Design, Construction and Testing Of A Remote Operation ... · PDF fileHeavy-Lift Model...
87
Design, Construction and Testing Of A Remote Operation Heavy-Lift Model Aircraft Capstone Final Report University of Maine SAE Aero Design: Benjamin Waller, MEE David Chandpen, MEE Joseph Travaglini, MEE Matthew Maberry, MEE Travis Cushman, MEE Zachary Veilleux, MEE Mechanical Engineering Class of 2013 Original: 07 MAY 2013 Reformatted: FEB 2015
Transcript of Design, Construction and Testing Of A Remote Operation ... · PDF fileHeavy-Lift Model...
Design, Construction and Testing Of A Remote Operation
Heavy-Lift Model Aircraft
Capstone Final Report
University of Maine
SAE Aero Design:
Benjamin Waller, MEE
David Chandpen, MEE
Joseph Travaglini, MEE
Matthew Maberry, MEE
Travis Cushman, MEE
Zachary Veilleux, MEE
Mechanical Engineering
Class of 2013
Original: 07 MAY 2013
Reformatted:
FEB 2015
- 2 -
Abstract
The UMaine SAE Aero Capstone group is designed around an annual competition held by the Society of
Automotive Engineers. The main goal of the competition is to design and construct an aircraft that can lift
more payload than other teams while staying within the rules and restrictions of the SAE Aero Competition
guidelines. This capstone group represents the first iteration of this challenge attempted by the University
of Maine. This team did not attend the competition but rather the aim was to design and build an aircraft
that was competition worthy. The following report is a summary of the design process, decisions, analysis,
substantiation, and final results of the project.
- 3 -
1. Table of Contents
1. Table of Contents ............................................................................................................................... - 3 -
2. Table of Figures ................................................................................................................................. - 6 -
3. Contributions ..................................................................................................................................... - 8 -
4. Introduction ....................................................................................................................................... - 9 -
1.1 Project Beginnings .................................................................................................................... - 9 -
1.2 SAE Aero Design Competition ................................................................................................. - 9 -
1.3 MEE Capstone Project and Associated Opportunities .............................................................. - 9 -
1.4 Aerospace Studies at the University of Maine ........................................................................ - 10 -
1.5 Summary of Focus and Objectives ......................................................................................... - 10 -
5. Design Description .......................................................................................................................... - 11 -
2.1 Components and Systems ....................................................................................................... - 11 -
2.2 Engine ..................................................................................................................................... - 11 -
2.3 Wing........................................................................................................................................ - 11 -
2.4 Fuselage .................................................................................................................................. - 12 -
2.5 Wireless Systems .................................................................................................................... - 12 -
2.6 Controls ................................................................................................................................... - 13 -
2.7 Empennage .............................................................................................................................. - 13 -
2.8 Landing Gear / Externals ........................................................................................................ - 13 -
6. Design Concept Process................................................................................................................... - 14 -
3.1 Engine Selection ..................................................................................................................... - 14 -
3.2 Airfoil Selection ...................................................................................................................... - 14 -
3.2.1 Design Goal .................................................................................................................... - 14 -
3.2.2 Research Done ............................................................................................................... - 14 -
3.2.3 Operating Conditions ..................................................................................................... - 14 -
3.2.4 Resources Utilized.......................................................................................................... - 15 -
3.3 Testing .................................................................................................................................... - 15 -
3.4 Results Discussion .................................................................................................................. - 16 -
7. Wing Configuration ......................................................................................................................... - 17 -
4.1 Wing Design ........................................................................................................................... - 17 -
4.2 Wing Section ........................................................................................................................... - 17 -
4.3 Design Goal ............................................................................................................................ - 18 -
4.4 Research Done ........................................................................................................................ - 18 -
4.4.1 Wing Span ...................................................................................................................... - 18 -
4.4.2 Wing Span with Reynolds Number ................................................................................ - 18 -
4.4.3 Chord Length ................................................................................................................. - 20 -
4.4.4 Aspect Ratio ................................................................................................................... - 21 -
4.4.5 Wing Mounting Style ..................................................................................................... - 21 -
- 4 -
4.4.6 Dihedral Angle ............................................................................................................... - 22 -
4.4.7 Taper .............................................................................................................................. - 23 -
4.4.8 Twist ............................................................................................................................... - 23 -
4.4.9 Wing Sweep ................................................................................................................... - 23 -
4.4.10 Edge Design ................................................................................................................... - 23 -
4.4.11 Wing Loading ................................................................................................................. - 24 -
4.4.12 Externals ......................................................................................................................... - 24 -
4.4.13 Multiple Wings ............................................................................................................... - 24 -
4.4.14 Final Wing Size and Configuration ................................................................................ - 25 -
4.4.15 Construction and Components ....................................................................................... - 26 -
8. Website ............................................................................................................................................ - 26 -
9. Preliminary Modeling ...................................................................................................................... - 26 -
10. Fuselage ....................................................................................................................................... - 27 -
7.1 Sizing ...................................................................................................................................... - 27 -
7.2 Wing-Fuselage Connection ..................................................................................................... - 27 -
7.3 Empennage-Fuselage Connection ........................................................................................... - 28 -
7.4 Motor-Mount........................................................................................................................... - 28 -
7.5 Payload and Maintenance Accessibility .................................................................................. - 29 -
7.6 Fuselage Construction Process ................................................................................................ - 29 -
11. Empennage .................................................................................................................................. - 30 -
8.1 Tail Configuration Selection ................................................................................................... - 30 -
8.2 Horizontal Stabilizer Airfoil Selection .................................................................................... - 30 -
8.3 Horizontal and Vertical Stabilizer Sizing................................................................................ - 30 -
8.4 Empennage Position Relative to Wings .................................................................................. - 31 -
8.5 Incidence Angle ...................................................................................................................... - 31 -
12. Preliminary Testing ..................................................................................................................... - 32 -
13. Design Analysis and Review ....................................................................................................... - 32 -
10.1 Wing........................................................................................................................................ - 32 -
10.2 Wing Structural Analysis ........................................................................................................ - 39 -
10.2.1 Analysis With Design 2 .................................................................................................. - 39 -
10.3 Empennage .............................................................................................................................. - 40 -
10.4 Servo Performance .................................................................................................................. - 40 -
10.5 Weight ..................................................................................................................................... - 40 -
10.5.1 Predicted Performance ................................................................................................... - 40 -
14. Final Testing and Evaluation ....................................................................................................... - 41 -
11.1 Testing .................................................................................................................................... - 41 -
11.2 Results ..................................................................................................................................... - 42 -
11.3 Evaluation ............................................................................................................................... - 42 -
11.3.1 Engine ............................................................................................................................ - 42 -
- 5 -
11.3.2 Wings ............................................................................................................................. - 42 -
11.3.3 Fuselage.......................................................................................................................... - 42 -
11.3.4 Wireless Systems............................................................................................................ - 43 -
11.3.5 Controls .......................................................................................................................... - 43 -
11.3.6 Empennage ..................................................................................................................... - 43 -
11.3.7 Landing Gear/Externals .................................................................................................. - 43 -
11.4 Evaluation Summary ............................................................................................................... - 43 -
11.5 Conclusion .............................................................................................................................. - 43 -
15. References ................................................................................................................................... - 46 -
16. Appendices .................................................................................................................................. - 49 -
13.1 Appendix A - Plans and Specs ................................................................................................ - 49 -
13.2 Appendix B - Wind Tunnel Modification and Testing ........................................................... - 50 -
13.2.1 Design Goals .................................................................................................................. - 50 -
13.2.2 Research ......................................................................................................................... - 50 -
13.2.3 Design Process ............................................................................................................... - 52 -
13.2.4 Abandonment of Modification Idea ............................................................................... - 53 -
13.2.5 Final Results ................................................................................................................... - 53 -
13.3 Appendix C - Budget and Costs .............................................................................................. - 54 -
13.4 Appendix D - Mech. Lab III (MEE 443) Report ..................................................................... - 56 -
13.4.1 Introduction .................................................................................................................... - 58 -
13.4.2 Experimental Objectives ................................................................................................ - 59 -
13.4.3 Apparatus, Equipment, and Instrumentation .................................................................. - 60 -
13.4.4 Experimental Theory ...................................................................................................... - 62 -
13.4.5 Experimental Procedure ................................................................................................. - 64 -
13.4.6 Experimental Results and Conclusions .......................................................................... - 67 -
13.4.7 Mech. Lab Appendix A: Uncertainty Level Buildup ..................................................... - 73 -
13.4.8 Mech. Lab Appendix B: Beam Calibration .................................................................... - 77 -
13.5 Appendix E - Flight Simulation Code ..................................................................................... - 78 -
13.6 Appendix F - Servo Performance MathCAD Worksheet ........................................................ - 80 -
13.7 Appendix G - Engine Specs .................................................................................................... - 82 -
13.8 Appendix H – Project Timeline .............................................................................................. - 85 -
13.9 Appendix I – Team Photograph .............................................................................................. - 86 -
- 6 -
2. Table of Figures
Figure 1 - Finished Aircraft ...................................................................................................................... - 11 -
Figure 2 - Magnum XLS 0.61A ................................................................................................................ - 11 -
Figure 3 - Wing End View ....................................................................................................................... - 12 -
Figure 4 - Perspective View of Wings ...................................................................................................... - 12 -
Figure 5 - Final Fuselage; FWD End ........................................................................................................ - 12 -
Figure 6 - Fuselage FWD End with Hatch Open ...................................................................................... - 12 -
Figure 7 - Transmitter and Receiver ......................................................................................................... - 12 -
Figure 8 - Servo Illustration...................................................................................................................... - 13 -
Figure 9 - Adjustable Angle of Incidence ................................................................................................. - 13 -
Figure 10 - Completed Empennage .......................................................................................................... - 13 -
Figure 11 - Ground Stance ....................................................................................................................... - 13 -
Figure 12 - Example Cl and L/D Polars ................................................................................................... - 16 -
Figure 13 - Example Pressure Distribution .............................................................................................. - 16 -
Figure 14 - Example Data Spreadsheet .................................................................................................... - 17 -
Figure 15 - Relative Drag Contributions .................................................................................................. - 19 -
Figure 16 - Induced Drag Schematic ........................................................................................................ - 19 -
Figure 17 - Clmax vs AOA for Various AR ............................................................................................. - 20 -
Figure 18 - Stall Onset for Taper Design .................................................................................................. - 22 -
Figure 19 - Wing Edge Desing Variation ................................................................................................. - 24 -
Figure 20 - Lift Distribution Illustration ................................................................................................... - 25 -
Figure 21 - Wing SolidWorks Iso View ................................................................................................... - 27 -
Figure 22 - Wing SolidWorks End View ................................................................................................. - 27 -
Figure 23 - Hand-Built Wing Mockup ..................................................................................................... - 27 -
Figure 24 - Fuselage Spacing Model ........................................................................................................ - 27 -
Figure 25 - Aluminum Sheath w/ Dihedral .............................................................................................. - 28 -
Figure 26 - Wing Mounting Method ........................................................................................................ - 28 -
Figure 27 - Fuselage Tail Boom ............................................................................................................... - 28 -
Figure 28 - Fuselage / Tail Boom Interface .............................................................................................. - 28 -
Figure 29 - Motor Mount .......................................................................................................................... - 29 -
Figure 30 - Payload Bay Open ................................................................................................................. - 29 -
Figure 31 - Payload Bay Closed ............................................................................................................... - 29 -
Figure 32 - Wing Rib Illustration ............................................................................................................. - 29 -
Figure 33 - Empennage Diagram .............................................................................................................. - 30 -
Figure 34 - Tail Angle Adjustability ........................................................................................................ - 31 -
Figure 35 - Thrust Data Comparison ........................................................................................................ - 32 -
Figure 36- General Plan FBD ................................................................................................................... - 33 -
Figure 37 - Takeoff Simulink Schematic .................................................................................................. - 35 -
Figure 38 - Simulink Results .................................................................................................................... - 35 -
Figure 39 - Velocity Streamlines .............................................................................................................. - 36 -
Figure 40 - Boundary Layer ..................................................................................................................... - 36 -
Figure 41 - Velocity Contours .................................................................................................................. - 37 -
Figure 42 - Pressure Contours .................................................................................................................. - 37 -
Figure 43 - Turbulence ............................................................................................................................. - 38 -
Figure 44 - Airfoil Mesh. .......................................................................................................................... - 38 -
Figure 45 - FEA 3 in 1 Wing Deflection .................................................................................................. - 40 -
Figure 46 - Emergency Repairs ................................................................................................................ - 41 -
Figure 47 - Off-kilter empennage ............................................................................................................. - 42 -
- 7 -
Figure 48 - Broken Landing Gear ............................................................................................................. - 42 -
Figure 49 - Current Wind Tunnel Flows .................................................................................................. - 50 -
Figure 50 - Flow Straightening Proposal .................................................................................................. - 51 -
Figure 51 - Flow Mapping Rig ................................................................................................................. - 51 -
Figure 52 - Velocity Contour at Outlet ..................................................................................................... - 52 -
Figure 53 - Flow Streamlines Through Duct ............................................................................................ - 53 -
Figure 54 - Pressure Contours Along Duct Wall ...................................................................................... - 53 -
Figure 55 - Elevation View of Experimental Setup .................................................................................. - 61 -
Figure 56 - Plan View of Experimental Setup .......................................................................................... - 62 -
Figure 57 - LabView VI Schematic .......................................................................................................... - 66 -
Figure 58 - Data Table and Plot of Static Thrust Coefficient v. RPM for APC 11x7 .............................. - 68 -
Figure 59 - Theoretical Propeller Thrust vs. RPM (APC 11x7) ............................................................... - 69 -
Figure 60 - EVO 11x7 Static Thrust vs. RPM .......................................................................................... - 70 -
Figure 61 - EVO 12x6 Static Thrust vs. RPM .......................................................................................... - 70 -
Figure 62 - K Series Master Airscrew Static Thrust vs. RPM .................................................................. - 71 -
Figure 63 - Propeller Trendline Comparison ............................................................................................ - 72 -
Figure 64 - Calibration Data Plot ............................................................................................................. - 77 -
- 8 -
3. Contributions
David Chandpen (Lead)
Wing Design, SolidWorks of the wing, assisted in other SolidWorks parts, assisted in
airfoil selection, thrust testing, project poster, construction of the aircraft.
