior analysis also assumed the availability of storable
propellant propulsion systems (N2O4/MMH) ata speci�c impulse of 320
seconds; this is feasible, but may be optimistic for this type of
application. Atrade study was performed to investigate the
sensitivity of OPM module size achievable as a function
ofpropulsion system exhaust velocity. Figure 4 shows that the
vehicle size is not stronly sensitive to speci�c
4 of 15
American Institute of Aeronautics and Astronautics
-
Figure 2. Mass trends of OPMs with and without ded-icated LOI
stages
Figure 3. Total cost for OPM architecture with andwithout
dedicated OPM stages
impulse. Nonetheless, the decision was made to reduce the
assumed speci�c impulse to 310 seconds to beconservative.
Figure 4. E�ect of speci�c impulse on OPM performance
III. Mission Applications
As a result of the trade studies, it was clear that the
assumptions of the previous phases of this analysiswere essentially
correct, but optimistic. The logical next step was clearly to
revisit the baseline design forhuman lunar exploration in light of
these revised baseline parameters.
A. Lunar Exploration Program
The basic concept of operations for a logistics-based human
lunar mission is shown in Figure 5. A seriesof Delta IV Heavy
launches deliver three Orbital Propulsion Modules, one Lunar
Landing Module, and ahuman spacecraft. The assembly is docked
together in LLO and used for a lunar landing. The
landedcon�guration is shown in Figure 6. A key sizing factor was
the desire to use a standard fully-fueled OPM tolaunch the crew
module back into low lunar orbit, where it would rendezvous and
re-dock to the same OPMwhich brought it into LLO, which has
su�cient propellant to perform the TEI burn for return to Earth
anda direct entry and landing.
The size of the human spacecraft is bounded by two constraints:
keeping within the lunar launch payloadcapability of an OPM, and
keeping it spacecraft mass low enough that it can be launched with
su�cientpropellant in its OPM to perform both a lunar orbit
insertion and a trans-Earth insertion. This constraintwas selected
for crew safety: while a number of \spare" OPMs will be based in
LLO, this assumption ensures
5 of 15
American Institute of Aeronautics and Astronautics
-
Figure 5. Original baseline CONOPS for human lunar landing2
Figure 6. Original vehicle con�guration post-lunar landing2
6 of 15
American Institute of Aeronautics and Astronautics
-
that the crew has the capability to abort to Earth at any point
in the mission (up until lunar landing) withoutthe requirement for
precursor orbital operations.
1. Revised DIVH Mission Baseline
Table 1 compares the critical design parameters for the orbital
propulsion modules of the prior study, andafter the revised
baseline assumptions documented above. As can be seen, the revised
assumptions result inan OPM with reduced gross mass, but increased
inert mass. This should allow su�cient margin for detaileddesign
without extraordinary e�orts to maintain minimum mass. Table 2
shows the same data for the lunarlanding module. In both cases, the
LLM was essentially an OPM with an additional 50% inert mass
marginto account for landing gear, descent avionics, and other
systems required for the actual lunar landing phase.
Table 1. Comparison of previous and revised orbital propulsion
module design parameters
Parameter Old Baseline Current Baseline
Gross Mass (kg) 6950 5973
Stage Inert Mass Fraction � 0.10 0.160
Propellant Mass (kg) 6255 5105
Inert Mass (kg) 695 868
Speci�c Impulse (sec) 320 310
Table 2. Comparison of previous and revised lunar landing module
design parameters
Parameter Old Baseline Current Baseline
Gross Mass (kg) 6950 5973
Stage Inert Mass Fraction � 0.15 0.24
Propellant Mass (kg) 5908 4543
Inert Mass (kg) 1042 1430
Speci�c Impulse (sec) 320 310
At this point, it became possible to derive the constraints on
the human spacecraft. In the previous study,a spacecraft mass of
4966 kg was selected to allow a single OPM launch from the moon
back into lunar orbit,which also met the constraint for single-OPM
insertion and departure from LLO. The current baselineanalysis
produced a human spacecraft mass limit of 4886 kg for the LLO
insertion-departure requirement,but the new OPM design will only
deliver 3548 kg to LLO with a single module. The original
report2
discussed required spacecraft mass in some detail, with
reference to past human spacecraft; while 3548 kg iscomparable to a
Gemini spacecraft, it was felt at that stage that a spacecraft mass
of approximately 5000kg would provide for a three-person crew and
direct entry return to Earth from a lunar trajectory.
