See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/301628024

The Design and Organizational Approach to a Student-built Paraffin-Nitrous

Oxide Hybrid Sounding Rocket

Conference Paper · October 2015

DOI: 10.13140/RG.2.1.5053.9924/1

CITATIONS READS

0 2,388

9 authors, including:

Ashis Ghosh Jeremy C. H. Wang

G. B. Pant University of Agriculture and Technology, Pantnagar Ribbit
7 PUBLICATIONS   29 CITATIONS    17 PUBLICATIONS   26 CITATIONS   

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Autonomous aircraft and RPAS (drones) View project

Compressible flows with non-ideal thermodynamics View project

All content following this page was uploaded by Jeremy C. H. Wang on 26 April 2016.

The user has requested enhancement of the downloaded file.

66th International Astronautical Congress, Jerusalem, Israel. Copyright ©2015 by the International Astronautical Federation. All rights reserved.

IAC-15,E2,3-YPVF.4,7,x30369

THE DESIGN AND ORGANIZATIONAL APPROACH TO A STUDENT-BUILT PARAFFIN-NITROUS

OXIDE HYBRID SOUNDING ROCKET

Ashis Ghosh1, Adam De Biasi2, Jeremy Chan-Hao Wang3, Thomas Siu-Hong Leung4, Oleg Petelin5, Eric Jing-Bo
Yang6, Carl Pigeon7, Adrian Typa8, Mari Timmusk9

This paper presents the final design, testing methods and results, and organizational approach of Eos III (also
called Helios I), the University of Toronto Aerospace Team (UTAT) Rocketry Division’s third-generation sounding
rocket. Eos III was developed over a period of 10 months with the goal of delivering a 1.33kg 1U CubeSat scientific
payload to 3km above ground level, as part of the 2015 Intercollegiate Rocket Engineering Competition (IREC).
Because a standard carbon-fibre airframe was used, the design discussion focuses on propulsion, avionics, and payload
instead. The organizational approach is briefly discussed in terms of team structure and community impact.
Eos III was powered by the 8 100Ns (tested so far, with a target of 10 000-Ns) 'Bia III' hybrid rocket engine,
which used a mixture of paraffin-carbon black as fuel and nitrous oxide as the oxidizer. Fuel cartridges and shoulder-
bolted assemblies promoted ease of assembly and enabled multiple consecutive static test fires. Modular avionics
enabled independent system development, simplicity of design, and reparability. The payload contained an inertial
measurement unit, atmospheric sampling and weather-sensing units, and parachute recovery system, all arranged inside
a standard 1U CubeSat. Many of the components, including structures, internal flows and aerodynamics, engine, and
flight performance were simulated through in-house or commercial software. Ground tests validated these predictions.
The Rocketry Division was headed by two individuals (Lead and Chief Designer) and organized into five
subdivisions (Propulsion, Fluids, Avionics, Payload, and Structures). A team of four high school students was also
selected to develop the scientific payload under the mentorship of undergraduate and graduate students. Tandem with
the development of the rocket itself, the Rocketry Division was heavily involved in educating, inspiring, or simply
reaching out to members of the general public, high school students, and aerospace professionals.
Ultimately, careful simulation, strategic resource allocation, efficient organizational structure, and
collaboration with high school students led to a promising with valuable community impact. A limited launch window,
however, prevented completion of the launch procedures at IREC and a reattempt is scheduled for the future.

I. INTRODUCTION
Since the launch of the first man made satellite by
the Soviet Union, satellites have revolutionized our
civilization. For a large portion of human’s history in
space, the focus has been on developing large
multifunctional satellite missions. In recent years, the
global interest in nano (<10 kg) and microsatellites
(<100 kg) has increased. Beginning in 1999, California
Polytechnic State University and Stanford University
developed a standardized approach to nanosatellites
form factor referred to as a CubeSat [1].
Fig. 1: Distribution of Orbital Satellite Mass: 2000-
2009 for 0-10 kg Satellite Class [2]
As seen in Fig. 1, the 1 kg Cubesat is the most
popular due to its cost and short delivery time. Trends
show that small satellite development will continue to
grow and the need for launches will increase [2]. For
the trend to continue, developing services such as
launch services that support the microsatellite
community are needed.
Currently, small satellites rely heavily on piggy-
rides on medium to large launch vehicles such as
India’s PSLV and Russian rockets. The disadvantage
of being a secondary or tertiary payload on a large
rocket is that the availability, scheduling and orbit
parameters depend on the primary payload. Meaning
that preferred orbital locations cannot always be
achieved and compromises have to be made to obtain
a launch opportunity. New launch options dedicated
for small satellite payload would be a valuable service.
The payload and launch vehicle designed by the
University of Toronto focuses on demonstrating a
technology that could offer solutions to the scarcity of
dedicated small satellite launches. Moreover, a hybrid
engine was chosen due its safety and increasing
promise for sounding rocket applications [3].

