Landing Capsule and Rover Designs for the
Sooner Lunar Schooner Mission
D. P. Miller, M. Alfaro, A. Balasubramanyam, Z. Butler, J. Calero,
J. Coplin, B. DeKock, M. Dirckx, A. Huizenga, T. Hunt, M. Moffitt, M. Roman, A. Shah,
J. Stephens, E. Tirado, M. Welker, L. Wilmes, A. Winterholler & J. Yoon
School of Aerospace & Mechanical Engineering
University of Oklahoma, Norman, OK, 73019 USA
Abstract
The Sooner Lunar Schooner mission is a robotic mission to the Moon that will
hopefully launch in the latter half of this decade. The basic mission scenario is as
follows: The spacecraft will be launched and inserted into a Lunar transfer orbit; A
few minutes before impact on the Moon, a braking rocket (e.g., Star 27) will be fired
killing off virtually all of the relative motion between the spacecraft and the Lunar
surface; the spacecraft will come to a stop approximately 50 meters above the
Lunar surface; it will then freefall to the surface; the landing ellipse will be centered
around the Apollo 17 landing site; after landing, the landing capsule will split apart
releasing the two rovers; one rover, capable of doing materials studies will head
for the Apollo 17 LEM. The other, a high speed rover, will start on a traverse to
Lunakhod 2’s last known position, approximately 140km distant.
The mission described above has been developed and is being used as a focus for
a number of multi-disciplinary engineering and science projects at the University of
Oklahoma. During the courses, critical systems for the SLS missions are designed
or built. The SLS project is designed to be more than a paper study. In a few years,
when the design is finished, and much of the hardware has been completed, an
alumni campaign will take place to raise completion and launch costs.
During the 2002-2003 AY, two SLS classes of note took place: (1) The first course
was a graduate robotics course which explored rover architectures for a Lunar
mission. The results from this class were the rover mission scenario described
above and a detailed preliminary design, including mechanical form and fit models,
of each rover. (2) The Spring class was a ME senior capstone class where the
students designed and built a crash landing capsule to deliver the rovers through
the last 50-100m of free fall to the Lunar surface. A full size mockup was built
and dropped from a crane onto a concrete surface from 18m in height (yielding an
impact speed similar to a 100m drop on the Moon).
The Sooner Lunar Schooner:
The Sooner Lunar Schooner is a multi-disciplinary ongoing project at the University of Oklahoma
to plan, design, prototype, cost and (when funds become available) build/contract and fly a robotic
mission to the Moon. The core of the SLS mission will be two robot rovers that will traverse the
Moon’s surface and observe its geology, as well as locating and studying artifacts of previous Lunar
missions, including Apollo 17 and Luna 21 (Lunakhod 2). These rovers will be delivered to the
surface by an innovative landing capsule designed to absorb the impact of a rough landing, rather
than relying on a precision-guided powered descent. Upon arriving at the Moon, the spacecraft will
fire a retrorocket, slowing its velocity to nearly zero approximately 100 meters above the surface.
The capsule will then free-fall the rest of the way. After impact, the capsule will deploy the rovers.
The goal of the flight will be to explore a small section of the Moon; conduct a materials analysis
of the materials left there by an Apollo mission thirty years earlier; and to perform a selenographic
survey of areas that were too distant or considered too dangerous to be done by the Apollo crew.
The goal of the Sooner Lunar Schooner Project is to improve the science and engineering educations of the hundreds of undergraduate and graduate students working on the project [6]. The
participants, while primarily from engineering and physics, will also include representatives from
business, art, journalism, law and education. This project ties together numerous existing research
programs at the University of Oklahoma, and provides a framework for the creation of many new
research proposals.
The Sooner Lunar Schooner grew out of one of the author’s (Miller) earlier experiences. During
the Summer of 1999, the Summer session of the International Space University (ISU) did a design
project on extra-planetary human exploration. One of the key precursor missions proposed in this
study was to have a Lunar rover race [3]. The race was a refinement of ESAs EuroMoon 2000
project [7] which proposed having several rovers circumnavigate the Aitken basin at the Lunar
South pole. These were the latest in a series of proposals to do a variety of robotic missions
on the Moon in the wake of the popularity of the Mars Pathfinder mission. Another project with
similar origins was Blastoff! Corporation’s L1: Return to Apollo mission. Blastoff! was a spinoff
company from Idealab. The company’s mission was to do entertainment space missions – real
missions that would pay for themselves through the sales of advertising, media content, action
figures, etc. Blastoff! differed from other companies espousing similar goals (e.g., LunaCorp) in
that it was decided not to advertise until it was well into phase C & D of the mission, and Blastoff!s
first mission was fully funded (we thought). During 2000, a team of about thirty engineers created
a detailed mission design, including several design iterations of spacecraft and rovers. A significant
amount of prototype hardware was also created. Unfortunately, changes in the US stockmarket
caused Blastoff! to cease operations in early 2001, but the lessons learned during the design and
prototype studies have not been lost.