To meet this shortfall, the spacecraft mass was assumed to be
4886 kg from the LLO constraint. Tosuccessfully launch this
spacecraft from the lunar surface back into LLO, two OPMs will be
landed with itfor lunar launch; one fully fueled, and one loaded
with 1670 kg of propellant. While it seems wasteful touse an OPM
with only 28.3% of its propellant loaded, the only other choices
would be to either design aspecialized smaller version of the OPM,
or to develop a larger lunar-launch stage and use a larger
launchvehicle to deliver it.
The revised baseline lunar landing system can now be de�ned.
Where the original baseline requiredone human spacecraft, one LLM,
and four OPMs for six DIVH launches per human lunar landing, the
newbaseline requires six OPMs (one partially loaded), a LLM, and a
human spacecraft for eight DIVH launchestotal. Assuming the current
DIVH costs of approximately $300M per launch, the total launch
costs for thisapproach would be $2.4B per mission, which is still
competitive with expected launch costs of the SpaceLaunch
System.
In an active, robust program of lunar exploration, there will be
a signi�cant need for routine cargo deliveryto the moon. Whether
for landing exploration robots or pre-emplacing logistics and
equipment for future
7 of 15
American Institute of Aeronautics and Astronautics
-
human landings, the components developed for human missions can
also provide one-way cargo deliveries tothe lunar surface at a
variety of payload sizes, as detailed in Table 3.
Table 3. One-way cargo capabilities to the lunar surface
Vehicle Con�guration Cargo Landed (kg)
LLM only 1732
LLM + 1 OPM 5025
LLM + 2 OPMs 8292
LLM + 3 OPMs 11,553
LLM + 4 OPMs 14,811
2. Mixed Fleet Options
Since the prior publications of this study, SpaceX has announced
the development of Falcon Heavy, a three-core heavy-lift version of
the Falcon 9. Recognizing that all SpaceX vehicles are at an early
stage of maturitycompared to EELVs, Falcon Heavy is nonetheless a
heavy-lift vehicle which is both funded and in activedevelopment.
For that reason, it will be used to examine the e�ects of a
heavy-lift launch vehicle on thislunar exploration
architecture.
Falcon Heavy is projected to carry 53,000 kg to LLO and 16,000
kg to TLI. Using the same paradigmwith two identical OPMs, using
one partially fueled to deliver the fully-fueled OPM into LLO,
producesa Falcon Heavy-based orbital propulsion module (OPM-F)
design with a gross module mass of 9739 kg,propellant mass of 8483
kg, and inert mass of 1256 kg (�=0.144). This module, used alone,
will be capableof carrying 6082 kg from the lunar surface to LLO,
which represents the baseline human spacecraft and 1196kg of
optional up-cargo. This scenario reduces the total launches to
seven, replacing two DIVH launcheswith a single Falcon Heavy. Since
OPMs come in integer quantities, four fully-fueled OPMs and the
LLMprovide su�cient performance to land the spacecraft, OPM-F for
lunar launch, and 186 kg of down-cargo.
Two other cases in the mixed-eet model were also examined. In
case 3, all OPMs were replaced byOPM-F stages, with only the DIVH
version of the LLM still used. For the last case, an LLM-F was
designedbased (as in the case of the DIVH LLM) on a 50% increase in
inert mass for the OPM-F. The LLM-F massparameters are therefore a
gross mass of 9739 kg, a propellant mass of 7855 kg, and an inert
mass of 1884kg.
Table 4 summarizes the four cases considered with the revised
baseline architecture. It should be notedthat every OPM used for
lunar landing or launch requires a second OPM for the lunar orbit
insertionmaneuver, which is then expended. Similarly, LLM and human
spacecraft each require an OPM for LOI.The table also lists the
available up-cargo and down-cargo masses, which represent the
payload mass marginsfor landing and launch maneuvers based on
fully-fueled OPMs, as well as the total requirement for DIVHand FH
launch vehicles.