IAC-15,E2,3-YPVF.4,7,x30369 Page 1 of 11
66th International Astronautical Congress, Jerusalem, Israel. Copyright ©2015 by the International Astronautical Federation. All rights reserved.

Fig. 2: A solid model of the final Eos III rocket design.

II. ROCKET DESIGN simplified to a dynamics problem with a constant

but conservative drag coefficient.
The overall rocket design is discussed in terms of The final design parameters were in
the propulsion subsystem, avionics subsystem, and agreement between the MATLAB results and
payload subsystem. The mission profile was to launch spreadsheet outputs, and are covered in detail in the
a single-stage hybrid rocket to 3km and deploy a remaining subsections. Given that rocket weight
CubeSat 1U at apogee. Fig. 2 shows a solid model of (80 lbm) could not be further lowered before
the final design for Eos III. IREC—mainly due to the technical interest in the
propulsion system despite its inherent weight— the
II.I Propulsion Subsystem best performance predicted by MATLAB is shown
The propulsion subsystem consisted of all in Fig. 3 and features an apogee of 8000 ft (still
mechanical and chemical design considerations qualifying for IREC). The predictions agree with
associated with engine performance. Discussed here existing work demonstrating that optimal
are the MATLAB engine performance suite, oxidizer performance is reached near an oxidizer-fuel ratio
tank, plumbing bay, injector assembly, engine of around 7 [9]. The initial peak in thrust is due to
chamber, and nozzle assembly. an initial high “guess” required for engine pressure,
based on the thermochemical solution for rocket
II.I.I Simulation Tools: MATLAB and Excel inputs used by NASA CEA. The MATLAB tool
To estimate engine performance of proposed was validated by comparison with existing data
designs, a MATLAB engine performance suite was from the Rocketry Division, and with data from the
developed in tandem with a steady-state Excel University of Washington up to 15% error [9].
spreadsheet. Whereas the MATLAB program
provided a transient prediction of key performance
parameters like thrust and apogee via a rudimentary
aerodynamic model, the spreadsheet was a rapid
design tool in which promising designs could later
be inputted into the MATLAB simulation.
Key assumptions in both the MATLAB
simulation and Excel spreadsheet were: (1) the
oxidizer maintained liquid state until it entered the
engine; (2) the fuel core regression could be simply
modelled with the following equation [4]:
𝑛
𝑟̇ = 𝑎𝐺𝑜𝑥
(3) chemical combustion followed assumptions in
NASA Chemical Equilibrium with Applications
[5]; (4) isentropic expansion across the nozzle. The
nitrous oxide thermophysics was modelled
according to work by Fernandez [6]. Difficulties in
estimating Reynolds numbers led to the adoption of
industry practices in constraints for quenching or
blowout, borrowed from Humble [7]. In addition,
experimental or simulated values were used for the
airframe drag and injector discharge coefficient.
The overall analytical modelling and flow of Fig. 3: Predicted performance by MATLAB
calculations for the MATLAB engine followed that Program, including thrust and apogee.
found in work by Genevieve [8], where instead of
invoking NASA CEA, a multivariable polynomial
regression equation was used to speed up
calculations. The rocket trajectory model was

