PellegrinoHonorsThesis

Panel Processing of DebriSat
Marielle Pellegrino
University of Florida, Gainesville, FL, 32612
DebriSat is a representiaive low earth orbit satellite that was developed to provide data for
improving existing NASA/DoD satellite breakup models. It was subjected to a ground-based
hypervelocity impact test at Arnold Engineering Development Center where the satellite was
catastrophically destroyed generating numerous debris fragments. To minimize damage (i.e., size
and shape) to the fragments during the test, the walls of the test chamber were covered by layers of
foam panels forming a soft catch system. This paper discusses how these panels are processed at the
University of Florida using “archaelogical” techniques. Specifically, the paper addresses some of the
challenges associated with maintaining the integrity of the fragment pieces, tagging/cateloging the
debris fragment location within the panel (and the chamber). Other challenges discussed inlcude
using systems engineering techniques to develop training modules for new team members and
organizing/coordinating the efforts of large groups of team members to work efficiently and
harmoniously.
Nomenclature
AEDC = Arnold Engineering Development Complex
ACDS = Attitude Determination and Control System
CDH = Command and Data Handling
EPS = Electrical Power System
LEO = Low Earth Orbit
PRC = People’s Republic of China d
SOCIT = Satellite Orbital debris Characterization Impact Test
TSC = Thermal Conrol System
TTC = Telemetry Tracking and Command
I. Introduction
he presence of objects orbiting Earth has drastically increased since the launch of Sputnik. As shown
in Figure 1, the amount of objects has rapidly increased over time, and does not appear to be slowing
down. Using current technology, only objects of size 10 cm or greater can be tracked. Although space
exploration provides many benefits including communication satellites, enviornmental monitoring, and
space telescopes, of the approximately 15,000
trackable objects in low Earth orbits (LEO), only
6% are functional. The remaining objects
consist of retired satellites, spent rocket bodies,
and most notably fragments from on-orbit
collisions. The increasing amount of objects in
orbit is alarming because an increase in objects
in orbit causes an increase in chances of future
collisions. A collision with a fragment only
milimeters in size can cause significant damage
to operational satellites due to the high velocities
of the objects in orbit. The ISS sheilding, for
instance, can only protect against objects of 1.4
cm or smaller. This number shows that there is a
large gap between what objects can be tracked,
10 cm, and what objects can cause significant damage to the ISS, 1.4 cm.
T
Figure 1 Photo by NASA, only objects in the U.S. Space
Surveillance Network (SSN) catalog are shown. Sizes of dots
are not to scale.

Currently four accidental collisions have occurred, and in 2007 People’s Republic of China (PRC)
launched an anti-satellite assualt test. The PRC test nearly doubled the amount of objects [2]. The collision
of Iridium 33 and Cosmos 2251 in 2009 is particularly important. Post-collision analysis using the area to
mass ratio based on current breakup models accurately predicted the fragments generated by the Cosmos
spacecreaft as seen in Figure 2. However, the model did not accurately predict the fragments generated
from Iridium 33, as seen in Figure 3. The current breakup model is based on the SOCIT test [3], which was
based on the breakup of a 1960’s Transit satellite. This satellite is representative of the older Cosmos 2251,
hence the good correlation between prediction and measurement; however, Iridium 33 is a modern satellite
and thus is comprised of materials typically not found in a 1960’s satellite (e.g., multi-layer insulation and
composites), thus thepoor correlation. This discord in the expected data versus the actual data points to the
need for updated breakup models. To update the breakup models with dta from modern satellites, DebriSat
project was conceived. DebriSat is a 50 kg satellite designed to be representative of satellites found in low
earth orbits (LEO) and subjected to a hypervelocity impact test.