It should be noted that, even for mission architectures which
are otherwise exclusively launched on FalconHeavy vehicles, it is
assumed that the human spacecraft still ies on a Delta IV Heavy
with an OPM asmaneuvering stage. This is justi�ed on the basis that
DIVH is currently ying, and human-rating the DeltaIV family is
already under consideration. While SpaceX intends to human-rate all
of its launch vehicles, thelarger size of FH is ill-matched to the
relatively small size of the baseline human spacecraft.
Further analysis is important to examine the various lunar
options on the basis of cost and reliability.Even the most capable
of the options considered (all FH-sized modules) still uses six
ights as comparedto �ve for the old baseline; this is indicative of
the reduced performance (in terms of both � and Isp) ofarchitecture
components in the more recent analysis. Future e�orts will seek to
�nd optimum program-level plans which accommodate upgrades in
capability through phased development of new architecturecomponents
throughout an ongoing ight program.
B. Selecting Extended Exploration Parameters
With the revisions to the parametric description of the in-space
logistics architecture, the next issue to exam-ine is the
applicability of these components to extended space exploration.
Two next-generation evolutionary
8 of 15
American Institute of Aeronautics and Astronautics
-
Table 4. Summary of mission logistics for mixed-eet human lunar
surface missions
All DIVH OPM-F for All OPM-Fs All OPM-Fs
modules launch only and DIVH LLM and LLM-F
Mission Con�guration Spacecraft Spacecraft Spacecraft
Spacecraft
2 OPM (ascent) 1 OPM-F (ascent) 1 OPM-F (ascent) 1 OPM-F
(ascent)
1 LLM 1 LLM 1 LLM 1 LLM-F
4 OPM (descent) 4 OPM (descent) 3 OPM-F (descent) 3 OPM-F
(descent)
OPMs 14 10 2 1
LLM 1 1 1 0
OPM-Fs 0 2 8 9
LLM-F 0 0 0 1
Up-cargo (kg) 0 1196 1196 1196
Down-cargo (kg) 1414 186 3381 0
DIVH launches 8 6 2 1
FH launches 0 1 4 5
mission goals will be considered: a six-month trip to a
near-Earth object (NEO), and a human mission toPhobos, the inner
moon of Mars.
Prior to the detailed analysis of deep-space missions, it is
important to consider the choice of a construc-tion and check-out
site for the mission preparation. This could be anywhere in
cis-lunar space, excludingonly Earth orbits above approximately
1000 km altitude up to 65,000 km altitude, due to the
heightenedradiation environment due to the Van Allen radiation
belts. Other speci�c locations include the �ve librationpoints of
the Earth-Moon system.
The original trade study of optimum staging location showed that
the most advantageous staging sitefor lunar exploration is low
lunar orbit. Given the signi�cantly higher performance of the
LOX/LH2 upperstage of the Delta IV Heavy, it was highly
advantageous in that analysis to get the staging site as farout of
Earth’s gravity well as practical via direct injection of payloads
upon launch. Either of the stableEarth-Moon librations points, L4
or L5, require nearly equal performance for orbit insertion as the
selected100 km low lunar orbit used for lunar exploration staging.
For this reason, it was decided to consider theEarth-Moon L4
libration point as one possible option for space-based logistics
caching and vehicle assembly.The obvious alternative, as in the
original study of logistics site for the lunar program, is low
Earth orbit.Both of these potential sites will be considered in
this study.
C. Near Earth Object Missions
Earlier this year ENAE 484, the University of Maryland senior
capstone class in spacecraft design, performeda detailed study of a
human mission to four NEO candidates between 2020 and 2030. The
critical elementsof the mission design chosen were a round-trip
travel time of six months or less, and a total mission �V notto
exceed 7 km/sec from the LEO staging site assumed for that
study.
Two of the four candidate NEO targets from the ENAE 484 study
were adopted for this analysis:2007XB23 and 2001QJ142. These two
targets were the greatest and least �Vs from a LEO departure forthe
four considered. The calculated mission �Vs were recalculated for
Earth departure from L4 rather thanLEO, which brought all of the
L4-based mission �Vs into the range of 4.0-4.5 km/sec. The �V
requirementsby mission increments are detailed in Table 5.
