Cowboy Rocketworks at 2018 Spaceport America Cup

Team 54 Project Technical Report

Oklahoma State Univeristy AIAA Rocketry Team1

Oklahoma State University, Stillwater, OK 74078

Oklahoma State University’s AIAA Rocketry Team, Cowboy Rocketworks, is competing

in the 2018 Spaceport America Cup in the 10k – COTS – All Propulsion Types category. The
launch vehicle, Results May Vary, will fly to 10,000 feet AGL with an airbrake unit and camera
payload that will be compiled into 360° video after flight. The airbrake is an improved design
from the airbrake designed in 2017 with the ascent phase of the rocket monitored and
controlled by the autonomous system the Controls Team developed. This system will utilize
an array of sensors to provide acceleration, velocity, and altitude data to an Arduino which
will actuate the fins accordingly to reach the 10,000ft goal. In order to do so, the projected
altitude will be continuously calculated using equations derived from Newton’s Second Law.
The payload contains five cameras connected to an Arduino Nano that autonomously begin
recording video prior to launch and stops recording after touchdown. These two subsystems
are flying on a rocket made entirely of SRAD fiberglass airframes and CNC-cut fins and
centering rings. This rocket is entering the competition having flown twice before and bolsters
confidence in the rocket’s design and hardware.

Nomenclature
COTS = commercial off-the-shelf
d = cylinder diameter
l = length
m = mass
PLA = polylactic acid
SRAD = student-researched and developed
UHMW = ultra-high wolecular weight polyethylene

I. Introduction

T he Oklahoma State University AIAA Rocketry Team was founded in August 2016 with the intention of competing
in the Spaceport America Cup and revitalizing the interest in rocketry at the university. The Spaceport America
Cup served as the driving factor to certify the team’s members with the Tripoli Rocketry Association, incrementally
build larger and more complex rockets, and secure the support from the university and its faculty.
Following the entry in the 2017 Spaceport America Cup, and with the generous support of donors and sponsors,
the team was able to design a larger rocket for the 2018 Spaceport America Cup. Upgrading from a 4” diameter to a
6” diameter airframe, this allowed for a better designed payload and improved airbrake with fewer constraints as
dictated from a 4” diameter rocket. Multiple team members achieved Level 3 Certifications which helped the number
of test flights accomplished for this project.
OpenRocket was used extensively as the primary modeling software of the rocket and its flight simulations. All
performance metrics as stated in this paper are gathered from OpenRocket unless otherwise stated.

1
Hunter Billen, Jordan Chancellor, Garett Foster, Nicholas Foster, Nicolas George, Cole Henderson, Kyle
Hickman, Samantha Huckabay, Gerald McCullers, Katelyn Powell, Nicholas Rozell, Timothy Runnels, Austin
Stottlemyre, Garrett Townsend, Lucas Utley, Andrew Walsh, & Garrett Wilkens.

1
 The team’s program name is Cowboy Rocketworks, and they are flying a 12.5’-tall rocket named Results May
Vary in the 10k – COTS – All Propulsion Types category.
The Cowboy Rocketworks team was faced with a challenge: How to build a rocket that stops at 10,000 feet exactly?
There are many different ways to approach this problem. Some teams build a rocket and try to get it to weigh exactly
enough to hit the altitude mark. A major problem with this strategy is that it doesn’t account for weather cocking of
the rocket or variances in motor power or simulation accuracy. The next possible option is to build an airbrake that
just simply deploys. This is an option for shaving off altitude and slowing down the rocket, but ultimately suffers the
same shortfalls already discussed. That leaves one other scenario: create a smart altitude control device that knows
when to decelerate the rocket and when to let it fly. Such an idea is easier said than done, and creating a device that is
robust and sophisticated enough to handle the task can be quite difficult. To solve this problem, Cowboy Rocketworks
decided to go all in and created a Controls Team for the Spaceport America Cup. This team was tasked with designing
an airbrake system to control the ascent of the rocket intelligently. To do this, a series of prototype designs were
created by the team. They were then evaluated for efficacy by a series of test and further refined until the best design
emerged. From there, the airbrake was manufactured, tested, and programmed to perform optimally. This process is
further explained in the System Architecture Overview.
This year’s team consists of 14 members attending the Cup and 3 others who made valuable contributions to the
team’s progress, but who are unable to join at the competition. Austin Stottlemyre is the team’s Director. Hunter
Billen, Jordan Chancellor, Nicolas George, Cole Henderson, Kyle Hickman, Samantha Huckabay, Katelyn Powell,
Nicholas Rozell, Timothy Runnels, Garrett Townsend, Lucas Utley, Andrew Walsh, and Garrett Wilkens are attending
the Cup. Garett Foster, Nicholas Foster and Gerald McCullers also made significant contributions to ensuring the
team’s success in the competition.
Cowboy Rocketworks’ sponsors include Spirit AeroSystems, OSU Student Government Association, OSU CEAT
Student Council, and 100 friends and family members of the team who donated during our PhilanthroPete fundraiser
with the OSU Foundation in fall 2017.

II. System Architecture Overview

Cowboy Rocketworks’ launch vehicle, Results May Vary, consists of several subsystems which include Propulsion,
Aero-Structures, Recovery, Payload, and Airbrake. The airbrake is a significant technological development that will
ensure Results May Vary successfully reaches a precise target apogee of 10,000 feet AGL.
It is flying on an Aerotech M1939W motor, weighs 71 pounds on the pad, is 6” in diameter, and has a height of
12.5’. The minimum static margin is 2.62 as determined by OpenRocket.

Figure 1. Results May Vary Cutaway View. Graphic illustrating internal components in relation to exterior
airframe and structural components.

2
A. Propulsion Subsystem
The propulsion system for Results May Vary is an Aerotech M1939W composite solid-propellant rocket motor.
The performance of this motor as specified by the manufacturer is shown in the figure below.

Figure 2. Aerotech M1939 Performance Specs

This motor is in the form of multiple propellant grains and will use a reloadable Aerotech 98/10240 motor casing
with a plugged forward closure. In total, the propulsion system will contribute approximately 19.82 pounds (8988 g)
to the overall launch vehicle weight.
In order to maintain integrity and reliability with the propulsion system, no modifications to the motor will be
made. Furthermore, the given manufacturer instructions will be closely followed for motor assembly.

Figure 3. Aerotech M1939 Thrust Curve. Thrustcurve.com

B. Aero-Structures Subsystem
1) Fiberglass Tubes
This year, Cowboy Rocketworks began making SRAD rocket body tubes and sheets. The tube-making process
involved wrapping fiberglass weave around a mandrel. For convenience and ease, the casting mandrel was an extra-
long phenolic coupler section purchased from Public Missiles. This allowed for the purchase of commercial coupler
sections for couplers and electronic bays. In the future as the manufacturing processes become better and tolerances
become tighter, these couplers will be SRAD parts as well. The process that was used for Results May Vary is described
below.
First the fiberglass cloth was cut to size to correspond to the width and number of wraps needed to achieve certain
dimensions. 6 wraps were used, and this proved sufficient in strength as well as weight. The mandrel is prepared by
cleaning it with mineral spirits and adding a layer of wax paper wrapped around the mandrel. Special care was taken
to ensure there are no air bubbles, and the paper has the non-stick surface pointing outward. A generous layer of
petroleum jelly is distributed along the length for proper lubrication. Another layer of wax paper was added to the

3
outside of the mandrel with the non-stick surface facing inwards. Finally, the second wrap of wax paper is sprayed
with a non-stick cooking spray. This ensures the paper will easily come off the inside of the tube.
Fiberglass resin is prepared simultaneously by another team member. A tube of 6” outer diameter with 6 wraps
required 30 fluid oz of resin and a corresponding 300 drops of hardener. The resin is poured in two 12 oz containers
such as a standard plastic party cup, and the hardener premeasured in a separate container, such as an epoxy mixing
cup. When ready, the two constituents are mixed together at room temperature.
To begin wrapping the fiberglass cloth, some of the mixed resin is poured onto the top of the prepared mandrel.
The leading edge of the fiberglass cloth is wrapped onto the mandrel and the resin gently rubbed into the cloth with a
gloved hand until saturated. The mandrel is turned and fiberglass wrapped in slight tension while pouring resin
intermittently between the wraps. A downward sweeping motion keeps air bubbles from forming. This process is
continued until the last of the cloth is wrapped around the mandrel.
To finish the tube, it is set upright to harden. This takes anywhere from 20 minutes to an hour, depending on the
temperature of the room, ventilation conditions, and lighting, as all these factors have been observed to affect the
hardening time. Once the tube has cured, it is pulled off mandrel. The ragged ends are cut off with a table saw, and
the whole tube sanded to ensure a smooth finish.
2) Fiberglass Sheets
Sheets were made using pre-existing commercial fiberglass sheet (G10 Garolite) to press layers together while
curing. Sheets were made with 12 layers of fiberglass cloth for fins (yields 3/16” thick) and 8 layers (yields 1/8” thick)
for centering rings and bulkplates.
The existing fiberglass commercial sheets were covered in wax paper with the shiny side up and sprayed with
cooking oil. Next, the appropriate number of layers was cut out of fiberglass cloth with the layers oversized by about
an inch in each direction. This insured that the entirety of the 12x12 section was usable. Next, 12 oz of resin was
prepared. Then a small amount of resin was spread onto the bottom wax paper covered plate before the first layer was
set on. Resin was added between each layer and smoothed with a plastic scraper. After the last layer, wax paper
covering plate was placed on the layup and weights added to ensure an evenly packed layup.