IAC-15,E2,3-YPVF.4,7,x30369 Page 2 of 11
66th International Astronautical Congress, Jerusalem, Israel. Copyright ©2015 by the International Astronautical Federation. All rights reserved.
II.I.II Oxidizer Tank
The custom oxidizer tank was designed to
the following specifications: (1) 9L capacity; (2)
maximum pressure rating of 2,000 psi with an
additional safety factor of 2; (3) maximum outer
diameter 5.4” to fit past the airframe connectors.
The tank was made from 6061-T6 aluminum alloy
with ellipsoid caps welded to the main right
cylindrical body. The preliminary dimensions of
the minimum weight version were calculated
using thin-walled pressure vessel theory, using a
set inner diameter of 4.5 in. to generate the
required wall thicknesses for the caps and the
main body. Due to the ellipsoid geometry of the
caps and the difference in thicknesses between the
caps and the main body of the tank, the minimum
weight design was further refined through finite
element analysis simulations to eliminate stress
concentrators while maintaining minimum weight.
The ends of the tank's main body were modified to
facilitate the welding of the caps. The strength of
the welds were determined by the strength of the
filler metal used in the weld. The filler metal and
welding technique used were decided by the
welder according to the design specifications
provided. The tank was successfully hydro-tested
to 2,000 psi without issue. The tank solid model is
shown below in Fig. 4
Fig. 4: Oxidizer tank solid model, featuring top-view
with main and secondary outlets (right) and the
overall tank (left).
IAC-15,E2,3-YPVF.4,7,x30369
II.I.III Plumbing Bay
In between the oxidizer tank and engine was a
set of plumbing designed to enable diagnostics and
control the flow of oxidizer. There were two
systems running in parallel: (1) main feed system;
(2) safety & monitoring system.
The main feed system used a servo-actuated
1/4” ball valve, mounted on the valve with a 3D
printed actuation mount. ½” Swagelok
compression straights were used as the main feed
line for ease of assembly and minimization of
pressure losses. The stem of the valve and the
servo’s driving shaft were connected by a 3D
printed adapter, which used set screws to hold the
2 parts in place. The ½” straights were adapted to
a ¼” ball valve, which constrained oxidizer flow
but was implemented due to the higher torque and
size limitations with a larger ½” valve.
Safety, monitoring, and venting were achieved
with plumbing connected to a second, smaller
outlet beside the main ½” line. This second outlet
was connected to a coaxial vent line, pressure
transducer, thermocouple, 1100psi pressure relief
valve, and a 1/8” ball valve open to the
environment were included in this secondary
system, with ¼” Swagelok compression fittings for
space management and ease of assembly. The
oxidizer tank was filled by connecting a quick
disconnect fitting to the main nitrous tank, and
closing the 1/8” ball valve when liquid nitrous
entered the coaxial vent lines.
II.I.IV Injector Assembly
Oxidizer injection was designed with the
following major factors in mind: (1) mass flow
rate; (2) pressure drop; (3) oxidizer dispersion. In
the Bia III engine, the injector assembly consisted
of an injector manifold and an injector plate. The
plate diameter was 3.2” as compared to the
minimum 0.25” diameter of the main feeding
system. Sealing was achieved with O-rings placed
at various locations. A parabolic profile was chosen
for the injector manifold because recirculation
region would have been prominent in the case of a
rectangular profile. As well, the parabolic profile of
the manifold allowed radial bolts to be used for
securing the assembly to the engine chamber itself.
A cross section of the assembly is shown below.
Fig. 5: Section view of the injector assembly
Page 3 of 11
66th International Astronautical Congress, Jerusalem, Israel. Copyright ©2015 by the International Astronautical Federation. All rights reserved.

Description Pattern Cold Flow Test CFD Simulation

31 holes, φ=3/32’’
Straight.

23 holes, φ=1/8’
Inner and outer ring counter-
clockwise, at 14°. Middle
ring clockwise, also at 14°.

24 holes, φ=1/8’ Not conducted due to

All holes are at 14° pointing limited resources.
toward the center.

(red-blue:high-low velocity)
Table 1: CFD and cold-flow tests for 3 injectors.

To balance the need for high 𝑚̇𝑜𝑥 (~1kg/s

according to MATLAB) and sufficient pressure
drop (at least 30% against 𝑃𝑐𝑐 = 250psia, to buffer
against instabilities), fine tuning was required for
the size, number and angle of these holes. Initially,
three distinct arrangements were designed and
tested through unstructured k-ω CFD simulations
and cold flow tests using carbon dioxide. The three
designs considered were straight, impinging and
swirl (Table 1). Fig. 6: Final injector plate design.
Results from the cold flow test, although
affected by 𝑚̇𝐶𝑂2 due to limitations of valves on II.I.V Engine Chamber
industrial carbon dioxide cylinder, verified the flow The engine chamber had ¼” aluminum walls
pattern generated using CFD simulations, with an inner diameter of 4.5”. It was divided into
suggesting that the results obtained were fair a 1” long pre-combustor, 12” long fuel core with
approximation of the actual performance. Based on 2.5” single circular port, and a 12” polyurethane-
these flow patterns and engine testing data, a final fiberglass ablative postcombustor liner with ½”
injector was designed to incorporate the dispersive thickness—a more cost-effective alternative to
and azimuthal flow properties of the first and industry-grade liners. Ignition was achieved
second designs (Fig. 6). through three smaller B-size solid rocket motors
impregnated in the top of the fuel core.