Zachary Veilleux
CFD Analysis, Strength Analysis, LabVIEW VI Design, Mech Lab Electronics Design &
Fabrication, Pitch Stability Analysis, Empennage Design, Laser Cutting
Matthew Maberry
Airfoil Selection, Wing Design and Analysis including structural considerations, assisted
in empennage design and analysis, horizontal stabilizer modeling, assisted with thrust
testing, completed many aspects of Mechanical Laboratory writing, completed
empennage construction, assisted wing construction.
Joseph Travaglini
Thrust Test LabView VI Setup, CorelDRAW Files, Laser Cutting/Manufacturing,
Assisted with Fuselage and Wing Construction, Landing Gear Research, Fuselage
Research, Class Presentations, Fall and Spring Final Reports, Assisted with Structural
Analysis
Benjamin Waller
Project Budget and Record Keeping, Project Webpage Design and Maintenance, Wing
Mockup Construction, Fuselage Design and Construction Work, Empennage
Construction, Wing Construction, Mech. Lab Report Writing, Final Aircraft Monokoting
and Preparation
Travis Cushman
Class PowerPoint Presentations, Original Thrust Test, Fall Semester Final Report
Writing, Formatting and Organization, Fuselage Design, Solid Modeling, MechLab Test
Rig Design/Construction, Poster, Construction/Assembly, Final Report
Writing/Formatting and Organization.
- 9 -
4. Introduction
4.1 Project Beginnings
This project was initially conceived in a
fundamental form near the end of the 2011-2012
academic year. At a basic level, this capstone is
built on an engineering competition, the essence
of which is small scale powered flight. Similar
established projects were known to exist at the
time this undertaking was first envisaged, and it
was hoped that an Aero Design group could be
established for students interested in aerospace
and flight sciences. This was of course,
successful, and the first SAE Aero Design
project group has been formed.
4.2 SAE Aero Design Competition
The central component of the capstone project is
the SAE Aero Design competition, the rules of
which are the source of most of the project goals,
guidelines, and regulations. This is an event
created and administered by the Society of
Automotive Engineers, in which groups of
college level engineering students from around
the globe may participate. This event may
adequately be described as an aircraft heavy-lift
competition.
Three classes are available for entry: regular,
micro, and advanced. The class chosen by this
group is the regular class. The objective of the
regular class is to design, substantiate, construct,
analyze, and test a remote controlled aerial
vehicle of limited specifications with the
capability of completing a circuit with payload
beyond its own weight.
The most critical design limitations are the
specification of engine size, the total
length/width/height of the plane, the total
possible weight, the takeoff and landing
distances, and to some degree, the materials
allowed in construction. Creating an aircraft
based on these restrictions while maintaining the
ability to lift enough payload to effectively
compete requires evaluation of design tradeoffs
and development of a very specialized
configuration. Focus must be maintained on the
assembly of components such that the vehicle
will be simultaneously stable, structurally sound,
and task-effective.
Through effective analysis and research, the
group must determine which characteristics may
be sacrificed in order to achieve others, and then
how to build the aircraft to achieve the desired
performance characteristics. Therein lies the
challenge. No single aircraft is suited for all-
around performance, and in this case need not
be.
4.3 MEE Capstone Project and Associated
Opportunities
The feasibility of attending the actual
competition was discussed early on in the project
timeline. Ultimately, it was decided that having a
competitive aircraft by the competition date in
April was not guaranteed given that this group
represents the first iteration from the University
of Maine, and none of the team members have
aeronautics experience. In order to participate in
the competition, groups must register some
months ahead of time, and by the registration
deadline, the prediction of successful
competition entry was not strong enough to
warrant spending a large portion of the team
budget in registration fees.
However, it was also determined that that
success of the project was not hinged upon the
attendance of the competition, for many reasons
which will be cited below. The goal set forth
from the beginning remains to produce a vehicle
that is “competition worthy”.
To be “competition worthy” the aircraft would
need to conform to all restrictions outlined in the
official rules. A brief summary of these, for the
regular class, is as follows:
Maximum takeoff distance is 200 ft.
Maximum landing distance is 400 ft., in
the same direction as takeoff
Aircraft must me heavier than air, fixed
wing construction
Total combined length measurement not
exceeding 225 in.
Gross weight restricted to 55 lb.
Prohibition of fiber reinforced plastics
and use of lead
- 10 -
Prohibition of metal propellers
Mandatory completion of 1 360 degree
aerial circuit
Controlled flight and intact landing
Use of either O.S. 61 FX or Magnum
XLS-61A two-stroke power plant
Use of competition approved
nitromethane fuel (10%)
Designed access to fuel tank and single
payload bay
No gyroscopic stabilization
No pressurization of fuel tank other
than with stock exhaust outlet of engine
Score awarded for max payload carried,
as well as accurate prediction of
capability
Empty payload point bonus
In addition to meeting these physical
requirements, the group has modeled this
document to a large extent on the formal
competition report guidelines, including a large
portion of the substantiation required to be
present at competition.
It can certainly be seen from the description thus
far, that the Aero Design event presents a
significant engineering challenge and requires
honed design skill. The aircraft design process is
beset with facets of mechanical and aerospace
engineering including, but not limited to the
following:
Aerodynamics and the loads derived
from aerodynamic considerations
Usage and analysis of airfoil shapes for
determination of lift and moment
coefficients, pressure distributions and
the like
Fluid flows
Electronics and radio control
Linkages and machinery design
Force and moment transmission
Assurance of stability in three
dimensions
Engine performance, operation and
tuning
Structural design, analysis, and testing
Structural load paths and factors of
safety
Part fabrication
Solid modeling
Advanced planning in weight buildup
and component fitment
Work delegation, time budgeting, etc.
As such, it requires the usage of skills learned
prior to and within the fourth year of schooling,
as well as the initiative to self-educate where
new skill and concepts are necessary. It presents
engineering situations potentially seen in
industry and in many ways allows for effective
experience.
4.4 Aerospace Studies at the University of
Maine
Another consideration is the promotion of the
project for continuation in future years. The
University of Maine is currently not far removed
from the inception of its Aerospace
Concentration. While this is effective in and of
itself, this capstone group is interested in
heightening awareness of aerospace studies at
the university, and developing them yet further.
One way this can be accomplished to some
degree is to establish a project within the
aerospace field which acts as an outlet for
students to experience effective design and use
of knowledge gained in the classroom. It is
intended that this project be perpetuated in future
iterations, and that it work in conjunction with
the existing model aircraft club to form a well-
established group.
4.5 Summary of Focus and Objectives
Overall, the goal of this undertaking is to
compile engineering knowledge and skill, learn
new techniques and information, and apply said
- 11 -
items to the design and fabrication of a flight
capable radio controlled vehicle, from scratch,
which matches the requirements established by
the SAE for its Aero Design Competition.
The aircraft should be able to perform such that
it would be able to earn points at the event. That
is, it should be able to lift as much weight as
possible while remaining itself very light.
Aerobatic performance is of no concern; rather,
good lifting qualities at low speed and stable
configuration are key. The vessel should be
easily controllable and not hazardous to operate.
The specifics of the design and how it was
intended to accomplish these things as well as an
evaluation of whether it was able to do so are
contained herein.
5. Design Description
5.1 Components and Systems
The aircraft is a total of 65.25 inches length, 129
inches in width, and 15.5 inches in height. It is
powered by a two-stroke gas powered engine.
The entire thing is constructed out of balsa wood
and piloted wirelessly through a 6-channel
transmitter and receiver that control the throttle
and control surfaces of the aircraft. It is a total of
13.75 in weight without any payload. See Figure
1 for the finished product.
Figure 1 - Finished Aircraft
5.2 Engine
The engine used to power the aircraft is a
Magnum XLS.61A. It is a two-stroke engine that
runs on model aircraft glow fuel, 15%
nitromethane. A manual for the engine can be
found in the Appendices that contains all other
necessary information about the engine. See
Figure 2 below for a picture of the engine.
Figure 2 - Magnum XLS 0.61A
5.3 Wing
The airfoil used in the design of the wing is the
Selig 1223. The total wingspan of the aircraft is
10.2ft.
The wing configuration is designed with 4
different aspects in mind: taper, twist, angle of
incidence, and dihedral. For the final design,
each half wing (pinion) has a starting chord
length of 16 inches, tapers down to 12 inches and
remains a constant 12 inches chord length from
2.55 feet (middle of the half wing) out to the tip.
The wing is attached at an angle of incidence of
3°, has a dihedral of 2°, and the whole wing has
2° of twist.
The wings were constructed with balsa wood
using a standard spar and rib technique. Each
wing has one flap and one aileron.