Figure 4. Stages of preparing a 12”x12” fiberglass sheet

A single 24”x24” fiberglass sheet 3/16” thick was made that became the rocket’s 3 fins and extra set of altimeter
bay bulkplates. The 3 centering rings were cut from a 1/8” thick sheet. These parts were CNC cut with an Inventables
X-Carve machine.

Figure 5. X-Carve CNC cutting fins from SRAD fiberglass.

4
 3) Rocket Assembly
The assembly of the rocket began with sanding the inside of the aft, middle, and forward sections to allow the
coupler sections, motor mount, and nose cone to fit. All epoxied surfaces were also sanded and washed to maximize
surface area for epoxy adhesion. Next, fin slots were cut into the aft section by means of a routing jig designed and
built by OSU students. The jig uses a Dremel tool and allows for perfectly straight slots to be cut in a variety of tube
sizes.
To attach the aft 1515 rail button, a hole was drilled through the aft airframe section between two fins, and a wood
screw placed through the rail button’s hole and screwed into a small square of ¼” plywood on the inside of the aft
section tube. Epoxy clay was spread over the wood block to reinforce the rail button. This process was used for both
rail buttons which were designed and 3D printed from PLA plastic. They are airfoiled to reduce drag.
The 98mm motor mount tube is held in place by 3 fiberglass centering rings that were epoxied into the aft airframe
with G5000 RocketPoxy. For motor retention, holes were tapped along the aft centering ring. An Aero Pack 98mm
flanged retainer was then fastened onto the bottom centering ring and secured with the included screws.
The fins were sanded down to fit inside the slots and were held in place with G5000 RocketPoxy fillets, both
inside the airframe and externally.
The altimeter bay is a 14” coupler with a 2” affixed slip band. The slip band had a 5/8” hole drilled into to it and
a ½” hole in the coupler to countersink the rotary switches used for arming the altimeters. This helps reduce drag.
Results May Vary uses a metal-tipped 5:1 ogive filament-wound fiberglass nosecone from Madcow Rocketry.
The included nose cone coupler and bulkplate are used. Commercial 6” couplers are also used throughout for joining
the various sections.
To ensure that the sections stayed together during take-off, holes were drilled in all the sections attached to the
coupler sections to allow aluminum rivets to be put in place. These rivets are threaded tube fasteners as manufactured
by LumaDyne. They are used in the couplers attaching the aft and middle section, and in the coupler section attaching
the middle to the forward section.
3 nylon shear pins (2-56) hold the forward section to the nosecone and at the bottom the electronics bay (middle
coupler section) to the middle section. The holes for the plastic shear pins had to be tapped so the pins could screw in.
The shear pins are designed to break off when the black powder charges ignite. When the black powder charge ignites,
the rocket will break apart just below the electronics section and at the nose cone for the drogue and main deployment
events, respectively.
5/16” forged eyebolts are used at the nosecone and altimeter bay which have a load capacity of 900 pounds. A
3/8” forged eyebolt is used as the aft-most attachment point on the airbrake (rather than attaching to the 98mm plugged
motor closure). 1/8” vent holes in each compartment release internal air pressure.

Figure 6. Results May Vary following construction.

5
C. Recovery Subsystem
The recovery system is centered around a combination of three altimeters and four ejection charges to guarantee
the deployment of both the drogue and main parachutes. In this system the charges are: The Primary Main, the Primary
Drogue, the Backup Main and the Backup Drogue. The System as a whole is depicted below.

Figure 7. Altimeter wiring schematic.

The altimeters are an Altus Metrum TeleMega, an Altus Metrum EasyMini, and a PerfectFlite StratologgerCF.
All three are attached by screws into the plywood sled of the electronics bay with rubber dampeners to mitigate
vibrations.
The TeleMega is powered by a single cell Lithium Polymer (3.7 volt LiPo) battery capable of providing a minimum
of 800mili-Amp hours (mAh). While the other two altimeters are each individually powered by nine-volt batteries
secured in the electronics bay by zip ties, seated on the sled shelf, and snap connectors secured with electrical tape to
maintain the circuit through the motor boost phase and for all the reactionary forces from the ejection charges.
Also, each altimeter connects to its own igniter for each of the drogue and main ejection charges to guarantee their
ignition and deployment of a parachute. This built in redundancy ensures the recovery of the rocket as well as the
payloads within.
The ejection charges contain a precisely measured amount of black powder as described in Eq. (1) 1 to adequately
over-pressurize the body sections of the rocket to 15 pounds per square inch (psi) each for a successful deployment
event.
3ld 2
mblackpowder  (1)
500
In Eq. (1) mblack powder is measured in grams, and l and d are measured in inches. An additional amount of black
powder is added to top off the charge, equal to a quarter the original amount, in order to make sure the over
pressurization does force the rocket apart. The primary charges are contained in 3D printed cylinders that add a
directionality to the detonation of the charge. While skeptical at first the, the 3D printed pieces are entirely reusable
and little to no damage with each burn.
The backup charges are two times as large as the main charges. This is primarily due to the issue that comes from
an incomplete/inefficient burn of the black powder which has been experienced in a previous launch by the club,
resulting in an unsuccessful recovery. The backup charges are contained within a piece of surgical tubing and zip tied
closed on both ends upon the recommendation from the members of the Kloudbuster’s Rocketry Club.
The rationale behind such a large amount of black powder is centered around the knowledge of how strong the
fiberglass body tubes actually are. If the primary charges function perfectly, then with the body tubes open to the air,
when the backup charges go off, they are immediately vented to the atmosphere and cause no harms to the rocket. If
the Primary charges don’t go off and the redundancy fails in that set of systems, the charges have to be sufficiently
large enough to force the rocket apart when confronted with pressure induced by the rocket falling ballistically, which

6
would hold all the sections together. Or if either of the main charges go off and they don’t achieve what they were
supposed to, the significantly larger charge is more likely to force out the parachute recovery systems.
The main charges are kept as close to the electronics bay as physically possible, allowing for the shortest amount
of wire to be used in the connection and eliminate the chances of a disconnect from the altimeters. However, the
drogue charges had to be moved to under the payload section due to design and mounting constraints.
A SkyAngle Cert-3 Drogue parachute, 22 inches in diameter, is used for the drogue deployment event at apogee.
According to OpenRocket, to a terminal velocity of 94 ft/s from 10,000 feet AGL to 1,000 feet AGL. The use of a
SkyAngle Cert-3 extra-Large main parachute 90 inches in diameter was chosen to support the fifty-one and a half
pounds that the rocket weighs for the final descent following the main recovery deployment event. The primary
altimeter will fire at 1,000 feet with the backup firing at 800 feet. If all goes as planned, the backup will fire, but the
charge will have been used several seconds before as fired by the primary altimeter. This final event ensures a
controlled touchdown at 26 ft/s according to OpenRocket simulations.
All separable portions of the rocket are joined together by coupler sections with a six-inch minimum shoulder (to
match or exceed the six-inch body diameter) and secured by simple nylon shear pins. These shear pins provide the
support to remain intact during boost and prevent drag separation, but cannot hold against the black powder ejection
charge at which point they shear apart and allow recovery systems to deploy. Other sections that are joined also by
couplers but do not separate (such as the airbrake to the aft airframe and the electronics bay to the forward airframe)
are secured by aluminum 4-40 tube fasteners and rivets.
Following ejection charge-induced separation, the sections of the rocket are tethered together by lengths of 1”
nylon shock cord. These segments of shock cord are 40’ long to allow the force of black powder ejection to dissipate
somewhat before the cord is pulled taught in tension. Longer shock cord reduces the energy of the system (lost by
drag) and minimizes axial loading on the shock cord, eyebolts, epoxy connections, and airframe components. The
cord is connected by ¼” stainless steel quick links and affixed to 5/16” forged steel eyebolts secured to the fiberglass
bulkheads and all centrally connected at the electronics bay where ejection charges are located.