Figure
32
Breakup model vs Actual Breakup of Iridium 33

II. Hypervelocity Impact of DebriSat
To be representative of satellites typically found in LEO, DebriSat contains all of the subsystems
in the satellites, and at times redundancies to encompass the wide variety of modern components used in a
wide variety of LEO satellites. DebriSat was subjected to the same rigorous design and fabrication
processes typical of the space industry. satellite but not electrically functional. Due to cost constraints, it
was cost prohibitive to utilize flight hardware so efforts were made to acquire flight spares and surpluses
which resulted in the donations of some ADCS components (IMU, a reaction wheel, an torque rod cores).
All other subsytems/components used in DerbiSat were fabricated to emulate existing components in
geometrical, structural, and material characteristics. Since the primary concern of DebriSat project is
characterizing post-impact breakup, high importance was placed on ensuring the the type and quantity of
materials used were consistent with modern LEO satellites. For this reason, it was not essential for the
satellite to be electrically operational. DebriSat went through the same environmental tests (vibrations,
thermal bakeout) that satellites must go through before flight, in order to ensure its fabrication and
assembly process.
The subsystems included in
DebriSat were the Attitude Determination
and Control System (ADCS), Eletrical
Power System (EPS), Payload,
Propulsion, Structure, Command and Data
Handling (CDH), Telemetry Tracking and
Command (TTC), and Thermal Control
System (TCS). A diagram of the
subsystems on DebriSat can be seen in
Figure 4. ADCS included an inertial
Figure 2 Breakup model vs Actual Breakup of Cosmos
2251
Figure 4 Diagram of Some of the Subsystems Used in DebriSat

measurement unit, two star trackers, four sunsensors, four reaction wheels, and three magnetorquers.
Although components like star trackers are not typical of a 50 kg satellite, the goal of DebriSat is to create a
breakup model of a wide variety of components in LEO. The EPS included two eight cell battery boxes, a
power conditioning and distribution system,. The core of an individual battery cell was fabricated to
emulate a Li- Ion battery, sans the electrolyte (for safety purposes). The payload consisted of two
spectrometers, an optical imager, and an avionics box to support the payload. The propulsion system
comprised of a composite overwrapped pressure vessel and six emulated thrusters, along with associated
plumbing and avionics. The TCS was comprised of an emulated capillary pumped loop design for heat
pipes and multi-layer insulation, as well as a panel used as a radiator. The TTC consisted of an X-band, S-
band, and UHF/VHF antennas, associated electronics, and shielded cabling. The structure comprised of
aluminum components (two hexagonal plates, six longerons, adapter ring), honeycomb aluminum sandwich
composite panels (twelve panels, three solar panels). The CDH was made up of an emulated flight
computer and an emulated memory unit. All components were cleansed in an ultrasonic bath, air dried, and
spot cleaned with isopropyl wipes and were assembled in a Class 10,000 clean room. The components were
handled in this way to ensure the authenticity of the material being tested.
The hypervelocity test was conducted at the Arnold Engineering Development Complex (AEDC)
using Range G, their two stage gas launcher. A hollow 600 g aluminum projectile designed by the AEDC
to be representative of a large fragment or a small satellite in LEO. The goal was to launch the projectile at
a speed of 7.0 km/s which was to represent the speeds characteristic of objects in LEO. The projectile,
which used a nylon wrapped sabot design reached a maximum velocity of 6.8 km/s at AEDC (highest
recorded in the US). AEDC uses a sabot design for its projectiles to (i) minimize the friction between
projectile and the walls of the cannon, and (ii) to ensure that the entirety of blast was pushing the projectile
(i.e., there was no leakage past the projectile). The resulting collision was completely catastrophic,
releasing 13.2 MJ of energy.
To properly document the collision, nurmerous instrumentation was provided by AEDC, NASA,
and the Aerospace Corporation. AEDC provided Flash X-ray Systems which comprised of dual heads
mounted every twenty feet and two single head systems. AEDC also provided six digital cameras and two
laser camera systems with YAG laser used to illuminate the test object. NASA provided acoustic/vibration
sensors, high-speed color video, Agilent Exoscan Portable FITR, Nikon 5200 DSLR camer, and Analytical
Spectral Device field spectrometer. The Aerospace Corporation provided AHMI Infrared Hypertemporal
Imager, AERHy Infrared Hyperspectral Imager, Portable Mass Spectrometer, Gas Sampling Bottle,
Witness Plate Assemblies, Borescope, High-speed (ns) gated ICCD cameras, 3 m optical fiber, two camera
assemblies for ICCD cameras, Photron camera for the Borescope, and UV-Visible spectrometer. All of
these instrumentation were essential in recording the collision of DebriSat. It was not only important to
record all sections of the light spectrum, as seen with using Infrared to X-Ray cameras, but also cameras at
varying speeds and resolutions. Witness Plate Assemblies and Gas Sampling Bottles were to collect
condensable species and the observed brown smoke. The acoustic/vibration sensors were to pick up the
vibrations and shock waves emitted during the collision. It is absolutely essential for all information to be
documented for later use. Lack of documentation in the SOCIT test, led to difficulties in understanding
how to conduct and improve a newer test like DebriSat.
However, the information that existed from the
SOCIT helped improve some of the key aspects of
DebriSat. For example, in the SOCIT test only ten percent
of the information was collected which led to arbitrary
data scaling. In DebriSat the goal is to collect all of the
items from the test chamber and catalog ninety percent of
it for an accurate breakup model. Another example is that
the SOCIT test showed that the fragments changed shape
and size after colliding with the walls of the test chamber.
To prevent materials from distorting in shape and size
upon imact with the wall, a soft catch system was used.
This soft catch system consisted of increasing density
foam panels lined up against the walls, as seen in Figure 5.
These foam panels were designed to prevent the projected
materials from reaching the wall by intercepting them
before they made an impact with the wall. After the hypervelocity test, all contents of the test chamber
Figure 5 Test Chamber at AEDC With Soft Catch