- 12 -
The method in which each wing is affixed to the
fuselage is by a 1 inch x 1inch aluminum tube
that runs from inside of the wing and protrudes
from the first rib by 2.25 inches. This aluminum
tube slides into an aluminum tubing sheath that
is located in the fuselage and is fixed in place by
two bolts and nuts on each half wing.
Figure 3 and Figure 4 are pictures of the final
wing design.
Figure 3 - Wing End View
Figure 4 - Perspective View of Wings
5.4 Fuselage
The final fuselage is a very basic design with a
removable cover on the top of it that allows
access to the components contained inside of it.
The floor of the fuselage is made of balsa
hardwood where two bolts extend upward to
secure the payload plates during flight. The
engine mount on the nose of the fuselage is also
constructed out of hardwood balsa and is
designed in a way the engine is completely
exposed.
Like the rest of the aircraft, the fuselage was
constructed with a rib and spar technique with a
single boom extending back to the empennage. A
PVC pipe runs from the fuselage back to the
empennage that contains all of the wiring
controlling the surfaces on the horizontal and
vertical stabilizer.
Two plates mounted on the sides of the fuselage
are made of balsa hardwood with a square hole
cut into the side where the two wings mount into
1x1 aluminum tubing sheathes contained in the
fuselage. Aside from the pieces mentioned, the
entire fuselage was constructed using standard
balsa wood. See Figure 5 and Figure 6 for a
graphic of the final design.
Figure 5 - Final Fuselage; FWD End
Figure 6 - Fuselage FWD End with Hatch Open
5.5 Wireless Systems
The wireless system consists of a Spektrum DX-
6i 6 channel 2.4GHz transmitter and an AR6210
2.4 GHz receiver. See Figure 7 below.
Figure 7 - Transmitter and Receiver
- 13 -
5.6 Controls
There are 5 channels that need controlling on the
aircraft: 1 for the ailerons, 1 for the elevator, 1
for the rudder, 1 for the flaps, and 1 for the
throttle. These items are controlled by JR Sport
ST47 servos. See Figure 8 below.
Figure 8 - Servo Illustration
5.7 Empennage
The empennage of the aircraft is attached to the
tail boom of the fuselage through PVC fittings.
The connection located on the horizontal
stabilizer is adjustable (See Figure 9), allowing
for different angles of incidence for the
horizontal stabilizer.
Figure 9 - Adjustable Angle of Incidence
The horizontal stabilizer has a 10 inch chord and
36 inch span with an elevator that controls the
pitch of the aircraft. Two vertical stabilizers
extend upward from the horizontal stabilizer tips.
Each one contains a rudder that controls the yaw
of the aircraft. The entire empennage is
constructed out of balsa wood, using the spar and
rib technique and covered over with balsa
sheeting. See Figure 10 for a picture of the
empennage design.
Figure 10 - Completed Empennage
5.8 Landing Gear / Externals
The landing gear of the aircraft consists of a
fixed set of two wheels that are attached to the
underside of the fuselage, close to the center of
gravity. A single wheel is also located under the
empennage. As the aircraft sits on the ground,
the landing gear is oriented so as to give the
aircraft a positive angle of attack to assist in
producing lift during take-off. See Figure 11 for
an example of this set up.
Figure 11 - Ground Stance
All external surfaces of the aircraft are covered
in a material called Monokote that is ironed on
and shrinks to form fit with the aircraft surface.
For a visual of the Monokote finish, refer to
Figure 1 again.
- 14 -
6. Design Concept Process
6.1 Engine Selection
There was virtually no design choice to be made
for the power plant of the aircraft. The SAE Aero
Rules and Restrictions required anyone
competing in the regular class to choose between
two similar engines. The choices are the
Magnum XLS .61A and the O.S. 61 FX.
Both engines are the same size bore and there are
negligible differences in the performances of
each engine. The group chose the Magnum XLS
.61A for the engine based on cost and
availability as the O.S. 61FX is discontinued in
production and the group wanted a brand new
engine to work with.
6.2 Airfoil Selection
6.2.1 Design Goal
After the initial organization of the project, as
well as familiarization with the rules, the
tentative design process lead to the
characterization of the wing. Each component of
the aircraft is of great importance and intended to
accomplish a specific task. However, one must
always keep in mind one essential rule of aircraft
design – a process of trade-offs and optimization
is inevitable. There are as many aircraft
configurations as there are tasks for an aircraft to
complete, and the goal of the engineer should be
to arrange for a design that is tailored to the
specific goals at hand.
As such, it is vital to ensure that the various
components involved interact well and
strengthen each other. That being said, there
must be a starting point, and as mentioned above,
after some consideration, this is clearly the wing.
The wing is the primary mode of generating the
force required to lift the aircraft. In the end, it is
what gets the plane off the ground. It also plays a
large role in the stability characteristics of the
vessel, accounts for most of the weight, acts as a
payload containment system and several other
things.
The aerodynamic forces which need to be
corrected by the tail and control surfaces stem
from the action of the wing to a large degree, and
the behavior of the wing will limit other
components. For example, with a specific weight
goal in mind, the effectiveness of the wing to
produce enough lift within the limiting velocity
will determine how large an area is needed.
Along with the thrust available from the engine
and propeller system, it is one of the primary
limiting factors for the overall aircraft
performance. So, naturally, it was decided to
focus the early part of the project on the wing
design. It was thought that obtaining desirable
and workable performance from the wing was
deserving of time and attention and that due to
the logic above, many other decisions would
follow from the final choices made with regard
to the wing.
6.2.2 Research Done
The next question was of course, where to start
designing. Many vital characteristics of the wing
are determined by the two-dimensional section
chosen. That is, the airfoil. These include, the 2D
pressure distribution, total wing lift coefficient,
angle of attack performance range, associated
pressure drag, manufacturability, flow separation
behavior, etc. There are several tools available
that allow even a modest user to analyze and plot
airfoils.
Some consideration was given to designing an
airfoil for our purposes from the ground up.
However, given the plethora of preexisting types
and reliable test data readily available, this was
considered a wasteful endeavor in a constrained
timeframe.
Several computer programs that allow for
analysis of two-dimensional sections were
reviewed, and upon usage it was found that
XFLR5 was the most user friendly. Another
benefit of XFLR5 is that it also incorporates a
“Wing and Plane Design” feature that extends
beyond 2-D plotting. While intuitive, learning to
utilize this software took some time, and learning
the subtleties of its capability is an ongoing
process.
6.2.3 Operating Conditions
It was then time to study airfoil geometry from
the perspective of our project. That is, with
respect to design of a relatively slow flight small
scale craft intended to lift a maximum amount of
weight. To do this, certain information which
defines the operational flow must be known.
Given the size of the engine selected, and well-
- 15 -
known typical speeds of large models, airspeed
was expected to range from just a few feet per
second, up to perhaps 100 feet per second,
although the latter would be rather extraordinary.
Thus, for free and open flow over the wings, the
expected range of Reynolds numbers was
decided as 50,000 up to 900,000+. Individuals
may note that this is a very low Reynolds
number regime, over a good deal of which the
flow could reasonably be expected to be fully
laminar. Laminar flow is accompanied by a
lower shear stress on the surface of the wing,
while the churning effects of the turbulent flow
draw higher energy elements toward the
boundary layer and increase the skin friction.
This higher level of kinetic energy also induces a
greater tendency for the flow to follow the
contour of the surface over the adverse pressure
gradient on the wing. This limits the pressure
drag due to the separated flow. So while the idea
of laminar flow may sound desirable, a turbulent
flow would be more ideal. Indeed, model aircraft
often contend with flow separation at early
points on the chord, leading to lack of lift. A
common effect is known as a separation bubble,
which is an area of circulation which disrupts the
flow without leading to complete detachment.
Another issue to contend with was the fact that
atmospheric conditions are less than stable and
uniform at sea level. It is not difficult to imagin
an aircraft of small size being forced off-kilter by
a burst of wind, or encountering slight variations
in density and temperature over a single flight.
This meant that our choice had to provide
desirable characteristics over a decent range of
angles of attack.
Clearly the goal was to find and use an airfoil
section that provided the high coefficients
needed in a lifting competition over a range of
angles of attack, and was efficient at lift
production in the low Reynolds number regime.
6.2.4 Resources Utilized
Several books published on home-built aircraft
design were found and perused for information.
Often times, such references come with
comprehensive lists of airfoils which designers
commonly use. Perhaps the most well known
resource examined was “Theory of Wing
Sections” by Von Doenhoff and Abbott,
although this book is not intended for model
building. Several suitable sections were found in
“Model Aircraft Aerodynamics” by Martin
Simons.
Some rather interesting sections were found in a
little known book titled “Model Aircraft Design
and Theory of Flight”, written by Charles
Hampson Grant, an early pioneer of self-built
model aircraft design. No information could be
found with regard to this on the internet. Some
time was spent mapping the coordinates
provided in the book which define the shape of
the airfoils into a useable text format which
could be plotted and inserted into software such
as SolidWorks.
Another valuable resource for finding airfoil data
was the University of Illinois maintained
database, run by Professor Michael Selig (the
designer of many of his own sections). This
resource is of particular note, given that one of
its main focuses is the low Reynolds number
regime.
6.3 Testing
Based on preliminary data from each of the
above sources, 10 individual airfoils were chosen
for further examination and testing. Each was
either of very frequent usage among scale
aircraft design, or was recommended for low
Reynolds number flow. The airfoils which were
proceeded with are listed as follows:
Clark Y
Eppler E193
Eppler E197
Eppler E423
NACA 0009
Selig S8036
Selig S1223
Grant G8
Selig S1210
Eppler E64
The coordinates for these airfoils were all found
on the UIUC database mentioned above, with the
exception of the Grant G8 as previously
described. These were loaded into XFLR5 for 2-
D analysis with specific focus on several
characteristics. Namely, the lift, drag, and lift to
drag ratio polars, the boundary layer behavior,
the center of pressure movement, and the
pressure distribution across the section.
- 16 -
It was found that viscous approximation led to
non-convergent results a large percentage of the
time, so inviscid analysis was performed instead.
Based on a paper published by Professor
Michael Selig, values of lift coefficient can be
expected as approximately 10 percent different
due to this.
The analyses were done simulating a range of
free streams, not considering tip effects, with
Reynolds numbers ranging from 100,000 to
600,000 and at angles of attack from -3 degrees
to 15 degrees. This data was then compiled into a
spreadsheet for ease of comparison. An example
of a small part of the data collected is presented
below in Figure 14.
An example of the pressure lift to drag ratio, as
well as the pressure distribution plot on an
airfoil, is shown Figure 12 and Figure 13 as an
illustration of the capabilities of the program.
Figure 12 - Example Cl and L/D Polars
Figure 13 - Example Pressure Distribution
After the data was compiled a comprehensive
comparison of the group was made.
6.4 Results Discussion
Overall the Clark Y exhibited almost no
preferable qualities. It was included due to its
wide popularity among model builders on
forums. It was quickly discovered that while it
may work for general purposes, it is not suited
for the lifting target associated with this project.
The Grant G8 was also somewhat of a
disappointment. It was strictly middle of the
pack in terms of lift coefficient and drag
coefficient.
The best performances were in general, by the
Selig S1223, the Selig S1210, the Eppler E423,
and the Eppler E197. The compilation of the data
used for comparison can be seen in the
appendices of this report.
- 17 -
Figure 14 - Example Data Spreadsheet
A definitive choice between these four sections
was halted until after some knowledge
concerning the wing configuration was had.
Instead, all four were used in additional analysis
that will be described later.
It was initially planned to test several of the best
performing models in the wind tunnel. The
portion of this project dealing with the wind
tunnel endeavor follows this section. For the
purposes of completeness in this section, it can
be said that the S1223 was the ultimate airfoil of
choice.