IAC-15,E2,3-YPVF.4,7,x30369 Page 4 of 11
66th International Astronautical Congress, Jerusalem, Israel. Copyright ©2015 by the International Astronautical Federation. All rights reserved.
A standard 1” pre- combustor was selected
based on good practices by Sutton [10].
The fuel core dimensions were the result of
simulations from the MATLAB performance suite,
discussed above. A small amount of carbon black
(2.5g/kg of fuel) was added to opacify the fuel and
protect the engine from radiation heating—the
effect on combustion performance was deemed
negligible. Fuel cores were poured into ½” thick
cardboard tubes, with a PVC inner mandrill for the
port diameter. These cartridges could be loaded and
removed from the engine to expedite testing and
transportation.
The length of the postcombustor was chosen to
allow enough time for mixing and combustion, as
approximated with characteristic combustor length
[11]. The postcombustor thickness was determined
through informal small scale testing using a 2”
inner diameter steel pipe as the stand-in
combustion chamber and observing the remaining
thickness of postcombustor after various tests.
II.I.VI Nozzle Assembly
The nozzle assembly was also held in with
radial shoulder-bolts and utilized a graphite conical
nozzle due to cost-effectiveness and machining
simplicity. The area ratio was 4.2 with a half-angle
of 12 degrees. Due to the high thermal conductivity
of graphite, a steel and not aluminum backing plate
was used to hold the nozzle in place. One-
dimensional thermal calculations assuming a
worst-case scenario of stagnation temperature
(3000K) at the nozzle inner wall showed that the
steel would not exceed melting temperatures. The
nozzle assembly is shown in (Fig. 7).
Fig. 7: The nozzle assembly, demonstrating the
shoulder-bolted design with graphite nozzle.
II.II Avionics Subsystem
The purpose of the avionics system was three-fold:
(1) to provide sensor (telemetry) data about the
rocket’s internal state during both flight and static
ground testing; (2) to communicate with a remote
ground station and initiate the launch sequence by
actuating and igniting the flow of oxidizer; (3) to
deploy the payload and drogue parachute when apogee
IAC-15,E2,3-YPVF.4,7,x30369
is reached and to deploy the main parachute when the
rocket is descending and is almost at ground level.
This section describes parts of the avionics
hardware and software system that were necessary to
safely launch and recover the rocket. At a high level
the avionics hardware was split between the ground
station and the avionics bay. The remote ground station
received telemetry data from the rocket via a wireless
link and sent ignition commands to the rocket via the
wireless and hard physical links. The avionics bay
contained all the sensors, power supplies and
processing devices necessary for static fire testing and
for launching the rocket. A half-page systems diagram
is depicted in Fig. 9 on the subsequent page.
II.II.I Power Distribution and Motherboard
The avionics bay had two power sources: (1)
a 12V 2000mAh battery pack providing 12V, 5V
and 3.3V to the motherboard where 5V and 3.3V
are derived from the 12V through switching buck
converters (150kHz) on the power board (Fig. 9);
(2) a 9V battery provided power to the redundant
parachute deployment system—the Raven 3
commercial altimeter (which was a redundant
system for parachute and payload deployment).
As shown in Fig. 8, the power board supplied
12V, 5V and 3.3V to the motherboard which
distributed these voltages to all the daughterboards.
Note that the 12V, 5V and 3.3V rails had their own
dedicated ground—this helped mitigate ground-
bounce noise seen by devices when high current
was being drawn through another supply rail (i.e.
when oxidizer valve motor draws 5A on the 12V
rail the “ground” reference jumped more
significantly on the 12V rail than the other rails).
Fig. 8 also shows the I2C bus which provided a
communication interface between the
daughterboards. The I2C bus was used in a single-
master multiple-slave configuration with the arbiter
daughterboard acting as the master.
Fig. 8: Power board-motherboard interface.
Page 5 of 11
66th International Astronautical Congress, Jerusalem, Israel. Copyright ©2015 by the International Astronautical Federation. All rights reserved.

Fig. 9: Avionics systems diagram.

II.II.II Common Daughterboard Hardware connection if the microcontroller was blank and did
All daughterboards in the avionics system had not have a bootloader.
a common interface to the motherboard. As shown
in Fig. 10, each daughterboard connected to the
motherboard via a standard header. The header
supplied 12V, 5V and 3.3V, and connected the
daughterboard to the shared I2C bus.
A USB-to-UART (RS232) bridge allowed the
Atmel Atmega328p microcontroller to connect to a
PC and send/receive data via serial. Programming
the microcontroller could be done in one of two
ways: (1) through the USB connection if a
bootloader was present on the microcontroller; (2)
through the AVRISP2 six-pin header (standard Fig. 10: Common daughterboard hardware.
Atmel programming interface) or through the SPI

IAC-15,E2,3-YPVF.4,7,x30369 Page 6 of 11
66th International Astronautical Congress, Jerusalem, Israel. Copyright ©2015 by the International Astronautical Federation. All rights reserved.