Lessons that were learned during these activities include:
• The staff working on a Lunar project had to have a diverse set of skills, but needed to have a
common focus and some sort of bonding experience.
• It is difficult to make a convincing business model for a private robotic Lunar mission. One of
the very few exceptions to this is Celestis [1] whose business it is to put human cremains in
space.
• There will be community excitement and buzz about you and your mission, even if you don’t
do anything to advance your mission other than say you are going to do it.
Discussions that took place after Blastoff! closed its doors made it seem obvious that a major
university was a much more likely organization to perform a Lunar mission than say an Internet
spinoff. Universities have a highly skilled workforce that cover the technical, public relations and
business sides of a mission. Additionally, all of the staff is paid below industry average and much
of the staff (e.g., graduate students) are hardly paid at all while undergraduates actually will pay for
the privilege of working on a project. Universities as a whole operate on a completely different kind
of a business model than does any other industry. And universities make a large portion of their
money on buzz and pride – both of which translate into alumni funding.
The Sooner Lunar Schooner concept was presented to the College of Engineering faculty at the
University of Oklahoma. Discussions followed of the various roles that existing labs, capstones and
faculty could fill. A number of courses were set up or reorganized to take advantage of the SLS
domain [6]. The remainder of this paper will review the results of two of those classes.
Space Robotics & the SLS Mission:
Part of the graduate sequence in the robotics curriculum at the University of Oklahoma is a course
on space robotics. In this course we cover the unique aspects of robot mechanics, electronics,
software and design that are necessitated by the space environment. In the most recent iteration of
the course, we also explored the SLS mission and tried to flesh out the robot portion of the mission,
and the robot architecture necessary to achieve it.
Figure 1: Preliminary Design of Science Rover
The initial SLS mission was to land at one of the Apollo landing sites and then to inspect the
lower stage of the LEM and perform an analysis of the aging the materials of the LEM had experienced. Preliminary designs for the robot to carry out the inspection were created (see Figure 1)
and the scenario and early design were then presented at [5].
At the World Space Congress it was pointed out that after Apollo 12 and the return to Earth of
the Surveyor scoop, that NASA decided to document the LEM sufficiently so that a detailed aging
analysis could be performed if the LEM was visited at a later date [4]. This was inspired by the lack
of full detailed documentation of the Surveyor scoop immediately before launch, leaving uncertain
exactly what surface features were caused during production and preparation, and which were
caused by exposure to the Lunar environment. Unfortunately, because of the production schedule,
the only LEM that was formally documented in this way was the one flown on Apollo 17.
This revelation was relayed to the class and it was decided to select the Apollo 17 landing site
as the primary site for the SLS mission. This selection had the additional advantage that Lunakhod
2 is less than 150km almost due North of this site. The class decided that a secondary sprint rover
(see Figure 2) should be created to try and reach the Lunakhod during the mission. Further details
about the mission design and details of the rover are contained in [6] and [8].
Design of a Lunar Landing Capsule
The mission architecture chosen for SLS uses a simple but proven Lunar landing system modeled
after that designed by Ford Aeronutronic and used by the Block II Ranger missions [2]. The original
Figure 2: Preliminary Design of Sprint Rover
Ranger system used a landing radar to fire a solid motor that would kill most of the lander’s velocity
at an altitude of about 100m above the Lunar surface. A sphere of balsa wood was used to encase
the payload and protect it from impact with the Lunar surface during that final fall of 100m or so.
During the Spring of 2003 a team of five mechanical engineering seniors developed an updated
version of the landing capsule for possible use in the Sooner Lunar Schooner. The capsule was
designed to handle a one cubic meter payload massing 100kg (the projected size of the rovers
designed a semester earlier). The landing system could be no more that two meters in diameter
and two meters tall, in order to be able to fit in the launch shroud. The payload could not be
subjected to more than 100 g’s when falling from a dead stop 100m above the Lunar surface. The
Lunar surface itself is assumed to be solid rock at the point of impact. For terrestrial test purposes,
a drop height of 16.7m in Earth gravity onto a concrete slab was used to accelerate the capsule to
the same impact velocity it would experience from a 100m drop on the Moon. Accelerometers in
the payload compartment were used to measure the payloads deceleration.