The Selig S1223 is actually advantageous for
several reasons. Research indicates that it has
become quite favorable among competitors in
the SAE Aero design challenge, and has been
used with great success recently. This of course,
makes sense given that it was essentially
designed with similar restrictions to the
competition in mind. But how exactly does the
S1223 work so well. What about its shape and
design makes it perform better? Professor Selig
has described his design efforts. The high
camber and progression of both the upper and
lower camber throughout the chord produces a
convex adverse pressure gradient. The trailing
edge is more heavily loaded with lift production,
which generates a stronger moment over the
chord. Some additional drag is introduced here,
but the benefits outweigh this. Specifically, the
point of flow separation moves very slowly
forward from the rear as the angle of attack is
increased. Even after the flow separates, the
maximum coefficient of lift rises with angle of
attack. This allows the coefficient of lift to rise to
over 2 before stalling begins to occur.
So in summary, the Selig S1223 has a high
maximum lift coefficient, which is good over a
range of angles of attack. It should, while
lowering stall speed, shorten takeoff and landing
distances, and increase the payload capacity. It
can operate near its peak lift coefficient while
maintaining a mild stall characteristic.
7. Wing Configuration
7.1 Wing Design
There are three main parts which constitute the
wing design as a whole. These are, the airfoil(s)
defining the wing section, the configuration of
the wing, and the size, or dimensions of the
wing. Each of these parts work together and are
tailored for a specific objective. As has been said
thus far and will be stated again, one must
always remember the goals of flight and
operation for which the aircraft is being
designed, because they will directly influence all
design decisions. Most, if not all aircraft, are
specialized. That is, there is no Swiss Army
knife of planes.
A qualitative description of the flight regime for
which the aircraft is designed is given above.
7.2 Wing Section
The wing section (profile) ultimately decided on
was the Selig S1223. The reasoning for this is
discussed above.
The final design of the wing incorporates a
uniform wing section throughout the entire span
of the aircraft. That is, no aerodynamic washout
is included. All washout in the wing is provided
by mechanical twist. To establish the ailerons
- 18 -
and the flaps on the wing, a division is made in
the airfoil at three inches from the trailing edge,
and the rear portion is then used as the structure
of the control surface or high lift device,
respectively. No modifications are made to the
Selig in the way of slots or slats. The lift, drag,
and moment coefficient data for the S1223 was
available to some extent through internet
resources, and flow simulations using XFLR5
were also generated, which will be discussed in
more detail hereafter.
7.3 Design Goal
The plane was designed for a competition in
which as much weight is carried around an aerial
circuit as possible. There are limitations on
overall dimension, total weight, take-off and
landing distance, engine power, among other
things. The goal is to lift a maximum weight at
takeoff, fly the circuit, and successfully land. A
more concise list of the objectives in mind is
summarized here:
- Select and design a configuration which
prioritizes lifting capability by
sacrificing adroit maneuverability and
flight speed.
- Design the integral structure of the wing
to handle flight loads associated with
theoretical weight limitations imposed
by the thrust capacity of the specified
engine and the size limits, while
minimizing the inherent structural
weight.
- Minimize project costs
A large number of resources were used to
examine the benefits and drawback of various
configurations and to decide on their relative
importance. A compilation of the relevant data
gathered in this process, with comments
concerning the applicability of each item follows
below.
7.4 Research Done
7.4.1 Wing Span
We begin by examining the subject of wing
span. The span of the plane is closely tied to the
characteristics of wing area and aspect ratio. The
maximum sum dimension for the Aero
competition is 225 inches. Thus, given
reasonable fuselage length and plane height, a
maximum span of approximately 110 inches is
present. However, it is not necessarily advisable
to simply maximize the span of the wing.
Increased wing span has both positive and
negative aspects. A plane achieves flight due to
the reaction force that is lift. In simplest terms,
the wing diverts air downward. A change in
direction represents an acceleration, and this
requires a force. Lift as a phenomenon is
commonly attributed to Newton’s third law,
when in reality, the third law only asserts that a
force will be present of equal magnitude, and
does nothing to explain the origin of said force.
To understand wing span is involved in lift,
consider two planes flying under the same
conditions with the same air speed. The craft
with the larger span will contact more air for a
given amount of time in flight than will the
smaller vehicle. Essentially, more air diverted
results in more lift. (Remember there has been no
mention of wing area or aspect ratio to this
point). Thus under set conditions, a longer span
is a better producer of lift.
One of the chief drawbacks to large wingspans is
directly related to the discussion above. We have
established that the wing is coming into contact
with more air by virtue of its length. This means
in turn that more material is contacting a surface
of the plane. A longer wing will generate more
skin friction than a short wing due to this fact.
One might say that a long wing is terrible for
parasite drag. See Figure 14 below.
7.4.2 Wing Span with Reynolds Number
We now consider the conditions our plane will
operate under through a different lens. We will
be travelling at quite slow speed. That is, in the
Low Reynolds Number regime. A basic facet of
aerodynamics is that slow speed flight is
dominated by induced drag. Induced drag is the
small component of the lift force vector opposed
to the direction of travel, as illustrated by Figure
15. This arises due to the orientation of the
aircraft necessary for the generation of lift.
- 19 -
Figure 15 - Relative Drag Contributions
The problem of parasite drag is minimized by the
fact that the plane will be flying so slowly
compared to large scale aircraft, that by the
standards of say, even a single seat Cessna,
almost no material will contact the wing surface.
Parasite drag also rises with velocity for the
same reason that it increases with area: more air
mass is grazing the wing surface. Given the
dimensions and speed of our machine, parasite
drag is secondary to induced drag. That is, the
drag associated with the production of lift.
Because we will be flying so slowly lift will be
hard to come by (a faster plane will divert more
air in a given span of time). Thankfully, our
plane is also inherently lighter than large scale
planes, and there are other ways to increase lift
than raising speed.
Another drawback of an increased span is that
the wing may become more difficult to support.
The wing of a plane is essentially a cantilevered
beam, and having a long wing puts mass out at a
long moment arm. Long wings may require the
use of deep spars to support, and therefore may
be more of a challenge to build. Within this
Figure 16 - Induced Drag Schematic
frame of mind, it is not hard to see that a long
wing is more prone to aerodynamic flutter.
Another consideration of span is the effect it may
have on control of the plane. Control surfaces are
intuitively more effective at a larger distance
from the body of the craft because of their
longer moment arm. By the same token, one may
say that a large span plane is particularly
sensitive to control inputs if the control surface
position also extends. So it may be expected to
follow that a longer winged plane will be more
difficult to trim for flight.
- 20 -
Higher sensitivity does not necessarily correlate
to responsiveness, however. Actually, a longer
span will have the opposite effect. As an
example, consider an aircraft in an attempt to
roll. The rotational inertia of the plane will
increase with wing span. We all remember the
example of the spinning figure skater from
physics? Let us consider our goal in competition
once more. Quite frankly, we don’t care about
roll rate.
In summary, it is certainly a good idea to attempt
a long span wing. However, there are very real
limits on the maximum length. Perhaps it seems
that the negatives outweigh the benefits. But we
have not considered any other related parameters
at this point.
Figure 17 - Clmax vs AOA for Various AR
7.4.3 Chord Length
The chord length is an important parameter for
several reasons. First, for an airfoil, the chord
length is the characteristic length associated with
the Reynolds number. Thus, to match Reynolds
number in aerodynamic testing, we must have
some information regarding the chord length.
However, in terms of wing configuration of the
full scale machine, the chord length is most
influential in obtaining the desired wing area and
aspect ratio. The lift produced by a wing section
is proportional to the severity of the curvature at
any point and always projected perpendicular to
the surface. On highly cambered sections like
those seen in low Reynolds number flow, most
of the lift is produced within the first 25% of the
chord. Having excess area which contributes
little to the lift is undesirable.
The chord length can be changed to achieve a
suitable wing area for lift production. The chord
length at the fuselage of the craft will also
determine the physical base with which to
support the cantilevered weight of the wing.
Another consideration when discussing chord
length is the fact that it does not need to remain
constant throughout the entire wing. The chord
- 21 -
length may change with span, adding another
level of variability. This will be discussed in
greater detail in the section below concerning
wing taper.
7.4.4 Aspect Ratio
The aspect ratio of the aircraft wing is defined as
the square of the span, divided by the total wing
area. Changing the aspect ratio will therefore
have a direct effect on the span or the area, or
both. A high aspect ratio usually implies a
relatively long span, and conversely a small AR
typically implies a somewhat short span. Perhaps
the most pertinent relation is that of the span to
the chord.
Increasing the aspect ratio at given angle of
attack will raise the total wing coefficient of lift,
up to a point (See Figure 17). However, this
action will also decrease the stall speed of the
wing, which reduces the useful range of speeds
at which one can operate the machine. This is not
overly worrisome from the standpoint of the
competition, as performance models are not the
goal, and a very large range of operational speed
is not necessarily required.
The aspect ratio of the wing is tied to its
efficiency for lift production. High AR wings
may reduce span-wise flow. The circular wingtip
vortices resulting from the spilling of higher
pressure air from under the wing are minimized
in strength if the aspect ratio is higher due
largely to the shape of the downwash sheet.
These vortices are the primary cause of induced
drag, a type of pressure drag. So, higher aspect
ratio is better for induced drag.
If, as mentioned above, a higher aspect ratio
increases the lift coefficient for a given angle of
attack, the wing can meet a lift requirement with
a smaller AOA, which also helps to minimize the
vortex induced drag. It has already been
mentioned that induced drag is a consideration
given the design goals. The airfoils which have
been selected for further testing at this point are
of the high lift variety, and have been considered
because of their attractive lift coefficient
performance and lift to drag ratios. For airfoils of
this type the effects of induced drag are more
pronounced.
It is known that the induced drag on a wing
varies in proportion to (1 / V^2). This makes
sense when considering that a faster moving craft
will displace more air in a unit time. The aircraft
being designed will operate at quite slow speed.
This problem can be partially remedied by higher
ARs given that for a set speed, using the
knowledge above, the AOA can be smaller, and
thus the coefficient of lift necessary to operate at
said speed will be smaller.
It may seem as though high aspect ratio wings
are great for use at slow speed. This is in many
ways true. However, as a cautionary word, the
Reynolds number must be considered again. In
slow speed flight where the Reynolds number is
low, the boundary layer on the wing is not
energized the same way higher Re flows allow
for, and is prone to separation. Remember that
the unit length associated with an airfoil is the
chord. For a given wing area, a high aspect
ratio will lower the chord, further lowering
the Re and compounding the effect. Careful
design and consideration of the Reynolds
number regime is advised. It is due to this fact
and others that devices such as vortex generators
and turbulators are sometimes introduced on
model wings.
7.4.5 Wing Mounting Style
There are three primary methods for mounting
the wing in regard to position on the fuselage.
These are, intuitively, low mount, mid mount,
and high mount. The high mount technique is
typically the most desirable for a number of
reasons.
Thinking about the fuselage as a lumped mass, it
can be seen that the center of gravity of the plane
will lie below the wing if the wing is mounted
high. In effect the fuselage will be suspended
from the lifting points on the wing above. This
gives the wing a measure of effective dihedral.
Low mounted wings will have the direct
opposite effect.
One of the drawbacks of a wing mounted along
the center of the fuselage is that it makes the
modularization of the vehicle more difficult. For
the purpose of removing the wing quickly and
maximizing payload volume it may be advisable
to devise a method for attaching the wing at the
top where it can be lifted out. Another advantage
of the high mounted wing is that it more readily
allows for strut-bracing against the fuselage of
the aircraft.