The same basic human spacecraft from the lunar analysis
(mass=4886 kg) was kept as the launch andentry spacecraft. However,
the much greater mission duration of a NEO mission will require
substantialadditional human support infrastructure. A six-month,
three-crew mission requires 540 crew-days of con-sumables, along
with an expanded crew volume for long-term habitability. Typical
rules of thumb for crewconsumables range from 10 kg/day for an
open-loop system, down to 1 km/sec for an aggressive degree
ofrecycling of water and air loops. A detailed design study for an
extended-duration lunar habitat concluded
9 of 15
American Institute of Aeronautics and Astronautics
-
Table 5. �Vs for candidate near-Earth object missions (all
values in m/sec)
2007XB23 2001QJ142
LEO departure 4160 3460
L4 departure 3834 1687
NEO arrival 180 1260
NEO departure 310 2040
Total �V (LEO) 4650 6760
Total �V (L4) 4324 4987
that the installed life support system mass would be about 600
kg, with a consumables usage rate of 5crew-day, for a total life
support mass of 3300 kg for three crew for 180 days.5 The crew
would also needexpanded habitable volume for a longer-duration
mission. Recent SSL experience in the development andtesting of
inatable habitats resulted in the demonstration of a
space-quali�able inatable habitat with aninated volume of 60 m3 and
a total mass of 350 kg.6
With these parameters in mind, it was assumed that the human
payload for the NEO missions wouldconsist of the 4886 kg launch and
entry vehicle, along with a 6083 kg supplemental human payload.
Thiswould consist of long-term life support systems and power
generation, crew consumables, and an inatablehabitat module for
additional living volume. The 6083 kg limit is based on the maximum
payload deliverableby the DIVH-based logistics infrastructure,
where the payload is unmanned and does not require an onboardreturn
capability for crew safety. This brings the mission payload to
10,859 kg.
Results of the mission analysis are summarized in Table 6.
Table 6. Summary of mission logistics for NEO missions from
L4
2007XB23 2007XB23 2001QJ142 2001QJ142
DIVH Modules FH Modules DIVH Modules FH Modules
Mission Con�guration Spacecraft Spacecraft Spacecraft
Spacecraft
Hab Module Hab Module Hab Module Hab Module
9 OPMs 5 OPM-Fs 11 OPMs 7 OPM-Fs
OPMs 20 2 24 2
OPM-Fs 0 10 0 14
Payload Margin (kg) 1208 228 0 637
DIVH launches 11 2 13 2
FH launches 0 5 0 7
Clearly, these architectures require a large number of assembled
modules for a mission, with a corre-sponding number of heavy-lift
launch vehicle ights. Unlike the lunar landing case, where a module
failureduring descent will result in an immediate (and redundant)
abort back to LLO and safety, a module failureduring a NEO mission
could strand the crew a long way from home.
To address both the cost and safety concerns, the NEO mission
seemed to be a good point to considerthe incorporation of advanced
technology, speci�cally a long-duration cryogenic propulsion module
for theFalcon Heavy launch vehicle (CPM-F). Given LOX/LH2
technology for a deep-space stage, it is also notunreasonable to
assume the development of a cryogenic upper stage for the Falcon
Heavy (currently underearly development as the \Raptor" stage). A
brief calculation based on comparison to other launch vehicleswith
cryogenic upper stages produced an estimate of 24,000 kg delivered
to C3=0 (lunar orbit/L4).
Rather than continue the approach of ying two identical modules
and using one partially fueled todeliver the other, it would make
sense to examine the pre-existing designs (OPM and OPM-F) as
deliverystages to insert the CPM-F into the L4 logistics site.
Analysis shows that the OPM is too small, but thestorable OPM-F
with a 68% propellant load will deliver a CPM-F with a gross mass
of 16,964 kg, propellant
10 of 15
American Institute of Aeronautics and Astronautics
-
mass of 15,093 kg, and inert mass of 1872 kg.