were collected, bundled, boxed and sent to the University of Florida for post-processing. The process
utilized emulated the techniques used in archeological extrications; i.e, the items were baged and tagged to
designate their location within the test chamber at the time of retrival.The main focus of the test was not to
distort the shape and size of the fragments and also to keep the information where the fragment were
located within the test chamber. These concerns continued to be a focus in the panel processing portion of
the DebriSat project.
III. Foundation of Panel Processing Procedure
Before handling any of the panels,
nitrile gloves are used for protection
from contamination. This contamination
is important because optical imagers
used in the characterization portion of
DebriSat can pick up the oil on students
hands. Each box has a bundle in it
belonging to the same section on the
testing chamber. The bundle was
wrapped in plastic sheet to ensure no
loss of debris during transportation. The
foam panels are placed on a table with
the positive z axis facing up. A photo is
taken of the Panel ID on the side of the panel; the following pictures are all in reference to that Panel
ID. The grid is then placed over the panel. Then the foam panel is entered into the database given its Panel
ID, the database then assigns it a Foam ID number and creates a panel entry. At this stage, the color of the
foam panel and the box number in which the foam panel was transported from AEDC are entered into the
database, as seen in Figure 6. The foam panel is marked with the origin, the grid referencing lettering
system (1A), the Foam ID, and that the face lies in the positive z direction (+Z). A photo is then taken of
the entire grid with the foam panel inside to document the grid, as seen in Figure 7. This photo is later
uploaded to the database under its Foam ID. The panel is
now ready to process.
These steps are necessary to provide
full traceability of the debris.
Using tweezers of varying sizes, each loose piece of
debris greater than 2 mm is placed in its own bag. The piece
of debris is then uploaded to the database. The database
generates a Debris ID and the student fills out whether it is
loose or embedded. Because it is loose, the student enters the
fields where from the chamber the bundle is from rather than
the panel it is associated with. The section, row, and area fields are used to define where the bundle is. The
student writes the Debris ID on the bag with the debris inside
and places it in a plastic box. Each piece of embedded debris
is also placed in its own bag. Embedded debris is debris in craters or is embedded in the foam. Any pieces
that are secure in the foam are left for X-Ray imaging and extraction. The piece of embedded debris is
uploaded to the database and receives a
Debris ID from the database. The student
then enters the piece of debris as
embedded which leave the Foam ID and
grid number fields for the student to enter,
as seen in Figure 8. The student writes
the Debris ID on the bag with the debris
inside and places it in a plastic box with
the other bags of debris. After looking
through this face the panel is flipped and
the process is repeated keeping the same
origin and flipping the grid in the same
Figure 6 User View of Database Foam Entry
Figure 7 +Z Face Photograph of Foam Panel
Figure 8 User View of Fragment Entry