- 22 -
7.4.6 Dihedral Angle
The dihedral angle of an aircraft is defined as the
degree to which the wing is angled from a
perpendicular orientation with the fuselage. Only
in extreme cases of dihedral angle will the lift of
the wings be significantly affected.
Because angling the wings alters the direction of
the lift vectors associated with each pinion, the
total lift may be decreased, although for general
angles of dihedral, this is not pronounced.
Figure 18 - Stall Onset for Taper Design
The main purpose of adding dihedral angle is to
utilize the dihedral effect to enable lateral (or roll
stability). Consider an undesired disturbance
which changes the orientation of the aircraft. If
one wing is lifted up relative to the other, the
lower wing will therefore be generating less
relative lift. This induces a sideslip motion where
the air comes not only from the front of the craft,
but also at some angle off the axis of the
fuselage. The airplane slips to the side and down
if uncorrected.
This is where the dihedral angle is beneficial. If
the low wing is angled up, the aircraft presents it
to the off-kilter stream of air at a higher angle of
attack that the upward tilted wing. This produces
a restorative rolling moment which helps to
correct the aircraft. This will also increase the
amount of air striking the side of the plane as it
slips to the side.
Another benefit of dihedral is that it will lower
the center of gravity of the plane relative to the
lifting points on each pinion, which will also
have a positive effect on stability. This is
especially important in our project if our choice
of mounting style is the high mount, because this
will inherently raise the center of gravity.
One possible negative effect of adding dihedral
angle is that if one gets carried away and the
- 23 -
upward slope is too severe, yaw-roll coupling
may be present. This is commonly referred to as
“Dutch Roll”. Remember that the lift rises in
proportion to drag, and so the restorative lift will
pull the wing back as the lateral correction
occurs, yawing the plane. This can be countered
be designing for the minimum dihedral necessary
and in some cases by increasing the vertical
stabilizer area.
7.4.7 Taper
Wing taper is the adjustment of chord length
over the span of the wing. The most prominent
characteristic associated with taper is the wing
loading over the span. Physically, taper may be
used to decrease the amount of material and
therefore weight of the wing as the moment arm
grows from the fuselage.
It has been shown analytically that the lift
distribution over the average wing span is
greatest at the root, and trails off to near zero at
the tips, in an elliptical fashion. The most
efficient chord variation would therefore
decrease the amount of chord-wise material in
conjunction with the decrease in lift. Several
elliptical taper designs have been built. This
design may be difficult to manufacture however,
and often a straight taper is preferred. See Figure
18 for a variety of different taper distributions.
The taper, or amount of taper strongly effects the
distribution of loads on the wing. Highly tapered
wings are prone to negative effects such as
dangerous tip stall, especially at high aspect
ratio.
For our purposes, it is suggested that at most, a
moderate amount of taper be used.
Another consideration is the idea of partial taper.
That is, part of the wing could be constant chord,
while another part is sloped. Similar to this is the
notion of compound taper. This makes
calculation of the wing area and aspect ratio
more complicated. Typically, the region closest
to the fuselage is held constant while the outer
portion is tapered. If taper is to be used, it is
suggested that the leading edge slope backwards
while the trailing edge is constant. This will
make the implementation of the control surfaces
somewhat easier.
7.4.8 Twist
The wing may be twisted aerodynamically or
mechanically. In both cases, the amount of lift is
adjusted as a function of the spanwise location.
Aerodynamically this is done by selection a
different airfoil section for parts of the wing.
Mechanically, the angle of attack is adjusted as
the span progresses. Wing twist is most
commonly seen in low Re regimes as
compensation for highly loaded wing-tips.
Decreasing the angle of attack near the tips
offsets this to some degree. This is known as
“washout”. If the wing is tapered it may be
beneficial to design twist, although this is an
additional complication.
7.4.9 Wing Sweep
Wing sweep is a design consideration often seen
on high speed aircraft which approach mach
speeds. This is largely due to attempts to
minimize the thickness to chord ratio by
increasing the effective chord, and avoiding
interference with shock waves.
These are obviously not things we must consider
for our project. However, sweep has other
effects.
Sources report that approximately 5 degrees of
sweep is effectively equivalent to 1 degree of
dihedral. Sweepback will also provide some
measure of directional (yaw) stability. By the
same type of logic that correlates with dihedral -
under a disturbance, if the airplane is yawed, the
wing swung forward will face the incoming air
more directly than the other and will therefore
have more accompanied drag, pulling it back
around restoratively. Yaw-roll coupling is
evident here.
7.4.10 Edge Design
The design of the wing edge is important in
controlling the amount of air that spills around
the edge from the high pressure surface. To
understand this think about how a wing
generates lift. The shape of the wing is a means
for producing a pressure differential across the
upper and lower surface. Regardless of this
means, this difference is present, and the more
highly pressurized air underneath will tend to
spill up and over around the tip. This decreases
the efficiency for lift of the wing.
- 24 -
One method of reducing this spilling effect is to
square off the edge of the wing. An extension of
this is to extend sheeting on the top surface out
past that of the bottom surface, connecting the
two with a shallow angled piece. If one is to
visualize two streams of essentially span-wise
flow over the top and bottom, it can be seen that
the point of interception of these streams will in
fact be somewhat beyond the edge of the wing.
This will extend the lifting capability of the wing
closer to the tip, increasing its effectiveness, but
also loading it more severely.
Figure 19 - Wing Edge Desing Variation
Perhaps a more notorious device employed for
the same purpose if the winglet, which ultimately
accomplishes the same thing, although the
winglet itself has some weight. The design
proposed does not utilize winglets due to their
additional complexity. However, an angled edge
will be utilized. Figure 18 shows a variety of
wing edges.
7.4.11 Wing Loading
The loading of the wing is the total weight
carried by the aircraft, which the wing must
support, divided by the area of the wing.
The pros and cons of high and low wing loading
are as follows: A highly loaded wing requires
more air flowing over it to attain the same lifting
power as a lightly loaded wing. It will have a
higher stall speed but also will required faster
take off and landing speeds. Highly loaded wings
have in general less parasite drag. They are also
harder to maneuver with simple control surface
input and undergo larger centrifugal forces while
turning.
Lightly loaded wings are capable of slow
takeoff and landing, but have a lower stall speed
and more narrow range of angle of attack. They
are have less inertia in flight and are more easily
disturbed, which is a characteristic which blends
well with high aspect ratios. For our purposes, a
lower relative wing loading is deemed
advantageous. Figure 19 shows wing loadings
progressing to the tips of the wing.
7.4.12 Externals
External additives would, for the purposes of this
document, include devices such as stall strips
and turbulators. That is, items which may “trip”
the boundary layer flow on the surface of the
wing, rendering it turbulent at some predictable
location, and allowing it thereby to adhere to the
surface at higher angles of attack.
Stall strips are generally rough material lined
chord-wise at some point along the span, which
will ensure that stalling will occur first at a more
desirable location, generally the inner area.
These types of devices should not be necessary
with proper design and are usually added after
some testing has been done, and alterations are
found to be in order.
7.4.13 Multiple Wings
A vast amount can be said concerning the design
of aircraft with multiple wings. This class
includes biplanes, triplanes, sesquiplanes and
other types of multiplane vehicles.
Most of the design challenge here results from
the interaction of one wing with the other.
Remembering that a wing has by its nature a
lower pressure over its top, one can easily see
that adding another wing above it would hinder
the usefulness of that top wing. In fact for each
- 25 -
additional wing added the efficiency of the
sandwiched wings decreases. This effect can be
somewhat mitigated by staggering the wings
forward or back. Overall, each individual wing is
less efficient than any monowing plane.
However, because the lifting area is obviously
increasing, the total lift is raised. There are
several issues to account for however. Multiple
sets of wings would require additional bracing
structure leading to some increased drag, and
would also increase weight for the same area.
Figure 20 - Lift Distribution Illustration
In fact, one of the great ironies of aviation is that
in the early era, it was clear that monowing
planes would require internal bracing. It was
thought that due to this, the wings would need to
be so thick, the additional drag would ruin the
performance. So designers stayed with extremely
thin wings with cross bracing and wire. In
reality, the drag on these structures was far larger
than thick single wings!
Biplane design is often seen with aerobatic
planes due to the fact that for the same total
lifting area, the span can be reduced, which is
good for high speed planes which need large
alpha ranges for maneuvers and low roll inertia
for snapping the plane over. Although this is not
the only benefit, it was determined early on that
due to some of the drawbacks mentioned above,
designing a biplane was not worth the effort, and
that a properly designed single wing craft could
accomplish the job well.
7.4.14 Final Wing Size and Configuration
The total span of the wing is 10.8 feet. Each
pinion of the wing is 5.1 feet in length, from its
root to tip, and the wings are mounted to the
sides of a fuselage six inches in width. The wing
is tapered such that the ratio of tip chord to root
chord is 0.8. For the designed 16 inch root chord,
this means the tip section is 12.8 inches.
This tapering is not carried out in a constant
fashion, however. A basic type of compound
taper is used, where all of the decrease in chord
length occurs linearly in the first half of the
wing, and the outer half remains at constant
chord.
The wing is fabricated with all of the taper in the
leading edge. The trailing edge is straight and
perpendicular to the plane of the fuselage side
wall. The decrease in chord length is
accomplished by overall scaling of the section
about the centroid of the S1223, rather than
merely decreasing the chord-wise dimension of
the wing.
Otherwise the group would encounter attendant
changes in aerodynamic characteristics. A side
effect of this is that the wing is also tapered in
- 26 -
the z axis, meaning the lowest point of the airfoil
surface is elevated above the lowest point at the
root, if only by a small amount, even when
ignoring the dihedral angle.
A dihedral angle of two degrees is used in the
wing. This is a simple linear dihedral achieved
by physically tilting the wing upward by its
mounting tube, where it is fastened to the
fuselage assembly.
As mentioned above, a mechanical twist exists,
and is a relatively small 2 degrees. All of the
twist occurs in the outer half of the wing, where
the section is constant chord. Constructing the
wing with the designed twist sees the trailing
edge raised linearly, with each rib rotating about
a fixed point at the leading apex of the S1223.
The outermost edge of the wing, at the tip, is
designed to be square without the upper and
lower surfaces converging to a single line, and
without any rounding of the planform.
7.4.15 Construction and Components
The frame of each wing is a fairly standard array
of balsa ribs connected by a network of beam-
like support structure and spars. Each half of the
wing is made up of 21 ribs spaced three inches
apart, with one alternate spacing accounting for
the discrepancy that a strictly linear pattern
would create with the overall span intended.
Three quarters of an inch has been removed from
the leading edge of the S1223 on every rib in
order to make room for a manufactured leading
edge of much stiffer basswood. The main spars
of the wing are placed in a pair at the quarter
chord. These are quarter-inch square balsa stock
pieces epoxied together for length extension.
Given the availability of a laser cutting tool, the
two main spars are connected with four balsa
sheets mounted on the spars as side walls.
Together, these four pieces run the span of each
half-wing and form a hollow rectangular beam.
As this beam structure passes through the seven
most inboard ribs, clearance is left at its center to
allow square aluminum stock to be inserted for
use in mounting the wing. Several of the ribs in
the vicinity of the aluminum beam are
manufactured from hardwood, rather than balsa.
Consistent spacing at the trailing edges is
maintained by the addition of several balsa
braces between the ribs, which will also serve as
mounting locations for the control surface
hinges. Both flaps and ailerons are incorporated.