D. Payload Subsystem
Our payload is a system of five independent cameras that record video out the sides of the rocket and can be
stitched together to create a single 360-degree video. The system is designed to begin recording once powered on. An
accelerometer on board is capable of detecting launch which then begins a 20-minute timer. Once 20 minutes have
passed, a mechanical relay will flip therefore cutting power to all five cameras. This is necessary so that the cameras
do not continue to record and write over the footage from the launch. This design and development of this payload
has enhanced the team’s ability to manufacture aluminum and steel parts as well as work with electronic parts not
directly associated with the rocket itself.

Figure 8. CAD model of payload

system

7
 1) Payload Structure
The payload team chose to use three steel plates and four aluminum 10” rods as the main body of the system.
Working with a local steel supply company, Stillwater Steel and Welding Supply, who was able to cut us three 6”
diameter circular plates that were 0.25” thick, we chose 0.375” diameter aluminum rods in order to prevent buckling
of the rods during high G environments throughout flight. From there the necessary holes on each plate were laid out
in order to mount the four rods and all the components to the plates. Though it would have been easy to simplify the
design down to only two plates instead of three plates, it was decided not to mount components from the bottom of a
plate. During ascent of the rocket it was preferred that mounted components be pushed towards the plate and rather
than pulled away from it. Using a print out of the CAD layout, the paper was placed on top of each plate and a drill
press was used to customize each plate as needed.
Two types of nuts were used for accessibility and strength purposes. Underneath each plate are four nylon lock
nuts followed by a washer that prevents vibrations from shaking those nuts loose throughout flight. On the top of
each plate are two standard nuts followed by a washer. Having two nuts greatly decreases the chance of them
vibrating and coming loose. However, all nuts also had thread locker applied in order to ensure a rigid structure
during flight.

Figure 9. Payload assembly and structural design

The structure is mounted directly on top of the electronics bay, so that a sturdy and rigid location for the payload
to be mounted to the rocket is achieved. This is important for the cameras, so that they will experience minimum
vibrations throughout flight. The upper body tube where the payload will actually occupy will then have five holes
for the camera lenses to protrude from and be able to record video properly. As the structure of the payload was
finalized and all the parts were assembled a final weight of 9.02 lbs was reached which is extremely close to our
targeted 9.00 lbs for the system.
2) Cameras
A number of cameras were considered for this payload. The original idea was to take advantage of the simple and
light weight cameras that could be used in conjunction with a RaspberryPi. However, after much research these
cameras would not provide a high enough quality or rigidity to justify their use. Next, GoPros were examined and
other similar action cameras. For the allotted budget cheaper options were sought out that had the same functionality
as that of these cameras. The RunCam Split cameras were selected. They are capable of recording in 1080p 60fps and
can utilize a 64GB SD card as well as record in five-minute segments. This ensures that even if something goes wrong

8
at some point during flight the entire video will not be lost. Another benefit to these cameras is that power loss to the
camera does not corrupt the current file and therefore produces only minimal losses of video. These cameras price fall
within the budget and are notorious for being durable since their original purpose is to be used on racing drones that
are prone to many crashes. Each camera comes with a removable Wi-Fi dongle so that settings can be modified through
a mobile app and files can be pulled from the cameras without having to remove each SD card.

Figure 10. Camera assembly

Another major component to the choice of cameras was the angle of the camera lens. GoPros are notorious for
their wide fish eye lens and make them a good fit for a situation like this. Fortunately, the RunCam Splits also are
equipped with wide angle lenses that even allow for them to be turned vertically so that each camera still has
overlapping video, but a much wider range of view from top to bottom is achieved. An early issue in design was the
requirement to be able to easily remove the payload from the rocket. This was difficult since the camera lens needed
to be as flush with the outside wall of the rocket as possible to get the best video, but this would inherently prohibit
the removal of the payload. The solution to this is using the RunCam Split camera system to our advantage and
unscrewing each lens prior to installation and then re-inserting the lens once the payload is successfully installed
within the rocket. This allows for optimum camera placement.
3) Initiation Trigger
With the initial concept of using RaspberryPi’s for the camera system, an accelerometer would be able to sense
lift off and save the previous ten seconds and the following 20 minutes so that the footage would be of just the launch.
However, with the RunCam Splits recording on the pad can begin and then using a similar accelerometer system with
an Arduino Nano to cut power to the cameras 20 minutes after launch, so that unnecessary footage isn’t recorded and
the camera does not accidentally write over the launch footage.
Two power bank batteries are wired in parallel so that the output amperage is upped to 4.2A, which is necessary
for all the cameras to work properly. One power bank will also power the Arduino Nano independently with 1A. The
power banks connect to the cameras through a mechanical relay that is used to turn off the cameras after launch. The
two power banks are designed so that when a device is plugged in the battery recognizes the device and automatically
turns on to begin providing power. However, when wiring the two batteries in parallel this proves to be a problem
since turning one battery off will inherently turn the other battery back on since it recognizes it as a connected device.
To mitigate this issue, a 3A diode was connected to each positive end of the battery. This therefore prevents current
flowing back into the battery and turning the battery on. From the relay, all five cameras are connected via micro USB
cables to their respective port on the device.
When power is turned on, all five cameras immediately begin recording. This should make syncing the cameras
when stitching them into a single 360 degrees video easier however, the Arduino Nano system also includes a buzzer
that will beep prior to launch and then sustain one long tone when ignition is detected from the accelerometer. This
will give the team a reliable auditory signal to sync the videos appropriately.
4) Testing
Testing has been completed on the payload and all the cameras function as intended along with the Arduino Nano
and relay system that shuts off power once the flight is completed. Currently, the team is working on stitching together
initial video recordings to practice for the final editing of the video. After the final launch prior to the SA Cup, the

9
team will review video footage and ensure all systems worked properly throughout launch without major issues or
vibrations during recording.

Figure 11. Battery endurance testing setup

E. Airbrake Subsystem
When designing the airbrake, there were a lot of aspects that had to be considered. Seeing as how the purpose of
the airbrake was to manipulate the airflow around the rocket, aerodynamics were a chief concern of the Controls Team.
One of the biggest choices was how many control surface fins to put on the airbrake. The Controls Team didn't want
to compromise the structure by cutting away too much of the body tube, so the size of the fins became limited. It was
decided that somewhere around 50% of the structure should remain at any point in the cross section where the holes
would be cut. This also limited how many fins there could be. Since the rocket itself was designed to have 3 aft stability
fins, it was only natural to have the airbrake have 3 fins as well. A 3 control surface fin design would also make
downstream air flow effects symmetric over the aft stability fins which was another big concern. That's why it was
decided to place the control surface airbrake fins “in-between” the aft stability fins. This greatly reduced changes in
the flow over those aft fins in CFD simulations which it was felt would keep the rocket more stable.
On the note of stability, it was also important to be able to show that the airbrake wouldn't change the location of
the center of pressure in such a way that the rocket became unstable upon deployment. For initial testing, the center
of pressure was assumed to be essentially wherever the airbrake fins deployed. This was because the surface area of
the fins would be larger than any other item extruding from the rocket causing pressure buildup at that location to
dominate any other pressure contributions. This was tested further using CFD testing for each design using the
methodology described in the next section.
Outside of aerodynamics, there were many mechanical considerations that had to be accounted for. Most
importantly, complexity played a big role in what design choices were made. It wouldn't matter how good of an
airbrake was designed if it was so complex that it wasn't realistic to build with the resources provided to the team.
Additionally, more complexity also leads to problems with reparability. Since the competition takes place in the
middle of the desert, the airbrake needed to be designed in such a way that it could easily be worked on with tools that
are easy to transport to such a locale. Another benefit to simplicity is a simple matter of being less likely to
malfunction. A more basic mechanism has less moving parts and less things that can go wrong. This also played into
the weight of the airbrake system because the less components inside the airbrake, the less it weighs. Having less
weight is important because it allows the rocket to reach a higher altitude with less propellant.
The final big consideration was which design would allow the control surface airbrake fins to deploy the quickest.
This is important because the airbrake is going to reactively deploy. This means it needs time to react to changes in
velocity and acceleration on the fly in order for the rocket to hit its target altitude. Simulations estimated that the entire
coast phase would last about 18 minutes, with only 5 seconds of that occurring at high enough speeds for the airbrake
to make a significant drag contribution to slow the rocket. If the deployment mechanism is too slow, this 5 second
window of significant drag opportunity will be missed making the airbrake’s control algorithm dramatically less
effective.