fashion as the panel was flipped. All debris is removed the same way as before. The panel is then place in a
bag and sent to the X-Ray facility.
Dust is generated by the mechanics of removing the debris. This dust is swept to the side after the
positive Z face is finished and after the negative Z face is finished. After the panel is removed the dust is
looked at once more for any loose pieces of debris and when there are no more left it is swept into a large
bag. This bag is labeled with the Foam ID, Panel ID, the color, “Dust”, and the box number the foam panel
is from. The bag is then placed with the other dust bags in a cardboard box.
IV. Challenges Met During Panel Processing
A. At the beginning of the panel processing stage of DebriSat, fourteen new students were hired to the
group to assist with the excess demands in amount of fragments being collected. The increase in student led
to an increase of occupancy during hourly shifts. Up to ten people would arrive for one shift; however, only
four people are able to work on one panel at once. This restriction is because there is limited space on the
actual panel for people to work. To work efficiently and keep to the schedule the new students were hired
for it is imperative that every student at the shift is processing.
B. Another issue that arose is when
panels have highly concentrated craters in
one location of the panel, as seen in Figure
9. After the three other students working on
their sections have finished, they would
have to wait on the final person to finish
their section of the panel before moving on
in the processing process. As said in the
previous paragraph, it is imperative that the
students’ time is used efficiently to keep
with the preplanned schedule. Having
students waiting and not working while
another student is finishing up a section is
not an efficient use of the group’s time.
C. Many of the new students came from different educational backgrounds and years spent in college.
Therefore, the students did not all have the same knowledge of the materials being worked with during
processing. Although at this stage the debris is not characterized, to efficiently use time, students should not
be bagging items that are not from DebriSat. Burnt foam was easy for all the students to distinguish;
however, burnt wood was hard to distinguish from actual fragments for the students who had not had
experience with the materials in a lab. The inability to distinguish between fragments and the accidental
bagging of material that did not need to be characterized or measured resulted in efficiency.
D. When bagging the debris, many times students would have the fragments get caught in the ribs of the
bag. This problem was the result of static from the bag pushing the fragment upwards, while the student
freely dropped the fragment in the bag. Getting the debris caught in the ribs of the bag poses two major
concerns. The first concern is that the process of putting a fragment in a bag took close to three minutes
when a student had to retrieve the fragment from the ribs of the bag, unlike the usual amount of time which
was approximately ten seconds. The time difference was a major inefficiency of the panel processing
program. The second concern was that the process of removing the fragment from the ribs was a very
aggressive one. Rarely did fragments survive whole after
retrieving them from the ribs. This is a colossal problem
for the project because the major focus is characterizing
the size and shape of all of the fragments and that
information is lost once the fragment is broken.
E. From the heat of the collision, the material that
comprised of the band that held the foam bundles together
would get so hot that it would break and melt the foam of
multiple panels together, as seen in Figure 10. Because
each panel is processed separately this posed a problem
for panel processing. The separation process had to be
documented in such a way that it can be repeated in a later
Figure 9 DebriSat Foam With a Concentrated Area of Debris
Figure 10 Image of Two Panels Sealed together by
Melted Bundling String and Foam