The ribs are made of 3/32 inch this balsa sheet,
and are fitted with lightening holes in an attempt
to reduce as much mass as possible.
The spars making up the beam structure are
sunken into the edge of each rib far enough that a
layer of 3/32 inch sheeting may be laid over the
desired portions of the wing while maintaining
the original contour of the S1223. In fact, much
of the airfoil is reduced by this thickness such
that when the sheeting is added, the section
surface meets the basswood leading edge with no
rise or drop in need of sanding.
8. Website
As part of the capstone requirements, a website
for the design project was required. The first
time around, the group tried to construct a
website from a website called
http://www.moonfruit.com/ where templates of
websites were already made with easy to
navigate drop-down box menus and image
spaces etc.
However, there was some speed issues
associated with the website. Per request, the
website was re-designed and formatted into an
html code. This code can easily be modified and
updated whenever some changes need to be
made to show the progress of the project. The
website contains all important documents that
were written throughout the academic year,
competition rules for the year that project was
designed around, some pictures of progress
made, final designs etc. Below is the link to the
UMaine SAE Aero website.
http://www.umaine.edu/MechEng/mo/2012-
2013%20Capstone/AeroDesign%20Website/Aer
odesign.html
9. Preliminary Modeling
In the design process, some modeling was
needed to get an idea of how things fit together
and how things looked, space-wise. Preliminary
SolidWorks was done for one-half of the wing to
help determine how large and how complicated
the wing would be to fabricate. The preliminary
model incorporated the twist, the taper, and the
rib/spar technique that was used in the
construction process.
- 27 -
This model was then used to build a mock-up of
the half wing and to guide the construction of the
mock-up. The model helped the group determine
how long the construction would take for at least
the wing of the aircraft. See Figure 21 thru
Figure 23 for graphics of the preliminary model
and the final product of the mock-up.
Figure 21 - Wing SolidWorks Iso View
Figure 22 - Wing SolidWorks End View
Figure 23 - Hand-Built Wing Mockup
Preliminary modeling was also done on
SolidWorks of the fuselage. The group needed to
know that the design of the fuselage was large
enough to fit all of the components. Below,
Figure 24 shows the original spacing SolidWorks
model of the fuselage.
Figure 24 - Fuselage Spacing Model
10. Fuselage
The fuselage of the aircraft is the component that
essentially everything else is attached to and
where most of the components necessary for
running a remote control aircraft are contained
such as the fuel tank, battery, and receiver.
Thus the following considerations were taken
into consideration when designing the fuselage:
Sizing
Wing-Fuselage Connection
Empennage-Fuselage Connection
Motor-Mount
Accessibility
Construction Process
10.1 Sizing
The fuselage needs to have enough space in it to
be able to fit all of the different components
required to run the aircraft inside of it, plus the
payload that the aircraft is to carry.
As illustrated earlier, a rough SolidWorks model
was drawn up with all of the components to be
contained in the fuselage placed inside of it. The
sizing used in the model was big enough for our
purposes and left plenty of extra room to work
with so it was decided that the original design
was indeed big enough.
The main part of the fuselage was designed as a
6x6x19 box that had would have enough space to
contain the payload plates, the fuel tank, the
battery, and the receiver.
10.2 Wing-Fuselage Connection
The wing is designed to be two separate pieces
that come together at the fuselage, as stated
- 28 -
earlier. Thus, there needed to be some way to
connect these two pieces to the fuselage that was
sturdy enough to support the weight of the wing
and the forces it would experience while still
remaining easily assembled and disassembled.
At first, the idea of using the wing as the actual
cover to the main part of the fuselage was
considered but this idea was dismissed because
this method would be too difficult to effectively
fabricate without sacrificing strength.
Then the idea of using plates of some sort
mounted to the fuselage with a hole through it
arose. This hole would have the dihedral of the
wing built in and some sort of piping would
make a sheath for the main spar of each wing to
fit into. The spar and sheath would then have a
bolt driven through them and be nutted on both
sides to secure the spar and sheath.
PVC was discussed for the sheathing however
after much debate, it was decided that the
sheathing and wing spar would be made out of
1x1 aluminum tubing. The plates attached to the
fuselage are made of balsa hardwood for extra
support in the fuselage when holding the weight
of the wing.
See Figure 25 and Figure 26 for an illustration
of the wing-fuselage connection.
Figure 25 - Aluminum Sheath w/ Dihedral
Figure 26 - Wing Mounting Method
10.3 Empennage-Fuselage Connection
The empennage-fuselage connection has one of
two options that are commonly accepted: single
boom and twin boom. Both, if done correctly,
serve to stabilize the empennage during flight.
Due to the simpler design and fabrication of a
single boom, the group opted for the single boom
option. The boom is a PVC pipe that extends out
from the rear of the fuselage back to the tail
empennage. The PVC pipe is supported by the
rib design of the fuselage (discussed later) which
gives enough rigidity for the tail assembly to not
be moving around during flight.
The PVC being hollow also gives an advantage
to running the wiring from the fuselage to the
control surfaces of the empennage as the wire
can be run right through the piping.
See Figure 27 and Figure 28 for the empennage-
fuselage connection graphics.
Figure 27 - Fuselage Tail Boom
Figure 28 - Fuselage / Tail Boom Interface
10.4 Motor-Mount
The motor mount is located on the front of the
fuselage and the set up leaves the engine
completely exposed. This set up allows for easy
access to the engine for maintenance and engine
priming etc. Having the engine exposed also
helps in cooling it during flight when it is
running. The motor-mount material is made out
of balsa hardwood for extra support as the engine
is one the heavier components of the aircraft.
See Figure 28 for the motor-mount design.
- 29 -
Figure 29 - Motor Mount
10.5 Payload and Maintenance Accessibility
One of the key design considerations for the
fuselage was the accessibility. Someone needed
to be able to access all of the components held
inside with ease in case of troubleshooting. Also,
according to the SAE Aero Competition Rules
and Regulations, the payload of the aircraft
needed to be able to be loaded and unloaded in
60 seconds or less. Thus there had to be a way
to easily remove a panel from the fuselage in
order to load and unload the payload quickly. A
solution was reached where a cover was made
and hinged to one end of the access area and
Velcro attached the free end to the fuselage to
close the access area during flight. This made
accessing the inside very easy and quick which
was the main goal. See Figures 29and 30 for a
graphic of the concept.
Figure 30 - Payload Bay Open
Figure 31 - Payload Bay Closed
10.6 Fuselage Construction Process
A common technique in building remote control
aircraft is a rib and spar technique. This way of
fabricating the different components of the plane
saves a lot of weight while sacrificing a minor
amount of strength and rigidity.
This process was used in designing the fuselage
of the aircraft. As seen in the previous graphics,
multiple balsa ribs make up the shape of the
fuselage and these ribs are fastened together with
balsa spars. This process would prove to be
difficult and time consuming when construction
began, but the group had access to a laser cutter
located in the AMC building on campus.
The machine made it possible for every single
rib that was designed on SolidWorks to be
accurately cut each and every time while saving
a lot of man hours. See Figure 31 for an example
of a fuselage rib.
Figure 32 - Wing Rib Illustration
- 30 -
11. Empennage
The empennage, also known as a “tail
assembly”, is the part of the aircraft responsible
for four important flight characteristics:
Pitch stability
Yaw stability
Pitch control
Yaw control
Pitch stability is achieved by aerodynamic forces
on the horizontal stabilizer (see Figure 33)
producing a restoring pitch moment on the
aircraft when it is angled away from trim.
Yaw stability is achieved in a similar way by
using aerodynamic forces on the vertical
stabilizers to restore the aircraft when perturbed
from the trim state.
Pitch control is provided to the pilot by actuation
of the elevator, a control surface located at the
trailing edge of the horizontal stabilizer.
Deflecting the elevator downward will tend to
pitch the entire aircraft downward and vice
versa.
Yaw control is provided to the pilot by actuation
of the rudders. Rudders are control surfaces
located at the trailing edge of the vertical
stabilizers. Deflecting the rudder right tends to
yaw the aircraft right and vice versa.
Figure 33 - Empennage Diagram
With all four of these desired flight
characteristics in mind, the following design
choices needs to be made for the empennage:
Tail configuration
Horizontal stabilizer airfoil
Horizontal and vertical stabilizer sizing
Empennage position relative to wing
Incidence angle of horizontal stabilizer
11.1 Tail Configuration Selection
The twin-tail configuration was selected as the
tail configuration for the advantages it offered
for our particular aircraft. This configuration
eliminates the effect of the prop wash (a helical
wind around the fuselage caused by the
propeller) on the yaw of the plane. Furthermore,
it allows for a horizontal stabilizer that is wide
and, consequently, has a lot of surface area. The
more surface area the horizontal stabilizer has,
the less distance there needs to be between the
tail and the wings. Shortening that distance is
advantageous since it reduces the length, width,
and height constraint.
11.2 Horizontal Stabilizer Airfoil Selection
Based on advice given by experienced remote
control aircraft hobbyists, the choices of airfoils
were narrowed down to symmetrical airfoils.
These airfoils are called “symmetrical” because
of their symmetry along the chord line.
There are several airfoils in the NACA 4-digit
series that are symmetrical and have
aerodynamic data readily available. The
NACA0012 was chosen because it had just
enough thickness to house the elevator and
rudder servos with a 10-inch chord.
11.3 Horizontal and Vertical Stabilizer
Sizing
The horizontal stabilizer was chosen to have a
10-inch chord length, which make its thickness
enough to hold our servos, as mentioned above.
The width of the horizontal stabilizer was chosen
to be 36 inches simply for convenience of
construction. The balsa leading edges come in
36-inch lengths, so no modifications would need
to be made to its width with a 36-inch horizontal
stabilizer.
With the chord length and width of the
horizontal stabilizer known, the distance known
as the tail moment arm was to be selected. This
is the distance between the quarter-chord points
of the wing and horizontal stabilizer. The choice
of tail moment arm was made so that another
parameter, called the horizontal tail volume
coefficient, falls within a range typical of
successful existing aircraft. The horizontal tail
- 31 -
volume coefficient is defined in equation
Equation 1 below:
Equation 1
where:
SH is the horizontal tail surface area
(chord x width)
lH is the tail moment arm
SW is the surface area of the wings
MAC is the mean aerodynamic chord of
the wing.
This non-dimensional coefficient on well-
behaved aircraft is in the range of 0.3 to 0.6
according to [Equation 1]. The tail moment arm
was chosen such that it gave a value close to the
higher, more stable value of 0.6. The final tail
moment arm is 40 inches.
Vertical stabilizers have an analogous non-
dimensional number called the vertical tail
volume coefficient, with the vertical stabilizer
area in place of the horizontal stabilizer area in
the equation above. This coefficient is typically
in the range of 0.02 to 0.05. The size of the
vertical stabilizers is such that the vertical tail
volume coefficient is 0.16. This was done to
allow for extra yaw control when attempting to
taxi.
11.4 Empennage Position Relative to Wings
The position of the tail assembly is partially
described by the tail moment arm, which is
described above. The height of the tail assembly
above or below the plane of the wings is chosen
so that the tail is not shadowed by the flow
behind the wings.
Since the velocity of the flow in the wake of the
wings is slower, the surfaces in the tail will have
less of an effect if placed there. So, the height of
the empennage is bumped up so that it is well out
of the wake of the wings. A height of 3 inches
was found to be sufficient based on the method
outlined for estimating the position of the wake.
11.5 Incidence Angle
The incidence angle of the horizontal stabilizer is
the angle between the chord line of the
horizontal stabilizer and the roll axis of the body.