Table 7. Summary of mission logistics for cryogenic FH NEO
missions from L4
2007XB23 2001QJ142
Mission Con�guration Spacecraft Spacecraft
Hab Module Hab Module
2 CPM-Fs 3 CPM-Fs
OPMs 2 2
OPM-Fs 2 3
CPM-Fs 2 3
Payload Margin (kg) 3175 5893
DIVH launches 2 2
FH launches 2 3
Although the cryogenic propulsion module based on a Falcon Heavy
launch brings the mission scenariodown to 4-5 launches, the large
number of launches required for the storable systems intuitively
would seemto suggest that the propulsion modules are too small for
the larger �Vs of the NEO mission. In order to getthe largest
propulsion modules possible, it makes sense to compare the L4
logistics site to a more traditionalLEO site.
For completeness, Table 8 list the nominal mission
characteristics of the three stages designed and ana-lyzed so
far.
Table 8. Orbital Propulsion Stage parameters for L4 staging
OPM OPM-F CPM-F
Gross mass (kg) 5973 9739 16,964
Propellant mass (kg) 5105 8483 14,150
Inert mass (kg) 868 1256 2814
Speci�c Impulse (sec) 310 310 460
Stage inert mass fraction � 0.1453 0.1290 0.1659
For LEO staging, components such as OPMs can be designed to take
up the entire LEO payload massof the launch vehicle: 23,000 kg for
the DIVH, and 53,000 kg for the Falcon Heavy. For completeness,
bothlaunch vehicles will be considered with propulsion stages using
storable and cryogenic propellants. (DIVHCPMs were not considered
for L4 staging, as it was assumed the very high number of OPMs
required wouldnot be signi�cantly reduced by the higher performance
of the small DIVH CPM stage allowed after transportto the L4
staging site.) The design characteristics for all four of these new
propulsion module designs arelisted in Table 9. Note that the pre�x
\L" is added for a LEO-speci�c propulsion module, so an \LOPM-F"
isa Falcon Heavy-sized orbital propulsion module with storable
propellants designed for use following deliveryto low Earth
orbit.
With all these parameters determined, the various components of
the architecture can be examined todetermine the necessary
arrangement of propulsion modules to accomplish each NEO mission
with each ofthe four candidate approaches to LEO systems logistics.
This is summarized in Tables 10 and 11.
One major di�erence between the architecture tables for the
lunar surface missions as compared to theNEO missions is the
absence of additional modules for transport to a remote staging
site. All of the payloadof the launch vehicles goes directly to a
LEO staging site, eliminating expended stages prior to the
mission.Also, both the crew spacecraft and the additional habitat
module �t easily within the LEO payload capabilityof a single DIVH
launch.
In summary, it is obvious that the LEO logistics base is far
superior to the L4 site for NEO missionsand, by extension, for all
non-lunar missions. By staging from LEO, even storable DIVH
architectures canreach NEO targets for 4-6 launches, and cryogenic
Falcon Heavy systems will allow similar missions with 1
11 of 15
American Institute of Aeronautics and Astronautics
-
Table 9. Orbital Propulsion Stage parameters for LEO staging
LOPM LCPM LOPM-F LCPM-F
Gross mass (kg) 23,000 23,000 53,000 53,000
Propellant mass (kg) 20,631 19,391 48,692 45881
Inert mass (kg) 2369 3609 4308 7119
Speci�c Impulse (sec) 310 460 310 460
Stage inert mass fraction � 0.1030 0.1569 0.0813 0.1343
Table 10. Systems architectures for 2007XB23 missions using LEO
staging
DIVH DIVH Falcon Heavy Falcon Heavy
Storables Cryos Storables Cryos
Mission Spacecraft Spacecraft Spacecraft Spacecraft
Con�guration Hab Module Hab Module Hab Module Hab module
3 LOPM 2 LCPM 2 LOPM-F 1 LCPM-F
Payload Margin (kg) 2973 6124 10,995 7437
DIVH launches 4 3 1 1
FH launches 0 0 2 1
Table 11. Systems architectures for 2007XB23 missions using LEO
staging
DIVH DIVH Falcon Heavy Falcon Heavy
Storables Cryos Storables Cryos
Mission Spacecraft Spacecraft Spacecraft Spacecraft
Con�guration Hab Module Hab Module Hab Module Hab module
5 LOPM 3 LCPM 3 LOPM-F 2 LCPM-F
Payload Margin (kg) 254 904 1606 24,531
DIVH launches 6 4 1 1
FH launches 0 0 3 2
12 of 15
American Institute of Aeronautics and Astronautics
-
manned DIVH launch and 1-2 FH launches of propulsion modules,
depending on the speci�c NEO target.(It should be noted that the
LEO case does not assume the development of the Raptor cryogenic
upper stagefor Falcon Heavy, as it is unnecessary for this
architecture.)