During the initial research and planning phase, the team investigated previous attempts at delivering a science payload to a planetary surface. Specific strategies examined included a powereddescent, soft landing similar to Surveyor, an airbag system such as that used by Mars Pathfinder,
or an impact-reducing capsule such as the Ranger 3-5 or the Soviet Luna 9, which returned the first
pictures from the lunar surface. All the ideas fell into one of three preliminary conceptsa powereddescent using a rocket motor, an airbag system, or a capsule made of energy absorbing material.
A rocket-powered descent was deemed too complex with its issues of guidance and control, and
issues of cost and safety associated with testing a prototype. The Pathfinder-style airbag system
was quite attractive based on its earlier high-profile success. However, an adequately strong and
durable airbag material such as Kevlar or Vectran turned out to be prohibitively expensive. Furthermore, an airbag system would need to be tested in a vacuum chamber to replicate its behavior
on the Lunar surface. As a final problem, the airbag systems bounce. The team was unable to
determine any advantage to bouncing across the Lunar surface, and many disadvantages, so this
strategy was abanadoned.
An energy-absorbing capsule presented several advantages. Its design and construction would
be fairly straightforward. It presented few difficulties in testing – a simple drop test would suffice.
Several viable materials were available at feasible prices, and these materials would not require any
special handling. Finally, if the material was chosen correctly the capsule would “stick” its landing,
limiting the payload to a single shock. This concept was selected by the team.
The landing capsule was based around a 1 meter tall octagonal frame with a 1 meter diameter.
In the flight version, this frame would spilt lengthwise after landing, freeing the rovers to climb out
and start their mission. For the tests purposes, the frame was made in part from steel box tube
to give a rigid fixture for the accelerometers. Additional mass was added to the system to bring
capsule mass to 100kg – the target mass of the rovers.
Research prior to the selection of the external cushion material revealed that viscoelastic polymer foams had quite complicated behavior under dynamic loading, strongly dependent on temperature and strain rate. Furthermore, many polymer foams dynamic properties come from the action
of gases escaping from the cellssomething that would not be possible on the surface of the Moon.
Compared to polymer foams, honeycomb materials had several advantages. They tend to be quite
strong and lightweight, and since their cell sizes are much larger, their properties are not dependent on gas dynamics or an atmosphere. Honeycomb materials dynamic properties are relatively
insensitive to strain rate, so their behavior is consistent across a range of impact energies. Since
honeycombs exhibit “collapsing column” failure modes, they have nearly flat stress-strain curves,
simplifying design calculations.
Preliminary design calculations were then performed in order to determine what range of properties would be feasible. After the necessary properties were specified, various manufacturers of
honeycomb materials were contacted. Honeycombs can be made of a wide variety of materials,
from aluminum to resin-impregnated carbon fiber membranes and even paper. Interestingly, many
of the paper honeycombs had comparable properties with more costly materials. The final selection, made based on properties, price, and availability, was a 33-wt paper honeycomb with 1/2 inch
cells, manufactured by Hexacomb, and supplied by Innovative Enterprises Inc. The material was
available as blocks 4 feet by 5 feet and 4 inches thick. Knowing the geometry of the capsule and
the properties of the material, it was then possible to make a detailed design for the cushion. This
was performed by equating the energy absorbed by the cushion with the kinetic energy of the falling
capsule, and solving for the necessary cushion area in various orientations. After calculating the
needed volume of honeycomb material, its mass was computed and added to the capsules mass in
the energy equation. The equation was thus solved iteratively until the required volume converged.
This calculated volume was then used to specify the amount of material to be purchased.
Knowing the shape and size of the cushion, the next step was to design the individual pieces
of material that would be assembled into the cushion. Since the material would be laminated five
layers thick, and had to accommodate the octagonal geometry of the capsule, this produced 11
unique pieces, of which as many as 32 needed to be made. Once a drawing had been produced
for each piece, diagrams were made that detailed how the pieces were to be cut out of the blocks of
honeycomb, in order to ensure that all the necessary pieces could come from the material on hand.
The blocks were then marked and cut, and the components were assembled into large segments,
which were then bonded to the capsule.