The tail boom is parallel to this axis and may be
considered the roll axis for all intents and
purposes.
Incidence angle affects the pitch-up or pitch-
down force on the empennage the same way the
angle of attack of the wing affects the lift on the
wing. The best angle of incidence to have is very
sensitive to the weight of the aircraft and the
position of the aircraft’s center of gravity (CG).
That is to say, choosing a good incidence angle
for the aircraft without any payload might make
the plane very difficult to fly with payload
added.
It is complicated further by the fact that the
location of the CG is difficult, at best, to know
with certainty before the entire plane is designed
assembled.
That difficulty is circumvented by designing for
adjustable tail incidence. With adjustable
incidence, the desired incidence angle can be set
for different payload scenarios and CG locations.
The tail assembly was designed for a range of
incidence angles between positive and negative
30 degrees. shows the mechanism behind the
tail incidence adjustability. The entire tail
assembly pivots around the upper bolt and the
lower bolt is tightened in a slot to fix the angle.
Figure 34 - Tail Angle Adjustability
- 32 -
Thrust vs. RPM Propeller Comparison
0
1
2
3
4
5
6
7
8
6500 7500 8500 9500 10500 11500 12500 13500
RPM
Th
rust
(lb
f)
E 11x7 E 12x6 K Series
12. Preliminary Testing
Preliminary testing was done on the engine of
the aircraft to determine how much thrust force it
could produce. This value was crucial in
determining the size of the wings that were to be
used in order to produce the lift that the aircraft
needed.
This process was done in a crude manner early
on in the 1st semester and then the experiment
was reproduced in the spring as part of the
requirement for the Mech Lab III experiment.
Essentially several different propellers were
tested on the engine that was attached to a cart
with minimal friction that was attached to a
cantilever beam with a strain gage. The beam
was then used as a scale to determine how much
thrust force the engine was producing.
A copy of the Mech Lab III report is attached in
the Appendices. The results of the experiment
are summarized below in Figure 34 and Table 3.
Figure 35 - Thrust Data Comparison
Table 1- Max Thrust Data
Propeller Maximum Thrust
Obtained
@
RPM
Evo 11x7 5.87916 lbf 13338.2
Evo 12x6 5.1496 lbf 11050.4
K-Series Master
Airscrew
5.1824 lbf 11098.1
13. Design Analysis and Review
13.1 Wing
The analysis done with regard to the wing
section utilized is mentioned above, as it was
part of a larger comprehensive analysis
performed to assist in the selection of one
particular airfoil. It is not given here directly, but
instead is treated as part of the wing design as a
whole.
After a careful review of the above compiled
information, it was determined that a relatively
high aspect ratio was desirable. An engineering
decision was made to start by fixing the aspect
ratio a nine for comparison purposes.
It was desired that some experimentation be done
with regard to the effects of twist, taper and
dihedral. Iterations of design analysis was
performed for these parameters as described:
The general characteristics were clear, but their
relative amounts (i.e. amount of twist, starting
point along the span for the taper, etc.) were not.
An abundance of possibilities existed in this
regard. For some measure of accessibility the
starting/ending point for both taper and twist was
set at half of the span. The dihedral was always
started directly at the root, or at the half span.
- 33 -
Increments of adjustment (i.e. 2 degrees, 4
degrees) were kept relatively large and at integer
values.
Upward of fifty different combinations of these
parameters were tested under the same
conditions with XFLR5.
The data was compiled and the total wing lift
coefficients were compared. This data was
required for other purposes in the project as well.
The comparison showed that the minor
differences in the configuration of several of the
desired characteristics had very small effect on
the total lift produced by the wing. Instead, the
largest factor was the airfoil section itself. This
allowed some freedom in adjustment to suit
other needs.
The configuration chosen was ultimately decided
on for a few key reasons. Given that a high
aspect ratio was desired, the wing was tapered
only moderately to avoid tip stall tendencies.
This worked well with efforts to control the stall
progression from the leading edge, as it was
decided to leave the outer half span constant
chord in avoidance of strong taper. A rectangular
section will, in general, stall from the root corner
first, where a tapered section is prone to
separation all along the trailing edge.
The wider chord at the root also provided a
sturdier section from which to support the
moments on the wing/fuselage joint. Excessive
dihedral was not opted for to minimize
construction complexity.
Configuration however, is independent of size.
The final size of the wing was controlled by the
need to generate the necessary lift predicted.
A MATLAB simulation was created to predict
takeoff performance given wing characteristics
and predicted weight. This simulation utilized a
number of values already generated. Preliminary
static thrust tests were performed with the engine
to obtain values of pulling force. The overall
wing lift and drag coefficients as configured
were estimate in XFLR5 for the size range
expected.
A summary of the simulation strategy is given
below:
With the Selig 1223 chosen as the best possible
airfoil, the amount that it can lift needs to be
calculated. The lift is defined as:
Equation 2
The coefficient of lift of 1.48 found as a good
average expected value from the XFLR testing
described above is used for the lift calculation.
The only unknowns in the lift equation are the
velocity and the area. The position (X) can be
defined as a function of time: X(t). The
derivative with respect to time can be defined as
the velocity:
The derivative of the velocity is the acceleration.
So
A free body diagram of the plane is constructed
in Figure 36.
Figure 36- General Plan FBD
Thrust(t) is the thrust of the engine of the plane
as a function of time; however, the thrust will
likely remain constant during takeoff as the
maximum engine thrust.
The Drag(t) is the drag as a function of time
which will increase as the velocity of the aircraft
- 34 -
increases; the drag is a function of the velocity,
which is a function of the time.
FN(t) is the normal force acting on the plane
from the ground. As the vessel is on the ground,
the normal force is the difference between the
weight and the lift of the aircraft. The lift, of
course, is a function of the velocity.
Newton’s second law states the net force acting
on a body is equal to the mass*acceleration of
the body. By summing the forces on the aircraft
with their respective component, the following
equations are formed:
Equation 3
Equation 4
M is the mass of the airplane.
The acceleration in the y-direction will remain 0
until the lift force is higher than the weight of the
aircraft. The acceleration in the x-direction can
be written as below:
Equation 5
Ordinary differential equation methods can be
used to solve equation Equation 5 for the
position, but a MatLab Simulink model was used
instead shown in .
The other unknown in the calculation of lift is
the area. With the aspect ratio at 9, the program
iterates through a series of chord lengths, and the
tapered area is calculated from the chord length
is used.
The mass of the aircraft is also currently
unknown and is iterated in the program. Since
the thrust from the experiment is only 4.88 lbf
and other reports have the same engine going up
to about 8 lbf, the program iterates through a
range of thrusts as well.
With the position, velocity, and acceleration
given as a function of time, the lift can be
calculated as a function of time. The program
calculates the lift over time, but the most
important time is that at which the aircraft
reaches the takeoff distance of 200 feet. Figure
38 shows the result of the combination of thrust,
weight, and chord length that has the lift force
higher than the weight of the plane.
The team chose a chord length of 1.333 feet and
wingspan of 10.2 feet. With the thrust provided
from the experiment, the plane should
theoretically be able to lift over 35 lbf. If the
thrust experiment was off, and the actual thrust
of the engine is more, perhaps closer to 8 lbf, the
plane will be able to lift almost 57 lbf with the
selected dimensions. Based on later results, this
appears to be a gross overestimate.
However, the calculations were sufficient to
serve as a starting point for continuing design, as
well as refining the aircraft at later points.
After the final geometry was decided on,
computational fluid dynamics software was
utilized to generate results. Specifically, this was
done with Ansys Fluent. Two different programs
were used as a check on the reliability of the data
generated. The results garnered from this were
values of lift coefficient, drag coefficient, and
moment coefficient.
In order to do this analysis, a solid model of the
wing was needed, so as to be imported. The wing
was modeled using SolidWorks to accomplish
this. The final SolidWorks model of the wing is
shown below.
Analysis was done the final geometry throughout
the same range of Reynolds numbers used in the
previous CFD considerations. Angles of attack
were 1, 5, 10, and 15 degrees. Given that the
results from Ansys were based on a finite wing
geometry, it was expected that coefficient data
would be somewhat lower than that of the 2-D
analysis.
This is due to the inefficiencies associated with
air spillage and other wing-tip effects, as well as
the other induced drag effects. Numbers seen
with Ansys are approximately 80 percent of
those from XFLR5, which is sensible.
- 35 -
Figure 37 - Takeoff Simulink Schematic
Figure 38 - Simulink Results
- 36 -
Figure 39 - Velocity Streamlines
Figure 40 - Boundary Layer
- 37 -
Figure 41 - Velocity Contours
Figure 42 - Pressure Contours
- 38 -
Figure 43 - Turbulence
Figure 44 - Airfoil Mesh
- 39 -
Ansys gives results for the total lift
produced by the wing, the total drag on
the wing and many other
considerations. However the data output
is in a coordinate system referenced by
Ansys during computation.
These results were converted to an appropriately
configured coordinate system using a MathCAD
worksheet. Comparisons have been made to both
the data gathered for the airfoil sections by our
group, as well as published data for the Selig
S1223. As expected, the 3-dimensional wing
underperforms the 2-D analysis. However, the
lift figures are aligned with our necessities.
A full scale mockup of the wing was constructed
to gain experience in practical fabrication. The
layout of this differed considerably from the final
design. A wooden dowel was used in place of
stock leading edges, the ribs were all ¼ of an
inch, and the elevator was considerable larger.
An angled jig was used to set the appropriate
twist and dihedral, where in later models, the
mechanical twist was accomplished solely by
proper implementation of rib height.
The largest difference was seen in the cutting
technique. The mockup was fabricated
completely by hand whereas the final design was
cut using the laserjet printer housed at the
advanced manufacturing center. This was
accomplished by importing the solid models
made in SolidWorks to software known as
CorelDraw. These files are read by the laser
printer and cut to scale. Overall, the mockup was
an effective tool for the group to refine its
construction technique before the final built.
13.2 Wing Structural Analysis
Another primary consideration of the wing
design was the capabilities of the wing structure.
Balsa wood, while very lightweight, is inherently
weak, especially transverse to the grain. It was
critical that the wing be able to support adequate
distributed loads upward. The wing as designed,
and as built, is in reality quite a complex
geometry. The joint at the fuselage where the
beam members sleeve together supports bending.
Examining the load path, one will see that the
distributed load generated by aerodynamic forces
causes the wing to lift upward. This load is
transferred to the aforementioned beam members
and sustained internally. The wing is effectively
a cantilever beam fixed at one end. The largest
stresses are in the structural components which
allow for connection to the fuselage.
The leading edge is a stiff member which gives
the wing rigidity, but there is no load path from
the leading edge directly to the fuselage.
All twisting loads are transferred through the
ribs. Aside from that consideration, the main
purpose of the ribs is to provide a platform for
the spars. Thus, while the wing can be
conceptualized as a cantilever beam, there are
inherent differences.
Initially, the attachment apparatus consisted of
two sleeved PVC pipes. This arrangement was
set at some distance from the main spars of the
wing. The load transfer between these two
components was a large source of difficulty in
analysis. The parts of the wing (the PVC and the
balsa) also obviously have different properties.
The balsa itself is an anisotropic material,
meaning that its strength is dependent on the
direction in which it is loaded. Simplifications
were made by finding average properties
weighted by the cross sectional area percentage
along the span, and also by finding an effective
moment of inertia of the “beam”.
Attempted analyses using the fluid/structure
interaction capabilities of Ansys were
unsuccessful. Eventually, for several reasons, the
beam structure was altered.