D. Phobos Mission
To further extend the applications of this modular architecture,
the same vehicle components derived forNEO missions in the
preceding section will be applied to a human mission to Phobos, the
inner moon of Mars.Beyond the interest in Phobos in its own right
and as a moon probably gravitationally captured by Marsin the past,
Phobos could provide su�cient mass for highly e�ective radiation
shielding for long-durationhuman missions, using the proximity to
the Mars surface to teleoperate high-performance robots with
verylow latencies in the control loop.
�Vs were calculated for a Phobos mission using standard patched
conic techniques, and are summarizedin Table 12. This table also
presents the �V calculation results for reaching a 100 km altitude
orbit aboveMars rather than Phobos orbit; this lower orbit would be
ideal for staging a Mars landing program, whichis beyond the scope
of the current paper.
Table 12. �V requirements for Mars missions to Phobos and 1000
km orbits
Maneuver Phobos orbit 1000 km orbit
�V (m/sec) �V (m/sec)
LEO departure 3165 3165
L4 departure 2264 2264
Mars orbit arrival 1885 2030
Mars orbit departure 1885 2030
Total �V (LEO) 6.934 7.225
Total �V (L4) 6.034 6.342
While a Phobos mission is operationally just another trip to
another asteroid, the fact that it is farfrom Earth introduces its
own set of challenges. Foremost among them is the additional trip
time; theincrease in mission duration from 6 to (approximately) 24
months drives the life support consumables massup considerably.
Keeping the 5 kg/crew-day parameter used earlier, this would add
8250 kg of consumablesto the mission payload. Longer missions
optimize for higher levels of closed-loop life support, which
shouldreduce this parameter by half. For convenience, it was
assumed that another DIVH-delivered cargo module inL4 would be
devoted to life support consumables, making the total payload mass
for a minimum three-personMars mission 16,832 kg.
Initially, a complete set of L4-staging scenarios was analyzed
for feasibility in a Phobos mission, althoughthe NEO mission
results cast doubt upon the e�cacy of this approach given a 50%
larger payload andan additional 2000 m/sec �V. Results demonstrated
that an L4-based mission using Falcon Heavy-sizedpropulsion modules
would require 16 OPM-F modules to complete the mission, and a
DIVH-based missionwould require 27 OPMs and 30 DIVH ights. Even
though the Falcon Heavy system with cryogenic propul-sion modules
could complete the Phobos mission with only four modules, the much
greater e�ectiveness ofthe LEO staging site demonstrated in the NEO
section led to the abandonment of L4-based scenarios fordeep-space
missions.
Table 13 details the analysis results for the four LEO-based
approaches developed in the NEO section. Allthree of the payload
modules can reach Earth orbit on a single DIVH launch. Although the
DIVH approachwith storable propellants is somewhat cumbersome, six
ights for a storable propellant solution with FalconHeavy launches
are completely in line with baselined missions for the lunar
surface program. Cryogenic FHpropulsion modules can bring a human
Phobos trip down to a three-launch mission, although the
practicalimplications of maintaining liquid hydrogen without boilo�
for a two-year mission would require advancedtechnologies, un-front
nonrecurring development funds, and additional mission risk.
13 of 15
American Institute of Aeronautics and Astronautics
-
Table 13. Systems architectures for human Phobos missions using
LEO staging
DIVH DIVH Falcon Heavy Falcon Heavy
Storables Cryos Storables Cryos
Mission Spacecraft Spacecraft Spacecraft Spacecraft
Con�guration Hab Module Hab Module Hab Module Hab module
Logistics Module Logistics Module Logistics Module Logistics
Module
10 LOPM 5 LCPM 5 LOPM-F 2 LCPM-F
Payload Margin (kg) 1358 3166 4463 0
DIVH launches 11 6 1 1
FH launches 0 0 5 2
IV. Future Work
The \holy grail" of the human space exploration program is a
human landing on the surface of Mars. Thispaper has demonstrated
that Phobos is an entirely feasible objective for human space ight
with minimalnew development required, but analyses of feasible
logistics trains for Mars surface exploration will have towait
until a future paper.