II.II.III Oxidizer Actuation

The oxidizer actuation system was composed of
an H-bridge (TLE5206) and the relevant pins used
to interface with the encoder and limit switches of
the oxidizer actuation housing. The encoder pin
was fed into the microcontroller’s timer/counter
input, which enabled counting of the square-wave
pulses of the encoder without the use of interrupts.
The switches were fed directly into GPIO pins of
the microcontroller.

II.II.IV Pyrotechnic Actuation

The ignition circuit was an implementation of a
‘firing circuit’, a custom modular-design solid-
state pyrotechnic actuation circuit (Fig. 11). The
circuit had two different MOSFETs that controlled
the current flow, labelled ‘arm’ and ‘fire’. When
the circuit was armed (the ‘arm’ FET switched into
saturation mode) the e-match attached in series was
energized to 9V, and when the ‘fire’ FET was
switched on also, sufficient current flowed in order
to activate the e-match. The circuit was designed Fig. 11: Pyrotechnic actuation circuit (i.e. firing
such that before the arming, there was no voltage circuit for test or launch).
on the e-match, and therefore no risk of a short
circuit. The reason behind the solid-state
construction was that relays (typically used for
pyrotechnic actuation) are susceptible to launch
vibrations. An addition feature of the firing circuit
was continuity detection across the e-match, which
allowed for the detection of incorrectly attached e-
matches. The circuit was designed to be modular,
and could be implemented in such a way that Fig. 12: Temperature signal amplification circuit.
multiple channels of e-matches were armed
together and fired separately. Each channel could, II.III Payload
in turn, fire three e-matches. The objective was to demonstrate that it is
technically feasible to deploy a small satellite from a
II.II.V Sensors small rocket. The system was made for small diameter
The pressure transducers present were 1000 psi rockets capable of suborbital flight. It would be
models that fed their data output in the form of a feasible to adapt the system to medium sized rockets
current between 4-20mA full scale. In order to capable of achieving orbital mission or air launch
convert that to a voltage that the microcontroller’s missiles carried to high altitudes by aircraft before
on-board ADC could read, a noninverting amplifier being launched. The flexibility and versatility of this
op-amp circuit was used, ultimately amplifying the payload aim to open conversation and possibilities of
signal to a range of 0.8~4V. commercial adaptation of the technology.
The thermocouple amplifier used was Key requirements of the CubeSat included: (1)
MAX31855, which interfaced with the mass under 1.33kg; (2) conformation to the 1U
microcontroller over the SPI protocol (Fig. 12). CubeSat form factor; (3) presence of a remove-before-
This particular chip performed cold-junction flight pin and inactivation of all electronics when
compensation, enabling the use of a thermocouple engaged; (4) non-transmissions and no power while
without a constant temperature junction on the
other side. As well, since the thermocouple
amplifier also had an internal temperature sensor, it
could be used to sense the temperature of the
interior of the avionics bay.