10
 1) Hardware Design and Testing Methodology
The design process for the airbrake was largely iterative, following the same general pattern of steps over a series
of rough designs until the best final design emerged. First, a rough design was created on paper. This includes general
shapes of control surface fins, deployment mechanisms, and a rough internal layout.
Next, a Solidworks model was created of just the control surface fins and body tube. This model was then placed
into a Solidworks assembly of the team’s rocket so that CFD testing could be done. A series of test were conducted at
different Mach numbers, densities, and temperatures to simulate different heights and stages of the flights. A control
test with no control surface fins deployed was also conducted at the same test points. These results were then compared
to an OpenRocket simulation with a plain body tube in the spot of the airbrake which simulates a no control surface
deployed configuration. The drag results from the CFD control simulation and the OpenRocket Simulation were within
10% of one another which often equated to 0.5 lbf or less. This was determined to be close enough, because CFD in
general isn't the most accurate. Once the mesh and settings were validated by the control case, the same settings were
used with the airbrake control surface fins fully deployed. Drag produced and general airflow patterns were monitored
during these tests. This was used to both see which designs produced the most drag, and what the effect of that drag
disturbance was on downstream air. The best design was a combination of enough drag to effectively slow the rocket,
but not so much so that it significantly disrupted flow over the aft stability fins. These results were used to show what
size of control surface fin was needed for a design to be worth pursuing, and that information was used in the next
step
Once a design was considered aerodynamically possible, the information on minimum control surface size was
then plugged back into the rough design. At this point, the mechanism for deployment was evaluated to see if it could
still work with the new sizing. If a mechanism wouldn't work, possible alternatives or modifications to the design were
explored. If it still wasn't possible, the design was discarded at this point. Of the 6 proposed designs, 3 of them were
stopped at this point. The designs that were left moved on to the next stage for even more computer aided testing.
The next round of CFD was done to gather more data on the designs integrated into the rocket body. Primarily,
this was done to monitor the stability of the rocket so the Controls Team could see how the airbrake impacted the
center of pressure location. As discussed earlier, the initial thought was to put the airbrake control surfaces directly on
top of the non-airbrake center of pressure location in its farthest forward point. The integrated airbrake and rocket
were evaluated over a range of Mach numbers to see how the center of pressure changed. The airbrake was evaluated
at 25%, 50%, 75%, and 100% deployment of control fins. Of course, the settings from the verified control mentioned
earlier were still used in an attempt to reduce error and increase accuracy of the simulations. At each test point, the
torque in each axis was divided by the force normal to it relative to an axis at the tip of the nose cone as seen in the
equations below. Of course, the rocket is asymmetric so calculating the CoP about only 2 axis did not provide an
accurate representation of what was really happening aerodynamically. To combat this, a series of axis were set up at
the tip of the nose cone, each offset by 22.5 degrees from the last. This allowed for the center of pressure to be
calculated about 6 independent axis, and for the values to be averaged together to create one more realistic number.
𝐶𝑜𝑃𝑥 = 𝐶𝑜𝑃𝑦 =
The second purpose for the additional CFD was to create a CD vs Mach curve for the airbrake. This information
was needed to accurately model how effective the airbrake would be at reducing altitude. Initially, the Controls Team
attempted to use OpenRocket for this task. There is a plugin for using a custom CD vs Mach curve for a rocket, but
there was no way to be able to switch between the different curves for different levels of airbrake deployment. To get
around this hurdle, the Controls Team went through the source code of the OpenRocket program and used the same
methodology to create a custom launch simulation in MATLAB. The program takes initial time, position, velocity,
and acceleration data generated by OpenRocket from the second that the boost phase stops and iterates through at a
defined time step to calculate a final altitude. Once again, using a control case of no control surface fin deployment
the simulation estimated within 2 meters of what OpenRocket did which was deemed an acceptable level of error. The
reasoning for choosing MATLAB to do this was because initially the Controls Team planned to use a PID algorithm
which MATLAB would more easily be able to tune.
At this point, the remaining airbrake designs could truly be compared to one another. Simulations were run using
the MATLAB code with the airbrake fully deployed to see exactly how much altitude it would shave off. The design
that performed best was then selected as the final choice. A parts list was put together and the needed materials for
the design were ordered. At this point, the airbrake was assembled and the MATLAB code was used to test different
algorithms for control surface fin deployment and find the optimal way to slow down the rocket in flight with the
chosen design.

11
 2) Design Iterations – Design 1

Figure 12. Design 1 in the fully deployed state.

The Controls Team’s first design started with a linear actuator place at the center of the airbrake system. A center
ring was used to support the airbrake system. There was also a plus-shaped component placed on top of the actuator.
The component is tied, by wire, to a corresponding rod below it, at the aft end of the airbrake system. For each fin,
there are four total rods. One rod attached to the wire, another attached to a bar on the back side of the fin, and the last
two attached to a bar that is connected to the ring. These rods had holes at both ends that connected to a center rod
connecting the four different rods to allow for rotation. Each fin was slightly curved to stay streamline with the rest
of the body of the system and is connected to the body using a hinge. When the linear actuator activates it would push
the plus-shaped component upwards, pulling the wire as well as the hinge upward. This would swing the fins upward
to a location almost perpendicular to the airbrake system.
What made this design likable was that was designed to be completely modular. It also allowed for a large amount
of fin surface area to create a significant amount of induced drag. However, due to the amount of drag that was created
a powerful actuator was required. The Controls Team had trouble finding a linear actuator that reached the parameters
that were needed and that was within the team’s budget. Another restraint with this design is that it was not able to
deploy quickly.
In the end, the Controls Team decided not to use this design and look into other viable options for an airbrake
system for our 2018 Spaceport America Cup entry.
3) Design Iterations – Design 2

Figure 13. Design 2 in the fully deployed state

12
 This design focused on maximizing surface area that could be used for induced drag of the rocket via the airbrake
while allowing for a rotational stepper motor to be the means of actuation. This resulted in a design that was made up
of three “fins” that would slide out of the airframe, perpendicular to the airframe, in a linear fashion. The motion and
amount of deployment was to be controlled by the rotation of a cylindrical, geared “hub” that would serve as the
interface between the drive motor and the fins. The fins also would be stacked on top of each other, with some spacing
in between each fin to allow for support “tracks” that the fins would ride along and in turn creating fairly tight
tolerances that the assembly would need to conform to. The components were intended to be manufactured out of
6061-T6 aluminum and assembled using off-the-shelf hardware.
This design seemed very feasible due to the availability of parts and similarities between various parts that would
needed to be manufactured. However, upon completing the 3D/CAD model of the hardware, looking at the
manufacturing equipment available and determining how the hardware would be assembled, the team came to the
conclusion that this specific design would be very difficult to complete and implement with the available time and
resources while also creating a piece of hardware that could be easily serviceable, should parts need to be altered or if
parts were to break.
Although this design was not chosen to be a part of our Spaceport America Cup entry for the 2018 competition,
the controls team intends to refine the design and explore alternative manufacturing techniques so that the design has
the possibility of becoming actual hardware on future rockets.
4) Design Iterations – Design 3

Figure 14. Cutaway capture of Design 3

This design was chosen as the hardware that would be manufactured, assembled, and installed onto the Cowboy
Rocketry entry for the 2018 Spaceport America Cup competition.
This particular design was originally destined to be “thrown out” as a design option for the Spaceport America
Cup rocket that our team would be putting together. This was because the amount of surface area that would be
available for use in creating induced drag on the rocket was much lower that the amount of surface area that was
expected to be obtained on alternate designs. However, this design was kept as a possibility and ultimately selected as
the final design due to the simplicity of the design, compared to other options that were explored.
The final design consists of three fins that rotate out of the airframe in a radial fashion, perpendicular to the
airframe. The actuation will come from a central stepper motor that interfaces with the fins by a central gear and
matching gear teeth on each of the fins. The fins and central gear will be assembled between two discs and everything
will be fastened together with off-the-shelf hardware. The fins, central gear, top and bottom discs will be machined
out of ¼ inch 6061-T6 aluminum plate, by a three-axis CNC machine. The stepper motor will be attached to and be
supported by the top disc in the assembly.
Another integral component that will be cut to size and epoxied to both the top and bottom discs, allowing for
interfacing with the fins, will be a 1/16 inch layer of Ultra-High Molecular Weight Polyethylene sheet (UHMW). The
UHMW sheet will serve as a glide surface for the fins to move along, with very little friction. The UHMW sheet is
also a durable material and will be able to stand up to heat of the New Mexico desert during competition.
This “sandwich” of components (bottom disc, fins, UHMW sheets, and top disc) will be affixed into the rocket
with the help of repurposed bulkhead plates and segments of 1x1 inch t-slot framing. The bulkhead plates will be