This paper also expanded the paradigm of this ongoing research
project from absolute adherence to aphilosophy of \no new
development" (or \spend the money ying") to show that some
destinations willrequire advanced technologies to keep the overall
architecture manageable. Future work should examine theoptimum
choices for additional technology developments. Is cryogenic
propellant maintenance more e�ectivethan aerobraking? When (if
ever) do larger launch vehicles make sense, and does the answer
change if launchvehicle development costs have to be borne by the
human exploration program?
In the same way that this paper re-examined simple assumptions
from earlier work and resulted in moreconservative (and hopefully
accurate) estimation practices, there is a real need to perform a
detailed designstudy of critical elements of the proposed
architecture, such as human spacecraft and baseline
propulsionmodules, to ensure that the parametric estimation used
here results in feasible design estimates. This isespecially true
in the case of Mars landing architectures, where published
estimates of mission infrastructuremay vary in mass by an order of
magnitude or more.
Lastly, while this paper has focused on parametric design and
performance calculations, it is in peril offalling prey to the
"conventional wisdom" of fewer ights is always better. While even
the author would notargue in favor of a mission scenario requiring
30 launches, it should be kept �rmly in mind that the goal hereis
to optimize for the maximum opportunity to y missions within the
strict funding limitations apparentlybeing applied to NASA today.
Costing of these advanced scenarios should proceed to inform
decisions onwhich architectures to select, as well as to perform
optimal sequencing of development programs so that asteady stream
of new technologies and capabilities ow into the program at regular
intervals without starvingan robust ight program from the funding
required for sustenance.
V. Conclusions
As before, the basic conclusion of this paper is that human
space exploration does not have to waita decade or more for the
development of super-heavy lift launch vehicles. Existing,
ight-proven DeltaIV Heavy vehicles, if modi�ed for human-rating,
are fully capable of supporting near-term human lunarsurface
missions without waiting a decade for super-heavy lift launch
vehicles, long-term cryogenic storage,or propellant handling at
orbital depots. While the low lunar orbit staging location, or any
similar high orbitin cislunar space, appears to be disadvantageous
for longer-range exploration, modular spacecraft assembledin low
Earth orbit are fully capable of taking humans to near-Earth
objects or to explore the moons of Mars.
References
1-, \ESD Integration: Budget Availability Scenarios" NASA
Internal Brie�ng, August 19, 2011
14 of 15
American Institute of Aeronautics and Astronautics
-
2David L. Akin, \In-Space Operations: Developing a Path to
A�ordable, Evolutionary Space Exploration" AIAA 2010-2293, SpaceOps
2010 Conference, Huntsville, Alabama, April 25-30, 2010.
3David L. Akin, \Build a Little, Fly a Lot: An A�ordable
Evolutionary Approach to Flexible Path, Lunar Surface, andBeyond"
AIAA 2010-8610, AIAA Space 2010 Conference and Exposition, Anaheim,
California, August 30-September 2, 2010.
4David L. Akin, \Mass Estimating Relations" ENAE 483/788D
Principles of Spacecraft De-sign Lecture Notes, Department of
Aerospace Engineering, University of Maryland, Fall Term
2011hhttp://spacecraft.ssl.umd.edu/academics/483F11/483F11L09.MERs/483F11L09.MERs.pdfi
5David L. Akin, Massimiliano Di Capua, Adam D. Mirvis, Omar W.
Medina, William Cannan, and Kevin Davis, MinimumFunctionality Lunar
Habitation Element, Final Report for NASA Grant NNJ08TA89C,
University of Maryland Space SystemsLaboratory, July, 2009.
6Massimiliano Di Capua, David Akin, Kevin Davis, and Justin
Brannan, "Design, Development, and Testing of an Inat-able Habitat
Element for NASA Lunar Analogue Studies" AIAA-2011-5044, 41st
International Conference on EnvironmentalSystems, Portland, Oregon,
July 17-21, 2011.
15 of 15
American Institute of Aeronautics and Astronautics