Concurrent with the design and fabrication of the honeycomb cushion, the instrument package
was designed and assembled. The instrument package had to measure acceleration along all
three axes, with a range of at least +/-100g’s and enough sensitivity to distinguish one or two
g’s. We selected Motorola 1200D single-axis accelerometers. We further decided to go with a
double-redundant system, so we would ensure that the deceleration would be measured by at least
one sensor in each direction. The six accelerometer sensors were assembled according to the
manufacturers recommendations, with a regulated 5V power supply and RC conditioning on the
output signal. To record the deceleration data, we selected an Onset Tattle-Tale data logger, which
had a sufficient number of input channels, a capable analog to digital converter, a relatively high
sample rate, and a great deal of data storage capacity.
Figure 3: Drop Capsule Falling from a Great Height
In order to test the prototype, we calculated the scale height on earth that would produce the
same impact energy as the 100 meter free fall on the moon. We arranged to use a basket lift
provided by the OU physical plant to hoist the capsule to the test height of 17m (see Figure 3). In
order to ensure the safety and feasibility of this system, we dropped a concrete block having similar
mass to our prototype capsule. Having verified the drop test method, we arranged for a drop time
and location. Since the test would be quite a spectacle, the public was invited and advertised the
drop’s time and location.
The capsule was hoisted and dropped flawlessly, and the cushion performed as expected, deforming significantly in the impacted areas. Upon impact, the adhesive bonding the cushion material to the metal capsule failed, and the material came off. This was a desirable feature – lowering
the deceleration shock and making exit from the capsule area easier for the rovers. The cushion
material between the capsule and the ground of course was trapped in place. The honeycomb
material exhibited no elastic properties and the capsule had absolutely no visible bounce, so the
shed material was not needed.
The landing shock was estimated by measuring the deformation of the cushion material. According to our measurement, the material was deformed by a maximum of 9 inches, corresponding
to 73gs. This was confirmed by data from the accelerometers (see Figure 4). Additional drops,
using cushion material only on the bottom of the capsule yielded similar results. This shock is
within the survival specifications for standard hard drives, and consequently most electronics and
mechanical systems.
A significant mass savings can be achieved by stabilizing the capsule so only one end needs
full padding, the rest of the capsule can have lower levels of padding should the capsule capsize
after the initial impact, or roll from landing on an incline or rock. Additional design work is need to
determine the best way to stabilize the capsule for landing (e.g., an RCS, or spinning), should this
approach be used.
Conclusions
The Sooner Lunar Schooner is an ongoing multi-disciplinary design build test and eventually fly
Figure 4: Accelerometer Data from Drop Test
engineering project at the University of Oklahoma. During the 2002-2003 academic year two significant milestones were accomplished. 1) A mission architecture was chosen along with the high level
design of the payload. 2) An important element in the architecture, a crushable landing capsule,
was prototyped and tested successfully.
The latter milestone demonstrated that inexpensive and simple solutions, using readily available
materials can be substituted for very high tech, and consequently highly expensive methods that
are more commonly used. Much of the SLS mission will follow this philosophy, and so we hope to
keep the entire mission costs within the reach of alumni funding.
REFERENCES
[1] Celestis home page. http://www.celestis.com/, 2004.
[2] R. Cargill Hall. Lunar Impact. Number SP-4210 in NASA History Series. National Aeronautics
& Space Administration, Washington D.C., http://history.nasa.gov/SP-4210/pages/
Cover.htm edition, 1977.
[3] ISU-SSP. Out of the cradle: An international strategy for human exploration away from earth.
http://neptune.spaceports.com/%7Ehelmut/exploration99/main.html, 1999.
[4] Wendell Mendell. LEM surface documentation. Personal communication, 2002.
[5] D.P. Miller, D. Hougen, and D. Shirley. The sooner lunar schooner: Lunar engineering education.
In Proceedings of the 34th COSPAR Scientific Assembly - The Second World Space Congress,
2002.
[6] D.P. Miller, D. Hougen, and D. Shirley. The sooner lunar schooner: Lunar engineering education.
Journal of Advances in Space Research, 31(11):2449–2454, 2003.
[7] Wubbo J. Ockels. Euromoon 2000 - a plan for a european lunar south pole expedition. http:
//esapub.esrin.esa.it/br/br122.htm, 1996.
[8] M.J. Roman, T.S. Hunt, J. Yoon, and D.P. Miller. The sooner lunar schooner mission. In Abstracts
from 34th Lunar & Planetary Science Conference, page #2126. Lunar Planetary Institute, March
2003.
8