IAC-15,E2,3-YPVF.4,7,x30369 Page 7 of 11
66th International Astronautical Congress, Jerusalem, Israel. Copyright ©2015 by the International Astronautical Federation. All rights reserved.
Fig. 13: Deployment system, cross-sectional view.
stowed in the deployment bay; (5) inclusion of a
parachute in the design, to achieve a descent rate of
3.5m/s to 4.5m/s
II.III.I CubeSat Contents
The final design included a stowed parachute as
shown in Fig. 14 and an electronics compartment
which enabled the scientific aspect of the mission.
An Arduino Uno was incorporated, with weather
sensors for the measurement of atmospheric
temperature, pressure, humidity, acceleration and
rotational rates. A static line and deployment bag
enabled the parachute to be deployed upon exiting
the deployment bay. The deployment bag would
remain with the rocket once the parachute was been
deployed. The recovery of the CubeSat would be
done by radio tracking with one of IREC’s radios.
Fig. 14: The completed CubeSat 1U.
IAC-15,E2,3-YPVF.4,7,x30369
II.III.II Deployment
The CubeSat is deployed perpendicular to the
axial direction of the rocket, via four springs
located on the top and bottom of the satellite (Fig.
13). The springs were attached to an aluminum
back plate which the satellite is held up against in
compression when the door is in place. The door is
kept closed by two explosive bolts. Once apogee is
reached, the explosive bolts simultaneously
detonate thus releasing the stored energy of the
springs and ejecting the satellite.
Horizontal deployment was primarily done to
not interfere with the parachute bay stowed at the
nose cone. Additionally, the sideways deployment
meant that a number of these deployment systems
could be stacked onto each other in order to deploy
multiply CubeSats independently aboard a single
rocket.
II.III.III Structure
The structure of the payload bay was made from
6061 aluminum with a carbon-fibre shell for
aerodynamic streamlining. Rails machined into the
payload guide the CubeSat out of the bay. UTAT
female bay connector grooves were added for quick
integration with the rest of the rocket.
II.III.IV Camera Module
A camera module was an additional payload
carried on-board the rocket to serve as a means of
acquiring photographic evidence confirming the
deployment of the CubeSat at apogee. The camera
module used a GoPro camera system remotely
operated from the ground station. The data was
stored locally and could be viewed once the rocket
Page 8 of 11
66th International Astronautical Congress, Jerusalem, Israel. Copyright ©2015 by the International Astronautical Federation. All rights reserved.
has been recovered. Since the camera module was
connected with a standard UTAT bay connector to
the deployment system payload bay, the camera
module was optional and could be reused in future
years as a stand-alone unit.
III. ORGANIZATIONAL APPROACH
The organizational approach to Eos III
development may be discussed in terms of: (1) team
leadership; (2) key projects and division of labour; (3)
timeline; (4) community involvement.
Rocketry Lead and Chief Designer had already
spent two years with the Rocketry Division. The
former handled administrative elements and overall
project management. The latter handled systems
integration and providing design advice—especially in
terms of general mechanics and aspects of manufacture
or assembly—to the subsystem leads.
By contrast, the subsystem leads were responsible
for projects relevant to their expertise. Propulsion was
chiefly responsible for the oxidizer tank, plumbing,
and engine mechanical design as well as combustion
performance. Avionics was responsible for the on-
board sensors, microcontrollers, and effectors (e.g.
firing circuit for the engine), as well as the ground
station. Structures was responsible for the airframe and
recovery system (not discussed here due to standard
approaches) and advising propulsion structures. Fluid
mechanics was responsible for designing the injector
plate, nosecone, and fins. Payload was responsible for
the deployment bay and actual scientific
instrumentation, namely the CubeSat 1U. The test
facility was a collaborative effort between all
subsystems. Following this matrix organizational
structure, the team could harness the technical
knowledge of its members to accomplish the projects.
Of the 10-month period allotted for development,
testing, and integration, the first three months were
spent setting high-level requirements, performing
conceptual design, and developing or becoming
familiar with simulation tools. At that point,
approximately 40 new members joined the team as part
of recruitment activities taking place at the start of the
new Academic Year. These new members were
integrated into the team through ‘beginner’ projects
such as simple mechanical designs or fabrication tasks.
New members were also encouraged to attend the
University of Toronto Aerospace Team’s general
aerospace seminar series which covered topics ranging
from solid modelling to aircraft and spacecraft
electronics to machining and fabrication. The
remaining seven months witnessed two 3-month
engine testing phases, the first phase involving minor
iterations on the initial design and the second phase
allowing for major changes relative to the initial
concept. The schedule primarily motivated by engine
IAC-15,E2,3-YPVF.4,7,x30369
development—the most challenging technical
element—the other subsystems planned accordingly to
aim for integration at T-2 months before the
competition. Delays in manufacture and further
required engine testing resulted in integration taking
place at T-1 month before competition. This prevented
the possibility of any pre-competition launch attempt,
during which if a severe failure occurred there might
not be enough time to repair systems for the
competition.
Major bottlenecks throughout this time were: (1)
manufacture limitations as the entire rocket was
designed, built, and tested by students (i.e. all
mechanical parts machined or wet-laid-up by
students); (2) design and debugging of avionics
systems, which was typically the last to begin of any
subsystem because it could be done only after the
mechanical parts they controlled/were housed
in/sensed from were specified.
Lastly, being a student-led initiative, a secondary
motivation of this project was to educate and inspire
members of the local and international community.
Throughout the design and testing of Eos III, the team
interacted with members of the public and high school
students at events such as Science Rendezvous, guest
lectures at local secondary schools, appearances at
conferences (e.g. International Space Development
Conference), on-campus design showcases, university
applicant events or open-houses, and talks at
organizations affiliated with students or the team at
large, such as the German Aerospace Center (DLR).
These opportunities provided an easy way to engage
others requiring minimal preparation aside from
bringing items and multimedia for display and
explaining concepts well. It is estimated that the team
was able to directly interact (i.e. speak with) just under
1000 individuals across these settings.
IV. MAJOR TESTS AND RESULTS
This section discusses the end results of engine
testing, as well as qualitative challenges with avionics
and launch operations at IREC. Overall, the Bia III
engine could provide a maximum thrust of 280-lbf
and an average of 200-lbf over the course of the 9-
second burn. This is approximately 2/3 the thrust of,
and 4 seconds longer than, the MATLAB prediction.
The primary suspected reason is low oxidizer flow
rates due to choking at main valve in the plumbing.
IV.I Engine Testing
A number of static test fires were conducted at a
student-constructed, inverted-engine static test fire
facility at the University of Toronto Institute for
Aerospace Studies. A photo from the last test
conducted is shown in Fig. 15, along with engine thrust
and pressure data.
Page 9 of 11
 66th International Astronautical Congress, Jerusalem, Israel. Copyright ©2015 by the International Astronautical Federation. All rights reserved.