13
epoxied into the coupler section using high strength rocketpoxy, directly below the airbrake’s section of airframe and
directly above the motor tube. The t-slot sections will then be fastened to the bulkhead plates providing somewhat of
a frame for the airbrake. The bottom discs of the airbrake “sandwich” will then be fastened to the ends of the t-slot
framing sections, effectively securing the airbrake hardware. This method was chosen for mounting due to the fact
that the coupler, that the bulkhead plates will be attached to, can be separated from the rest of the airframe. This will
allow for easier access to the airbrake hardware for any maintenance or adjustments that might be necessary by
essentially being able to pull the entire airbrake assembly outside of the airframe.
While this had not been the intended final design from the beginning due to the lower surface area, it was the
design that was feasible for the team to manufacture and assemble with the resources and time available. This design
will also serve as a stepping stone for the team to create more advanced airbrake systems in the future.
5) Avionics Hardware Considerations
The avionics hardware is probably the area of the airbrake that seemed like it would be the easiest, but ended up
being one of the hardest. One of the biggest challenges facing the Controls Team was putting together a good enough
avionics that would do the job without spending a ton of money. It's easy to get carried away and spend thousands of
dollars on sensors that can detect the pressure change from an insect flying by at 1 MHz, but that amount of data would
almost be too much. It was also important to not get carried away getting a ton of fancy electronics that also weigh a
bunch for obvious reasons. Along with weight, it was imperative to get sensors that aren't bulky. The airbrake on the
teams rocket was only budgeted 20” worth of tube and that's a lot less than it sounds. That also plays into the fact that
the sensors needed to be low power. There wasn't much space available for large batteries, and with how hot the inside
of the tube may get in the desert, large batteries could pose an overheating problem inside the rocket.
A challenge to finding cheap and lightweight sensors was making sure they were good enough to accomplish the
goal of providing good information quickly to the Arduino. This meant finding sensors that had high enough accuracy
and resolution to calculate what the projected altitude is on the fly, and when motor burnout occurs. On the flip side,
sensors that read too quickly and produce noise are a bad thing in this case. The Controls Team strived to strike a
balance between the two areas. It was also very important to get sensors that could easily communicate to the Arduino
without significant post processing. This allowed for more simplistic coding and less potential for misreading or errors.
The last problem faced by the Controls Team was finding sensors that could withstand a rocket launch. The fact
is most products available for the Arduino were not designed with 20g’s of acceleration in mind. This was a problem
experienced first-hand by the OSU Senior Design team when launching their quadcopter. To solve this, an effort was
made to ensure the data sheets of every sensor used was checked to ensure failure would not happen from launching
alone. Additionally, every component or wire that could possibly vibrate loose is hot glued down to ensure it stays
put.
6) Software Layout Considerations
The software was really the heart of the whole airbrake. It harmonized the airbrake hardware and electronics
together into one functional unit. One of the biggest challenges that the Controls Team focused most on was optimizing
the software. Since the airbrake needed to react to what was happening in real time, every clock cycle of the Arduino
was important. The Arduino’s 16 MHz may sound like a lot, but anything outside of the simplest calculations can take
up a bunch of cycles. For example, the SD card logging software initially opened and closed the log file between every
data point it was writing, but after changing that to writing a whole line of data points at once the cycle time for polling
and writing the sensor data reduced by 3/10ths of a second. That change allowed the software to go from polling
sensors at 3 Hz to around 12 Hz.
Repeatability is another big factor with the code. Great effort was taken to ensure it was as simplistic as possible
to ensure it didn't do anything unexpected. This means more than just simply using the least amount of lines and letters
possible in the code, because when it comes down to it the Arduino probably won't get confused. The code was kept
simple for the benefit of those working on it. Especially with multiple people editing code, it's important that variables
are intelligently named, the code flows in a logical manner, and functions are used to reduce redundancy in the code.
This also made the code easier tune for the airbrake which once again leads to more predictable results.
Focus on simplicity doesn't mean that the code is too bare bones, however. A buzzer was implemented which will
sound unique tones for normal operation, malfunctioning sensors, abnormal readings, or low batteries. Such
redundancies allow the airbrake to check itself in the case of human error. This adds a second layer of security into
the airbrake operations and make it less likely to malfunction. Of course, the buzzer isn't the only diagnostic tool. An
SD card was also implemented which allows for the logging of all data that the airbrake gathers from sensors. It also

14
records what position the airbrake is in during flight. This is the “black box” of the airbrake ensuring that if something
does go wrong, it can be more easily diagnosed after the event. The electronics are set up in such a way that anything
short of the SD card physically breaking wont corrupt the log files on board ensuring flight data will always be
available.
7) Final Avionics Layout
The Controls Team decided to base the avionics around an Arduino platform. A full list of parts can be seen
below. The Arduino Uno R3 was chosen due to its wide range of support, shields, and libraries available making it
the easiest platform to work with. Powering the Arduino is a 2000mah 7.3Wh lithium ion battery that is regulated
through an EnergyShield 2 Basic. Initial testing has shown that this battery should provide power for at least 7 hours
which should be more than enough. Connected to the Arduino is a series of sensors. The first is an Adafruit BMP280
Sensor. It ranges from -500m to +9000m above sea level. Its accuracy is +/- 1m which should be more than sufficient
for what's needed for the code. The next sensor is the Sparkfun ADXL377 accelerometer which ranges from -200g to
+200g’s. This is an analogue sensor, and initially the Controls Team ran into issues with the Arduino because it only
had a 10bit ADC which did not provide sufficient resolution. To counteract this, the ADS1115 Board which provides
16bit resolution was used in conjunction with the accelerometer. In order to create logs of all the data being collected,
the Adafruit SD Breakout Board was used which provides a quick means for storing data. This board is particularly
nice because the SD card is able to be clicked into place instead of just resting in a slot. Power was routed to all these
sensors from a distribution board made from a simple perf board. This allowed one power port on the Arduino to be
distributed to 8 different wires. A Piezo Buzzer was also used to play noises for both normal operation and if there is
an error detected in the sensors on startup.
The airbrake itself is actuated by a NEMA 17 Stepper Motor that has 0.6N*m of holding torque. Simulations run
by the Controls Team show this is more than enough to actuate the airbrake, even at the highest force loading. The
motor is powered by a 3.7V lithium ion battery that plugs into a barrel jack connector. There is a 100 μF capacitor in
between the battery and the motor to handle any voltage spikes that may occur to protect the motor. A DRV 8825
stepper motor driver is used to interface between the motor and the Arduino allowing for precise control over the
airbrake’s movement. A full parts list can be seen below.
The entire airbrake was wired using standard electronic jumper cables. The diagram for this wiring can be seen
in the picture below along with a flowchart for the code running on the Arduino.

Figure 15. Final avionics wiring schematic.

15
 Figure 16. Software Flow Chart

8) Results of Avionics Testing

The first test of the avionics hardware was performed in March. It consisted of launching the Arduino, SD Board,
Accelerometer, and an altimeter on a rocket to make sure they could all handle the forces of a launch. For this test, a
SL100 was being used as the altimeter, but that has since changed. The reason for this was problems with the altimeter
being difficult to interface with the Arduino. All electronics did successfully held up and recorded data as expected
for this test.
Next, the Avionics package did an endurance test to make sure the code performed as expected and to check how
big the log files can get. The Arduino was plugged into a
power outlet and run for 3 days collected data the entire time.
Throughout the test, sensor data read remained constant and
the log file stayed at a reasonable size meaning the code was
good and the test was successful.
As of writing this, the final test performed was to see
how long the Arduino lasts under the battery power pack
powering it. The Arduino ran for 7.3 hours before the battery
pack was switched off due to concerns of the battery level
getting too low. This constituted a successful test due to
demonstrating the ability to run all day without being
recharged.
9) Final Airbrake Configuration
The final full airbrake configuration can be seen below.
The fiberglass plate on the bottom screws into the motor, and
then the 80/20 rods are connected by bolt to the plate. The
reason for these rods is to space the airbrake up above the
coupler tube connecting the airbrake tube to the aft sections.
Threaded rods are then screwed into the 80/20 rods and go
through the entirety of the airbrake. A forged eye bolt is
screwed into the top plate which is held on by bolts. This
design provides a means for the parachute to connect to the
motor without sacrificing structural strength.
Figure 17. Final airbrake

16
 III. Mission Concept of Operations Overview (CONOPS)
During flight at the 2018 Spaceport America Cup, Cowboy Rocketworks’ launch vehicle, Results May Vary, will
experience various stages of performance and functionality throughout flight.