part of the panel processing stage. Repeatability is a major concern when conducting research, especially
with a group of 20 students total working.
V. Panel Processing Improvements
A. To solve the issue of having too many people working at the facility at once, I was able to create a
systematic way for students to process two panels at once. Although definining the panel with a Foam ID is
a sequential process, I created a way for two groups to work at once, through using a white board to display
to both table what Foam ID each group is working on. With two tables running, eight people were now able
to work at the processing facility. For the remaining two students I had two different solutions depending
on the day of work. The first solution was if there were plastic wrapping from around the panel bundles to
process. I created a process for retrieving the debris from the plastic wrapping by angling the wrapping at
various positions and using paint brushes to guide the debris towards a hole. If there was no plastic
wrapping to process, I separated the panel processing steps amongst students into extracting and labeling.
Because the four person per panel restriction was due to the size of the panel, students could work on the
remaining portions of the table if they were handed debris in a bag to label and seal it. By separating these
processes an additional student could be added to each group processing panels.
B. For times when debris was located in one area of the panel, I developed a process for all the
students to work on one area. After all of the students had cleared their respective sections of debris, they
would remove some of the debris from the concentrated area with a paint brush and guide it towards their
respective section. To prevent any confusion with a later student thinking the debris was loose after it had
been taken out of the crater, a permanent marker is used to mark around the debris along with the original
location grid number of the debris.
C. To aid the underclassman and majors outside of the mechanical and aerospace degree in
distinguishing types of material, I developed a technique for students to test if the material was wood or an
actual fragment from DebriSat. This technique was done by applying a small amount of pressure to
different materials including carbon fiber and aluminum, then showing the students how each material
would respond. Wood would have more
give than the harder materials but less give
than materials like rubber. After moving
through the demonstration the students were
able to use this technique to correctly
identify DebriSat fragments.
D. To avoid the disfigurement of the
fragments and the time costs involved with
removing debris from the ribs of the bag, I
suggested the use of anti-static bags. This
would allow the debris to drop freely into
the bag without getting stuck in the ribs.
This change improved efficiency of panel
processing 20 percent.
E. To document the separation of panels, I used a process of one student taking photographs while the
other student uses his or her hands to provide leverage between the two panels, as seen in Figure 11. The
use of photographs to document the process is not only important for repeatablity, but it also important for
documenting any type of breakage of fragements that may occur during the separation process.
VI. Conclusion
As of April 4, 2015 the DebriSat project has procesed 234 panels and over 55,000 fragments of debris.
This number covers almost half of the 500 panels received from AEDC. With my aid, the project was able
to stay on schedule of completion of panel processing for the next stage of the project. My processing
improvements helped maintain that everyone on the project was working efficiently, thus allowing the team
to move on to the characterization and measurment of the fragments stage soon. The completion of the
entirety of the DerbiSat project is now on schdeule for the original planned year of 2017.
Figure 11 Documentation of the Separation of Sealed Panels

Acknowledgments
The authors would like to thank the NASA Orbital Debris Programs Office for their contributions, the
subject matter experts at Aerospace Corporation and the Air Force Space and Missile Systems Center for
their design input, and several members of the University of Florida Space Systems Group most notably
Dr. Norman Fitz-Coy.
References
Periodicals
Sources:
1
M. M. Castronuovo, "Active space debris removal-A preliminary mission analysis and design," Acta
Astrnonautica, vol. 69, no. 9-10, pp. 848-859, November-December 2011.
2
J.-C. Liou, “The Orbital Debris Problem and the Challenges for Environment Remediation,” Seminar Powerpoint
Slides 22, 29, 40, 41, September 2014.
3
J.-C. Liou, "The Man-Made Orbital Debris Problem and a New Satellite Impact Experiment to Characterize the
Orbital Debris Properties," Houston, 2011.
Reports, Theses, and Individual Papers
1
M. Werremeyer, “Design and Fabrication of Debrisat – A representative LEO satellite for Improvements to
standard satellite breakup models”, 63rd
International Astronautical Congress, IAC-12-A6,3,7x16098
2
M. Rivero, I. Edhlund, “Hypervelocity Impact Testing of Debrisat to Improve Satellite Breakup Modeling” 65th
International Astronautical Congress, IAC-14-A6.2.10x25834
3
M. Wilson, R.Dikova, “AEDC Ranges Test Plan – DEBRISAT HYPERVELOCITY IMPACT TEST”, March
2014

PellegrinoHonorsThesis

More Related Content

What's hot (20)

Viewers also liked (19)

Similar to PellegrinoHonorsThesis (20)

PellegrinoHonorsThesis