extended burn time the presence of more leftover fuel

than expected add to the pool of evidence supporting
this hypothesis. In total, this resulted in a total impulse
of only 8100Ns as opposed to the 10 000Ns originally
predicted, but it is strongly suspected that 10kNs if not
more is possible given future refinements and the
performance of engines of similar scale [9].

IV.II Avionics Challenges and Testing

Throughout prototyping, it was found that the
oxidizer actuation’s microcontroller was not very
consistent at reading the states of the switches. When
observed using an oscilloscope, it was found that the
switch lines had a considerable amount of noise. This
was attributed to the fact that the high frequency pulses
of the square-wave encoder were adjacent to the switch
lines. However, a solution was found in delaying the
switch polling code, perhaps minimizing interference
between the lines.
Thrust and Pressure vs. Time One potential concern with the pressure transducer
circuit was its temperature stability, as the resistors
1200 used in this analog circuit were temperature dependent
net force (N)
Thrust (N) or Pressure (psia)

and therefore the readings could change with changing

1000 PSI ambient temperatures. However, during multiple fire
tests in subzero and standard room temperatures, the
800
circuit continued to provide reliable pressure data once
600 re-calibration was performed.
Overall the modularity of the avionics system came
400 at the expense of complexity and whether the system
will provide its promised long-term value will depend
200 on future attempts to ‘evolve’ the system instead of
replacing it entirely.
0
0

10.8
11.7
12.6
13.5
14.4
15.3
0.9
1.8
2.7
3.6
4.5
5.4
6.3
7.2
8.1

9.9

IV.III IREC Performance

Time (s)
Fig. 15: Photo of last engine test on the Bia III engine
noting the perfectly expanded nozzle gases (top);
thrust and pressure versus time graph (bottom).
Smoke near the bottom of the engine (top picture)
was from e-match cables spit out of the engine.
It is evident from the thrust curve that there is a
rapid increase followed by a relatively steep and then
even and shallower decrease until the engine burn
concludes at 9 seconds. Compared with the MATLAB
predictions of Fig. 3, the experimental thrust curve
bears similar shape but with 1/3 less thrust. The
pressure, too, is consistently about 1/3 less than that
predicted by MATLAB.
The main suspected reason for this is that the ¼”
main valve, which was the largest valve that could be
accommodated, was likely choking the flow and
limiting the usefulness of the otherwise ½”
components in the feeding system. Furthermore, the
A limited launch window resulted in a delay of the
original launch of the Eos III due to the time required
for launch preparation and on-site debugging. The
rocket has not yet flown and in the future UTAT will
seek to make improvements to the design before
reattempting launch before July 2016. Pre-flight
ground testing did show however that the payload
deployment mechanism was successful, and that the
CubeSat parachute packaging was conducive to
opening up when dropped from a tall building.
V. CONCLUSION & NEXT STEPS
A promising design has been designed and ground
tested, pending future launch tests. The MATLAB/
Excel performance suite demonstrated accuracy as a
design and prediction tool. During ground tests, the Bia
III underperformed, providing only 8100kNs of the
target 10kNs, but this was largely attributed to
restrictive oxidizer plumbing that will be changed. The
avionics system was able to support rigorous test
campaigns, and was successfully implemented despite
its modularity leading to more complexity than