Figure 18. Cowboy Rocketworks Results May Vary Concept of Operations Overview.
The flight trajectory of Results May Vary begins with motor ignition which is achieved by electrical arming and
launch at the ESRA-provided launch control system. The student-prepared electronic match burns briefly, igniting the
forward-most end of the propellant grain. As the Aerotech M1939W composite motor comes up to pressure, and the
liftoff sequence is initiated. As thrust builds, the trust to weight ratio exceeds one, and the rocket begins ascent up the
ESRA-provided launch rail. The 3D-printed airfoiled rail buttons ensure direct vertical ascent as the rocket clears the
17-foot tall rail. Once the aft-most rail button is cleared, the rocket is free of all ground support equipment and begins
the powered ascent phase of flight.
During ignition and liftoff, the airbrake as developed by Cowboy Rocketworks, remains disengaged as the system
registers the current altitude remains below the expected altitude at motor burnout. Upon clearing the launch rail, the
rocket has achieved the necessary velocity for the fins to provide aerodynamic stability. This aerodynamic stability is
confirmed by a thrust to weight ratio at liftoff greater than 5:1. With Results May Vary launching under a 6:1 ratio,
the fins will provide the necessary stability for powered ascent. The rail departure velocity is 62 ft/s according to the
model’s OpenRocket simulation.
Powered ascent under motor boost lasts 6.2 seconds during which the airbrake is still mechanically disengaged,
and the actively operational subsystems are propulsion, payload, and recovery (altimeter recordings). Upon motor
burnout, Results May Vary is simulated to be moving at 518 miles per hour, having passed maximum velocity just a
half second beforehand at 561 miles per hour and an altitude of 3700 feet AGL.
Following motor burnout, the unpowered ascent phase of flight begins during which the majority of its altitude
will be reached. Propulsion systems are now inactive and reload hardware becomes unused weight. Presently the
airbrake will engage and become mechanically operable, taking altimeter input, running the software program, and
automatically extruding the fins to gradually induce drag and reduce the expected apogee to precisely 10,000 feet
AGL.
After successful airbrake operation, apogee is expected to be 10,000 feet at which point the redundant altimeters
register maximum altitude. The primary altimeter, an Altus Metrum TeleMega will fire the drogue charge to separate

17
the rocket into two halves and deploy the 22-inch SkyAngle Cert-3 drogue parachute. The secondary altimeter, a
StratologgerCF, will also fire at apogee. As the drogue parachute unfurls, downward velocity will increase to terminal
velocity of 141 ft/s. Here the airbrake will return to gridfin zero angle of attack (flush with airframe) and become
mechanically inactive for the remainder of the flight.
Results May Vary will fall from 10,000 feet to 1,000 feet under the drogue parachute with the redundant altimeters
monitoring descent, anticipating the altitude for main event ejection. At 1,000 feet AGL, the primary altimeter will
fire the main parachute ejection charge. The redundant altimeter will fire at 800 feet and separate the rocket should
the primary altimeter fail.
The main ejection charge pushes all components inside the forward airframe out of the rocket. This includes two
sets of shock cord, the SkyAngle Cert-3 XL main parachute, but not the payload. The main parachute will unfurl and
reduce the descent rate of the entire rocket to 14.5 ft/s at which point the segments of Results May Vary will touch
down gently near the launch site.
Mission success is characterized primarily by two recovery deployment events, touchdown and recovery without
damage to any components of the rocket including aero-structures, recovery components, and payload. Apogee close
to 10,000 feet is also important but only significant following a safe return to the ground.
These collective mission phases contribute to a successful flight at the 2018 Spaceport America Cup.

IV. Conclusions and Lessons Learned

Cowboy Rocketworks has come a long way in two years. The team is tremendously excited to be competing in
the largest collegiate rocketry competition in the world. Its members look forward to building more of the rockets
from scratch, especially now that they are making their SRAD fiberglass airframes and sheets. Regardless of the
outcome at the 2018 Spaceport America Cup, the team has achieved so much, and the road here has shaped the
student’s lives during their time in college and will affect their professional careers for years to come.

18
 Appendix A: Systems Weights, Measures, and Performance
Rocket Design

Rocket
Stages: 1
Mass (with motor): 63.2 lb
Stability: 2.36 cal
CG: 97.065 in
CP: 112 in

M1939W-P
Motor Avg Thrust Burn Time Max Thrust Total Impulse Thrust to Wt Propellant Wt Size
Altitude 10857 ft
M1939 1582 N 6.52 s 2084 N 10340 Ns 5.62:1 11.7 lb 3.86/29.6
Flight Time 181 s
W in
Time to Apogee 26.1 s
Optimum Delay 19.2 s
Velocity off Pad 45.2 mph
Max Velocity 627 mph
Velocity at 84.4 mph
Deployment
Landing Velocity 9.07 mph

M1500G-0
Motor Avg Thrust Burn Time Max Thrust Total Impulse Thrust to Wt Propellant Wt Size
Altitude 5061 ft
M1500 1491 N 3.5 s 1717 N 5217 Ns 6.14:1 5.8 lb 2.95/26.2
Flight Time 129 s
G in
Time to Apogee 18.5 s
Optimum Delay 15 s
Velocity off Pad 52.3 mph
Max Velocity 400 mph
Velocity at 86.3 mph
Deployment
Landing Velocity 9.4 mph
Parts Detail
Sustainer

Mad Cow 6" Nosecone Fiberglass Ogive Len: 30 in Mass: 3.42 lb

(1.07 oz/in³)

Nose Cone Coupler NEW Diain 5.775 in Len: 9 in Mass: 1.15 lb

FIBERGLASS Diaout 5.998 in
(1.2 oz/in³)

Body tube Fiberglass Diain 6.01 in Len: 42 in Mass: 4.45 lb

(1.07 oz/in³)
Diaout 6.17 in
SkyAngle Cert-3 XLarge Ripstop nylon Diaout 126 in Len: 11 in Mass: 2.81 lb
(0.22 oz/ft²)

Shroud Lines Tubular nylon Lines: 3 Len: 100 in

(25 mm, 1 in)
(0.312 oz/ft)

24" Parachute Protector Diaout 5.5 in Mass: 0.201 lb

Shock Cord 1" Blue Nylon Len: 480 in Mass: 1.1 lb

Wildman Shock
Cord
(0.44 oz/ft)

Payload Diaout 5.984 in Mass: 9 lb

Slip Band Blue tube Diain 6 in Len: 2 in Mass: 0.212 lb

(0.751 oz/in³)
Diaout 6.17 in
Electronics Bay Blue tube Diain 5.775 in Len: 14 in Mass: 1.14 lb
(0.751 oz/in³)
Diaout 5.998 in
Drogue Compartment Fiberglass Diain 6 in Len: 22 in Mass: 0.146 lb
(1.07 oz/in³)
Diaout 6.17 in
SkyAngle Cert-3 Drogue Ripstop nylon Diaout 22 in Len: 2 in Mass: 0.375 lb
(0.22 oz/ft²)

Shroud Lines Tubular nylon Lines: 3 Len: 24 in

(25 mm, 1 in)
(0.312 oz/ft)

Drogue Shock Cord 1" Blue Nylon Len: 480 in Mass: 1.1 lb
Wildman Shock
Cord
(0.44 oz/ft)

Tube coupler Cardboard Diain 6 in Len: 12 in Mass: 1 lb

(0.393 oz/in³)
Diaout 6 in
24" Parachute Protector Diaout 5.5 in Mass: 0.201 lb

Air Brake Fiberglass Diain 6.013 in Len: 20 in Mass: 2.12 lb

(1.07 oz/in³)
Diaout 6.17 in
Tube coupler Fiberglass Diain 6.013 in Len: 12 in Mass: 1 lb
(1.07 oz/in³)
Diaout 6.013 in
Air Brake Components Diaout 5.984 in Mass: 6 lb

Aft Section Fiberglass Diain 5.99 in Len: 32 in Mass: 3.39 lb

(1.07 oz/in³)
Diaout 6.17 in
Motor Mount Tube Blue tube Diain 3.88 in Len: 24 in Mass: 1.64 lb
(0.751 oz/in³)
Diaout 4.06 in
Forward Centering Ring Fiberglass - Diain 3.858 in Len: 0.187 in Mass: 0.234 lb
Wildman Diaout 5.99 in
(1.21 oz/in³)

Mid Centering Ring Fiberglass - Diain 3.858 in Len: 0.125 in Mass: 0.156 lb
Wildman Diaout 5.99 in
(1.21 oz/in³)

Aft Centering Ring Fiberglass - Diain 3.858 in Len: 0.125 in Mass: 0.156 lb
Wildman Diaout 5.99 in
(1.21 oz/in³)

Fins (3) Fiberglass Thick: 0.187 in Mass: 2.74 lb

(1.07 oz/in³)
 Appendix B: Project Test Reports

Tests for Cowboy Rocketworks in preparation for the 2018 Spaceport America Cup
Date Type Description Status Comments
Begin CFD optimization of Based optimization off
1/15/18 Other Successful
fins current research papers
First iteration of SRAD Smaller diameter for first
1/19/18 Other Successful
fiberglass tube iteration
Begin CFD validation of CP Rocket shouldn't destabilize
2/5/18 Other Successful
locations with fins when airbrake deploys
2/16/18 Ground Ejection charge test Successful For internal pressure validity
In- First flight test using SRAD
2/17/18 Successful Small scale test flight only
Flight tubes
In-
3/11/18 High-G test of SRAD tubes Successful 15G simulated, good recovery
Flight
Large-scale ejection charge
4/6/18 Ground Successful Took 2 attempts
test
In- First large-scale test flight Deploy of main at 1200m not
4/8/18 Successful
Flight on M1500G 1200ft
Validate airbrake and code Airbrake actuates successfully,
4/16/18 Ground Successful
work on ground Code ran for 2 days
4/16/18 Ground Check Battery life of avionics Successful Battery lasted over 7 hours
Endurance test for camera Batteries allow 5 cameras to
5/3/18 Ground Successful
batteries record for aproximately 3 hours
Full-scale ejection charge With airbrake and payload
5/30/18 Ground TBD
test hardware
In- Full motor, payload and
6/1/18 Full-scale flight test TBD
Flight airbrake functionality
Verify video stitching Video stitching using
6/2/18 Ground TBD
techniques commercial software

Altimeter Bay wiring schematic

22
 Appendix C: Hazard Analysis
Team Rocket/Project Name Date

Cowboy Rocketworks Results May Vary 5/25/2018

Hazard Possible Causes Risk of Mishap and Mitigation approach Risk of injury after
Rationale mitigation

Motor ignites early on launch rail Short in electrical ignition system Low; Most of launch Arm recovery Low
system will be set up by energetics before
Spaceport America Cup inserting igniter.
personnel and launch Should the rocket
control is overseen and
launch early it will not
operated by Spaceport
America Cup personnel. endanger spectators
by falling
in an uncontrolled
manner

Accidental launch ignition by Have only one crew

launch control crew member at launch rail
during and after
insertion of igniter.

Touch ends of igniter

wire together during
insertion to reduce
chance of accidental
ignition of igniter.

Fuel grains ignite during flight Flames or sparks impinging on Low; No preparation Store fuel grains in a Low
preparation or transportation fuel grains activities for Results flame-resistant case
May Vary require open until motor is loaded.
flames and few have the
potentiality of creating
sparks

Black powder ignites during flight Flames or sparks impinging on Medium; Active Store black powder in
preparation or transportation black powder altimeters can sometimes flame resistant
falsely ignite attached case until flight
charges preparation
Electronics unintentionally ignite Completely disengage
black powder all power sources
from electronics
during
loading of black
powder charges

Only arm altimeters

when rocket is in
launch position

Electrocution by electronic systems Short created by accidentally Low; power sources are Completely disengage Low
during flight preparation touching exposed metal of all low voltage (9V or all power sources
electrical systems less) and wires being while working on
used are insulated electrical systems

23
 Appendix D: Risk Assessment

Team Rocket/Project Name Date

Cowboy Rocketworks Results May Vary 5/25/2018
Hazard Possible Causes Risk of Mishap and Mitigation approach Risk of injury after
Rationale mitigation
Explosion of solid- Cracks in Propellant grain Medium-Low; Motor in Visually inspect grain Low
propellant rocket use is a large motor but is during reload
motor during launch Debonding of propellant a COTS motor. Motor Inspect motor casing for
with blast or flying from wall will be damage and cleanliness
debris causing injury
Gaps between propellant reloaded by experienced No non-essential personnel
sections and/ or nozzle team members in launch crew
Chunk of propellant Launch crew farther than
breaking off and plugging 200 feet from rocket at
nozzle launch
Motor case unable to
maintain operating
pressure Follow reload procedures
outlined in instruction
Motor end closures fail to manual
hold
Rocket deviates from Rocket does not achieve Low; Fins are secure, and Firmly attach fins with high Low
nominal flight path, proper lift off speed flight proven. TWR is in grade epoxy and sand
comes in contact with excess of 6.8 fiberglass at epoxy joints to
personnel at high speed ensure proper bonding
Failure of rocket fins Ensure Proper lift-off
weight
Recovery system fails Altimeter fails Medium; many timing Use redundant recovery Moderate
to deploy, rocket or Deploy charges are not critical events lead to systems with multiple
payload comes in properly prepared several chances for failure
contact with personnel
Test samples from
deployment charges
Drogue deploys too Perform static ejection
early/too late charge tests with parachutes
in flight configuration
Perform flight tests of
rocket in competition
configuration
Visually track all rockets in
flight until ground contact
Recovery system Parachute is improperly Medium; an improperly Spend time practicing Moderate
partially deploys and folded/stored folded parachute with folding parachute and
comes in contact with almost certainly cause the double check folding work
personnel recovery system to in field.
improperly deploy. Perform static ejection
Chutes packed into tube charge tests with parachutes
too tightly may possibly in flight configuration
not deploy when rocket
separates

24
Recovery system Improper handling of black Medium; Black powder is Store Black powder Low
deploys detonates powder a volatile material that can charges in metal
before launch and Glitch in electronic deploy be set off accidentally Keep minimal number of
causes an injury systems people near armed
Wait until ready to launch
to arm recovery system
Main Parachute High winds cause rocket to High; any amount of wind
Use dual deploy with a Low
deploys at apogee and drift can cause a rocket with a
small drogue at apogee
rocket drifts into an high-altitude apogee and main chute at lower
area where it could deploy to drift a longaltitude.
cause an injury distance Launch far from any
populated areas and
Ensure winds are not
above safe speeds prior to
launch
Rocket motor does not Poor propellant consistency Low; commercial motor Wait until range safety Low
ignite when command and igniter are being used officer gives go ahead to
is given but does ignite Damp propellant approach rocket
when team approaches
Have team approach
slowly and carefully and
Improperly installed igniter watch for signs of ignition

Rocket falls from Rail buttons are not Medium; a heavy rocket is Use proper construction Low
launch rail during properly secured to hold prone to slippage and puts techniques and materials
preparation weight of rocket more stress on rail buttons for rail buttons
Rocket slips from the hands Load rocket slowly and
of the team while putting it ensure that load crew has
on the rail control of rocket at all
times

25
 Appendix E: Assembly, Preflight, and Launch Checklists

Cowboy Rocketworks
Materials Checklist
*All materials should carry duplicates when possible*

 Results May Vary  Electronic matches

 98/10240 Motor casing  Black powder canisters
 98mm forward closure  Model rocket wadding
 98mm aft closure  Precision screwdrivers
 98mm forward seal disk  Arming screwdrivers
 Motor igniters  Grease
 Masking tape  Vinyl gloves
 Electrical tape  Epoxy
 Zip ties  Mixing cups for epoxy
 Cordless drill  Mixing sticks for epoxy
 Cordless drill bits and  First aid kit
attachments  Sandpaper
 9V batteries  Shop towels
 Wire strippers  Cleaning supplies
 Pliers  Knife
 Table scale  Personal Identification
 Hanging scale  Pen
 Measuring tape  Clipboard
 Black powder

26
 Cowboy Rocketworks
Assembly Checklist

 Ejection charges wired into altimeters

 Fresh battery connected properly
 Fresh battery secured properly inside altimeter bay
 Fresh battery with taped snap connector
 Altimeter bay sealed and secured with switch OFF
 Altimeter bay is screwed together and secure
 Black powder charges are secure on either end of altimeter bay
 Ensure static ports on altimeter bay are free and open to air
 Attach payload to the electronics bay
 Attach lenses to cameras
 Install airbrake
 Confirm that screws are tightened
 Recovery devices connected via quick links to (6) eyebolts
 Parachutes and shock cord packed on correct side of parachute
protectors
 Vent holes free of paint, gunk, dirt, and fuzz
 Shear pins installed through nose cone and forward airframe
 Shear pins installed through airbrake airframe and ebay
 Fasteners installed through altimeter bay and forward airframe
 Fasteners installed through airbrake and aft airframe
 Insert motor
 Secure motor retention

27
 Cowboy Rocketworks
Pre-Flight Checklist

 Motor installed
 Motor retention good
 Ensure static ports on altimeter bay are free and open to air
 Confirm all shear pins are in place
 Confirm all tube fasteners are in place
 Rail buttons secure
 Tracking on
 Flight card completed

28
 Cowboy Rocketworks
Launch Checklist

 Strip igniter wire leads

 Igniter on hand for installation once on the pad
 Confirm tracking on
 Slide rocket onto launch rail
 Arm electronics via external switches
o altimeters
o airbrake
o payload
 Install igniter in motor
 Connect igniter leads to power
 Good luck wraps
 Confirm continuity

29
 Appendix F: Engineering Drawings

(THIS PAGE BLANK, CONTINUED ON NEXT PAGE)

30
 2 1

.13

B B
.32 2.38
2.70
.26

6.00

.23

UNLESS OTHERWISE SPECIFIED: NAME DATE

A DIMENSIONS ARE IN INCHES

TOLERANCES:
FRACTIONAL
DRAWN

CHECKED TITLE:
A
ANGULAR: MACH BEND
ENG APPR.
TWO PLACE DECIMAL
THREE PLACE DECIMAL MFG APPR.

INTERPRET GEOMETRIC Q.A.