IAC-15,E2,3-YPVF.4,7,x30369 Page 10 of 11
 66th International Astronautical Congress, Jerusalem, Israel. Copyright ©2015 by the International Astronautical Federation. All rights reserved.

traditional avionics systems. The payload deployment [2] D. DePasquale and A. C. Charania, "Analysis
was effective in ground tests but flight qualification is of the Earth-to-Orbit Launch Market for
needed along with integrated qualification of Eos III. Nano and Microsatellites," AIAA, Anaheim
In the future, the current valves and piping will be CA, 2010.
replaced with a burst disk system for simplicity, [3] E. Doran, J. Dyer, K. Lohner, Z. Dunn, M.
weight-saving, and increased oxidizer mass flow rates. Marzoña and E. Karlik, "Peregrine Sounding
Flight telemetry may also be added to the avionics Rocket," Stanford University, Stanford, CA,
system to supply additional data regarding the rocket 2008.
trajectory and engine status. High school students may
work with the team again for community impact, but [4] G. Ziliac and M. A. Karabeyoglu, "Hybrid
the Rocketry Division will likely partner with another Rocket Fuel Regression Rate Data and
design team at the University of Toronto to develop a Modelling," AIAA, Sacramento CA, 2006.
more complex payload. Lastly, a flight will be [5] NASA, "Chemical Equililbrium with
attempted either in Canada or the United States, not Applications (CEA)," NASA, Cleveland,
only to demonstrate Eos III’s capabilities but to serve OH, 2014.
as a simulated launch operations sequence for the [6] M. Fernandez, "Propellant tank
Rocketry Division. If a successful payload deployment pressurization modelling for a hybrid
at 3km is achieved, then the Division will continue rocket," Rochester Institute of Technology,
increasing the target altitude and stepping up engine Rochester NY, 2009.
performance with it. [7] R. W. Humble, G. H. Henry and W. J.
Larson, "Space Propulsion Analysis and
VI. ACKNOWLEDGEMENTS
design," McGraw-HIll, New York, NY,
To the University of Toronto Division of
1995.
Engineering Science, Institute for Aerospace Studies,
Department of Electrical & Computer Engineering, [8] B. Genevieve, M. Brooks, P. Beaujardiere
Department of Mechanical & Industrial Engineering, and L. Roberts, "Performance Modeling of a
and Engineering Society, for funding or otherwise Paraffin Wax / Nitrous Oxide Hybrid
supporting the bulk of this project; to past and present Motor," AIAA, Orlando, 2011.
members of the University of Toronto Aerospace [9] T. Edwards, V. Hansen, T. Slais, C. Chu, M.
Team’s Powered Flight, UAV, Space Systems, and Hughes, G. Li, G. Finnegan, T. Ip, A. Hatt
Outreach Divisions for their advice and moral support; and B. Degang, "University of Washington
to Adam Paul Trumpour, for readily critiquing and DAQ Destroyer Hybrid Rocket," Seattle,
having assisted with cold-flow testing; finally, to the 2012.
numerous sponsors and partners who shall go unnamed [10] G. Sutton and O. Biblarz, Rocket Propulsion
in writing but were instrumental in enabling the team’s Elements, New York City, NY: Wiley, 2010.
hybrid rocket programme.
[11] B. T. C. Zandbergen, "Hybrid Rocket
Motors," Delft University of Technology,
VII. REFERENCES
Delft, 1999.
[1] R. Nugent, R. Munakata, A. Chin, R. Coelho
and J. Puig-Sairi, "The CubeSat: The
Picosatellite Standard for Research and
Education," AIAA, San Diego CA, 2008.

1 University of Toronto Mechanical & Industrial Engineering, Canada, ashisghosh@live.com

2 University of Toronto Mechanical & Industrial Engineering, Canada, aa.debiasi@mail.utoronto.ca
3 University of Toronto Engineering Science, Canada, jer.wang@mail.utoronto.ca
4 University of Toronto Engineering Science, Canada, siuhong.leung@mail.utoronto.ca
5 University of Toronto Electrical & Computer Engineering, oleg.petelin@mail.utoronto.ca
6 University of Toronto Engineering Science, Canada, jingbo.yang@mail.utoronto.ca
7 University of Toronto Institute for Aerospace Studies, Canada, carl.pigeon@gmail.com
8 University of Toronto Mechanical & Industrial Engineering, Canada, adrian.typa@mail.utoronto.ca
9 University of Toronto Mechanical & Industrial Engineering, Canada, mari.timmusk@mail.utoronto.ca

IAC-15,E2,3-YPVF.4,7,x30369 Page 11 of 11

View publication stats