PROPRIETARY AND CONFIDENTIAL TOLERANCING PER:
COMMENTS:
THE INFORMATION CONTAINED IN THIS MATERIAL
SIZE DWG. NO. REV

A base Rev A
DRAWING IS THE SOLE PROPERTY OF
<INSERT COMPANY NAME HERE>. ANY
REPRODUCTION IN PART OR AS A WHOLE FINISH
WITHOUT THE WRITTEN PERMISSION OF NEXT ASSY USED ON
<INSERT COMPANY NAME HERE> IS
PROHIBITED. APPLICATION DO NOT SCALE DRAWING SCALE: 1:4 WEIGHT: SHEET 1 OF 1

2 1
 2 1

106.77°
.37
.38 R2.20 .24
1.70
R.68

B 1.30
160.30° B
2.07

R.13 R3.01

R.13

.24

.24
.25

UNLESS OTHERWISE SPECIFIED: NAME DATE

A DIMENSIONS ARE IN INCHES

TOLERANCES:
FRACTIONAL
DRAWN

CHECKED TITLE:
A
ANGULAR: MACH BEND
ENG APPR.
TWO PLACE DECIMAL
THREE PLACE DECIMAL MFG APPR.

INTERPRET GEOMETRIC Q.A.

PROPRIETARY AND CONFIDENTIAL TOLERANCING PER:
COMMENTS:
THE INFORMATION CONTAINED IN THIS MATERIAL
SIZE DWG. NO. REV

A Fin Rev A
DRAWING IS THE SOLE PROPERTY OF
<INSERT COMPANY NAME HERE>. ANY
REPRODUCTION IN PART OR AS A WHOLE FINISH
WITHOUT THE WRITTEN PERMISSION OF NEXT ASSY USED ON
<INSERT COMPANY NAME HERE> IS
PROHIBITED. APPLICATION DO NOT SCALE DRAWING SCALE: 1:2 WEIGHT: SHEET 1 OF 1

2 1
 2 1

.44

R.12

B B
R.12

.31

.55

.24

UNLESS OTHERWISE SPECIFIED: NAME DATE

A DIMENSIONS ARE IN INCHES

TOLERANCES:
FRACTIONAL
DRAWN

CHECKED TITLE:
A
ANGULAR: MACH BEND
ENG APPR.
TWO PLACE DECIMAL
THREE PLACE DECIMAL MFG APPR.

INTERPRET GEOMETRIC Q.A.

PROPRIETARY AND CONFIDENTIAL TOLERANCING PER:
COMMENTS:
THE INFORMATION CONTAINED IN THIS MATERIAL
SIZE DWG. NO. REV

A small Rev A
gear
DRAWING IS THE SOLE PROPERTY OF
<INSERT COMPANY NAME HERE>. ANY
REPRODUCTION IN PART OR AS A WHOLE FINISH
WITHOUT THE WRITTEN PERMISSION OF NEXT ASSY USED ON
<INSERT COMPANY NAME HERE> IS
PROHIBITED. APPLICATION DO NOT SCALE DRAWING SCALE: 2:1 WEIGHT: SHEET 1 OF 1

2 1
 2 1

.35

1.69

B .85
.20
B
1.69

.85 .40

2.38

.26

6.00

.23

UNLESS OTHERWISE SPECIFIED: NAME DATE

A DIMENSIONS ARE IN INCHES

TOLERANCES:
FRACTIONAL
DRAWN

CHECKED TITLE:
A
ANGULAR: MACH BEND
ENG APPR.
TWO PLACE DECIMAL
THREE PLACE DECIMAL MFG APPR.

INTERPRET GEOMETRIC Q.A.

PROPRIETARY AND CONFIDENTIAL TOLERANCING PER:
COMMENTS:
THE INFORMATION CONTAINED IN THIS MATERIAL
SIZE DWG. NO. REV

A
Sensor base Rev A
DRAWING IS THE SOLE PROPERTY OF
<INSERT COMPANY NAME HERE>. ANY
REPRODUCTION IN PART OR AS A WHOLE FINISH
WITHOUT THE WRITTEN PERMISSION OF NEXT ASSY USED ON
<INSERT COMPANY NAME HERE> IS
PROHIBITED. APPLICATION DO NOT SCALE DRAWING SCALE: 1:2 WEIGHT: SHEET 1 OF 1

2 1
 2 1

B B

.25
2.38

.26

.23

5.78
UNLESS OTHERWISE SPECIFIED: NAME DATE

A DIMENSIONS ARE IN INCHES

TOLERANCES:
FRACTIONAL
DRAWN

CHECKED TITLE:
A
ANGULAR: MACH BEND
ENG APPR.
TWO PLACE DECIMAL
THREE PLACE DECIMAL MFG APPR.

INTERPRET GEOMETRIC Q.A.

PROPRIETARY AND CONFIDENTIAL TOLERANCING PER:
COMMENTS:
THE INFORMATION CONTAINED IN THIS MATERIAL
SIZE DWG. NO. REV

TopAand Bottom Plate

DRAWING IS THE SOLE PROPERTY OF
<INSERT COMPANY NAME HERE>. ANY
REPRODUCTION IN PART OR AS A WHOLE FINISH
WITHOUT THE WRITTEN PERMISSION OF NEXT ASSY USED ON
<INSERT COMPANY NAME HERE> IS
PROHIBITED. APPLICATION DO NOT SCALE DRAWING SCALE: 1:2 WEIGHT: SHEET 1 OF 1

2 1
 2 1

B B
.25

2.38

.26
.23
5.78

UNLESS OTHERWISE SPECIFIED: NAME DATE

A DIMENSIONS ARE IN INCHES

TOLERANCES:
FRACTIONAL
DRAWN

CHECKED TITLE:
A
ANGULAR: MACH BEND
ENG APPR.
TWO PLACE DECIMAL
THREE PLACE DECIMAL MFG APPR.

INTERPRET GEOMETRIC Q.A.

PROPRIETARY AND CONFIDENTIAL TOLERANCING PER:
COMMENTS:
THE INFORMATION CONTAINED IN THIS MATERIAL
SIZE DWG. NO. REV

A base Rev A
Top
DRAWING IS THE SOLE PROPERTY OF
<INSERT COMPANY NAME HERE>. ANY
REPRODUCTION IN PART OR AS A WHOLE FINISH
WITHOUT THE WRITTEN PERMISSION OF NEXT ASSY USED ON
<INSERT COMPANY NAME HERE> IS
PROHIBITED. APPLICATION DO NOT SCALE DRAWING SCALE: 1:2 WEIGHT: SHEET 1 OF 1

2 1
Payload Drawings

37
 Appendix G: Airbrake
Airbrake Hardware Bill of Materials Timeline of Flight Events
Item Quantity Time Event
1' long 1" 80/20 pole 3 T+0 Launch
1' x 2' 1/4" plywood sheet 2 T+~7 Burnout Occurs
1' long 1/4" threaded rod 3 T+7.1 Airbreak Initializes
1" long 1/4" bolt 3 T+7.1-24 Airbreak Auto Adjusts to Changing Conditions
Shoulder Screw 1/8" diameter 7/8" long 3
1/8" nylon nut 3
1/8" washer 12
1/4" diameter 1/4" long nylon spacer 6
1/4" lock nut 3
9" long jumper wires 42
1'x1' 1/4" fiberglass sheet 1
1" long 1/4" diameter nylon spacers 18
1/8" bearing 6
3/8" forged eye bolt 1
3/8" lock nut 1
1" long 3/8" bolt 1
3/8" fender washer 6

38
 Acknowledgments
The Cowboy Rocketworks team would like to thank its sponsors: Spirit AeroSystems, OSU Student Government
Association, and OSU CEAT Student Council. Without their financial support and advising, this program wouldn’t
have grown, let alone afford the resources to compete in the 2018 Spaceport America Cup.
An enormous thank you as well to our 100+ donors who contributed to our PhilanthroPete fundraiser with the
OSU Foundation in fall 2017. They helped the team raise $8255, even more than our $8000 goal. Their generosity has
made our participation in the Cup a reality.
The team would also like to thank the technical expertise, advisement, and support of Jamey Jacob, Andy Arena,
Bob Brown, Mark Logan, Alex Cooper, and those at Oklahoma State University who have helped the team achieve
reach this competition.
Finally, the team would like to extend their gratitude and appreciation to ESRA and the Spaceport America for
administering and hosting such an incredible event.

References
1
Canepa, Mark, Modern High-Power Rocketry 2, Trafford Publishing, Victoria, BC, Canada, 2005, pp. 108.
2
OpenRocket, Software Package, Ver. 15.03, Sampo Nickanen, Open Source Software, 2012.
3
SolidWorks Student Edition, Software Package, Dassault Systemes, Velizy,-Villacoublay, France, 2016.

39