The second design utilizes lengths of thin-walled
aluminum tubing mounted in the wings and
fuselage which nest together as with the PVC.
This tubing is rectangular in cross section. The
second version moves this into alignment with
the built-up beam structure in the outer portion
of the wing, and space is cut in the ribs to allow
for this. The built-up beam is comprised of the
upper and lower spars running the length of each
wing, bonded to balsa sheeting on either side of
the spars. This makes for a thin-walled
rectangular structure.
13.2.1 Analysis With Design 2
A finite element analysis was set up using the
FEA 3-in-1 code given to UMaine students in
their Finite Element Analysis course. The
analysis was limited to under 100 elements, and
due to the simplifications necessary, results are
- 40 -
Vertical Displacement vs. X Position
-1.00E-01
0.00E+00
1.00E-01
2.00E-01
3.00E-01
4.00E-01
5.00E-01
6.00E-01
7.00E-01
8.00E-01
0.00E+0
0
1.00E+0
1
2.00E+0
1
3.00E+0
1
4.00E+0
1
5.00E+0
1
6.00E+0
1
7.00E+0
1
X Position (inches)
Ver
tica
l Dis
pla
cem
ent
(in
ches
)
Displacement vs. X Position
ultimately not overly reliable. However, they do
provide some sense of scope. The results are as
follows in Figure 45:
Figure 45 - FEA 3 in 1 Wing Deflection
13.3 Empennage
The empennage sufficiently performed its duties
on its maiden flight on May 6, 2013. Pitch and
yaw stability and control were achieved in flight.
However, pitch control was hindered by the tail-
heaviness of the airplane causing the aircraft to
want to pitch up. One improvement that could be
made to the empennage would be to remove
weight so that the CG was further forward.
Another area of improvement would be the
control linkage for the rudders. As is, the rudders
do not deflect enough to yaw the aircraft while
taxiing at reasonable speeds. Redesign of this
mechanism would lead to better overall yaw
control.
13.4 Servo Performance
An analysis was conducted to determine whether
or not the servo would hold elevator position at
high airspeed. Should the drag on the control
surface impart a torque on the servo arm that
exceeds the servo’s rated torque, elevator control
would be lost.
The analysis, performed in a Mathcad sheet
shown in the Appendices, solves for link
positions based on user input for servo arm
angle. Then, it uses force and moment
equilibrium to solve for the applied torque on the
servo arm. Even for unrealistically bad scenarios
(high speed, concentrated aerodynamic forces,
etc.), the servo operates safely within its design
range as expected.
13.5 Weight
The total weight of the aircraft with no payload
inside it is 13.75lb
13.5.1 Predicted Performance
The overall performance of the aircraft is not
quite at the level that the group had hoped as the
engine that is used in the design does not
produce as much thrust as desired. Other teams
in previous years had engines that produced
upwards of 8lbs of thrust determined from
testing.
As stated earlier, the Magnum XLS .61A that is
utilized in this aircraft produced only 5.8lbs of
thrust at its peak. This could be an error in
testing, an error in the testing rig/interface of
LabView or even simply just a weak engine in
comparison to other groups. Never the less, the
thrust force that was determined through testing
drove the results that are summarized below in
the predicted performance of the aircraft.
- 41 -
These results are produced by a Flight
Simulation Matlab code generated by one of the
team members. This code can be seen in the
Appendices.
- Max Thrust: 5.8lbs
This number is generated from the 11x7
propeller at its peak. However, in the final design
the aircraft will be using the 12x6 propeller
because as stated in the conclusions of the Mech
Lab III final report, the 12x6 had overall more
stable and better performance.
That being said, the maximum thrust that the
final design configuration will produce is
5.15lbs.
- Predicted Lift: 35lbs
Unfortunately this number is a bit lower than
what was hoped for, 55lbs. Compared to other
teams, 35lbs is a competitive number. This will
be explained below.
- Predicted Payload
The aircraft itself weighs 13.75lbs and it can lift
35lbs. Thus, the total payload the aircraft should
be able to lift is 21.25lbs. Winning teams in past
years have won the competition by lifting 15lb-
25lbs so by design, this aircraft is competitive.
14. Final Testing and Evaluation
14.1 Testing
Brian Barainca, President and Founder of the
Black Bearons Flying Club on campus was the
test pilot of the team’s aircraft. He has been
flying for a number of years and is more than
qualified. No member of the group had even
remotely enough experience with flying RC
aircraft to be comfortable trying to fly.
Our testing plan included the following:
- Test the aircraft with no payload added.
Take off, complete a 360° circuit, and land
successfully.
- Adjust the aircraft as needed and repeat the
circuit.
- Load the aircraft with payload in
increments of 1lb until take off is
unachievable.
Let it be noted that the final aircraft as built is
not a perfectly balanced aircraft in regards to the
pitch axis. The center of gravity of the aircraft is
slightly aft of the quarter chord line of the wings
which causes the aircraft to be tail heavy. A
perfectly balanced aircraft could be held up at
the quarter chord point and be perfectly
balanced. But the final design as built tends to
have more weight toward the tail. This caused
the aircraft’s pitch to be difficult to control.
The test day was unfortunately very windy (7-
12mph winds) which as explained later on,
seemed to cause some difficulty.
Before testing, the empennage was off kilter for
some reason and needed to be adjusted so as to
align the horizontal stabilizer parallel with the
ground. Upon trying to adjust this, the PVC pipe
came loose and broke all of the support rings on
each of the fuselage ribs which would have been
catastrophic if it had happened during flight.
Some emergency repairs were made using
packing tape and quick dry epoxy that repaired
and stiffened the empennage-fuselage
connection. Unfortunately the repairs added even
more weight to the aircraft. See Figure 46
Figure 46 - Emergency Repairs
This repair fixed the connection but
unfortunately the empennage was still off at a
slight angle. See Figure 47.
- 42 -
Figure 47 - Off-kilter empennage
None the less, flight testing resumed.
14.2 Results
With the wind blowing into the face of the
aircraft, the aircraft had no problem taking off.
The competition requires take off within 200ft
and our aircraft with no payload in it took off in
less than 40ft.
With the crooked empennage and a heavy tail,
the aircraft was difficult to control in the air. Of
course, the high winds did not help this situation
any. But as competition required, a full 360º turn
was made and the landing approach began.
While trying to land the aircraft, the wind picked
up and blew the aircraft of course and spun the
tail end around, causing a very rough landing
which upon impact, broke off the landing gear
from under the fuselage. See Figure 48.
Figure 48 - Broken Landing Gear
However, the landing gear was still attached to
the plane by some Monokote, so by competition
standards, that would have been a successful
landing. As long as nothing on the plane
becomes completely disconnected from the
aircraft, it is considered a successful landing.
Needless to say, flight testing ended abruptly due
to no landing gear for the plane to take off or
land on. There was no chance to load the aircraft
with payload and attempt flight with extra
weight.
One successful flight was made with the aircraft
with no payload in the aircraft. According to
competition rules, the team would receive bonus
points for completing a circuit with no payload.
No other points would be rewarded during the
flight part of the competition however because
the aircraft was not fit to fly any more circuits
without major repairs.
14.3 Evaluation
14.3.1 Engine
The engine had plenty of power had more than
enough power for the purposes of getting the
aircraft off the ground.
14.3.2 Wings
The wings were by far the most thoroughly and
best designed components of the aircraft. They
held together well and connected to the fuselage
with ease and sturdiness. Every aspect of the
wings were designed well. According to the
pilot, Brian Barainca, “Even with no wind
assistance, the aircraft would have taken off no
problem. The wings produced more than enough
lift, the thing wanted to climb.” He also stated
that the flaps were sized correctly, but the
ailerons could have been sized a little bigger to
help alleviate the control issues with the tail
heavy design. The wings of the aircraft were
designed exactly how they needed to be
designed.
14.3.3 Fuselage
The fuselage design was not as strong as it
needed to be. A lot of weight was saved by
having as few ribs as possible and by not
sheeting the fuselage with balsa like much of the
wing was, but in doing so the structural integrity
of it was sacrificed and this lead to the weak tail
end of it that broke so easily during adjustments
pre-flight. The repairs had to be made on site,
using epoxy which also added to the weight of
the back end of the aircraft, making it even more
tail heavy. Looking further back to the design on
the fuselage ribs, no fillets were used at the
corners which led to high stress concentrations at
these sharp 90º corners and were a point of
failure. The fuselage was under-designed and as
a result, was a point of failure.
- 43 -
14.3.4 Wireless Systems
The wireless systems acted just as they should
have and provided quick, reliable response time
between commands.
14.3.5 Controls
As evident in the analysis of the servos, the
controls were sized properly and the servos could
easily handle the loads that were experienced
during flight.
14.3.6 Empennage
The final design of the aircraft was a little tail
heavy which lead to less than stable flight. That
being said, the tail boom and tail assembly could
have been designed to be a little lighter to solve
this problem. “A tail heavy aircraft is difficult to
fly” according to Brian, and a result the aircraft
was a little difficult to control during flight.
However, the elevator on the horizontal stabilizer
was definitely big enough if the aircraft had been
properly balanced. The same could be said about
the rudders on the vertical stabilizers. The
control surfaces on the empennage were sized
properly but the overall weight of it was too
much.
14.3.7 Landing Gear/Externals
The landing gear was an under designed portion
of the project as can be seen by their failure
during landing. The connection between the
landing gear and the bottom of the fuselage was
not strong enough to withstand large loads so
when the aircraft landed roughly, they failed and
broke off.
As for the decision to use Monokote for the
outside of the aircraft, this was an excellent
choice as the surfaces of the aircraft are very
smooth and the final design is very aesthetically
pleasing.
14.4 Evaluation Summary
Overall, the aircraft was designed to competition
specifications. The aircraft would have no
problem passing the inspections of the
competition judges. This was a successful design
process. During testing however, some
components proved to be under-designed, as
explained above, and this led to issues.
This project met the standards of the competition
and guidelines it was supposed and although the
aircraft might not have scored well, the project
can be looked at as a success.
14.5 Conclusion
Upon completion and evaluation of the project, it
can be concluded that the design met all
specifications required by the SAE Aero Design
competition rules and speculations. However,
there are a couple of reflection points to touch
on.
The design process, as stated was executed
correctly and was well suited for the tasks at
hand. That being said, the design process was
carried out slower than it should have been. The
entire first semester set the team back
considerably in terms of design. As the first
iteration of the SAE Aero team that UMaine has
seen, the group had to get familiarized with all
rules and restrictions of the competition. Then
the group had to determine how this project
would be carried out and the process of
designing an aircraft from the ground up.
Once a process was determined, the largest
chunk of the first semester went into the design
of the wing. By the end of the first semester, a
thorough, well thought-out design of the wing
was complete. These extensive efforts were
evident in the success of the wings in flight. So
being the component that had the most time and
effort put into, the wing indeed was the most
successful piece of the project.
By spending so much time on the wing design,
other components of the aircraft lacked the time
and effort that the wing got. Both the fuselage
and the empennage, as stated in the evaluation,
were under designed and as a result had some
issues during final testing.
If a team were to come into this project in
later years there are a few areas that could be
improved on. These areas are as follows:
Strengthen the Fuselage:
Sacrificing strength for weight was not a good
trade off. Filet all corners, add more ribs, spars,
and sheet everything. Balsa is a very weak wood
so the more ribs, the better. Also, pay attention
to the grain orientation of each rib when
laser-cutting and assembling, balsa is
- 44 -
anisotropic so it is stronger with the grain
than it is transverse to the grains.
Lighten the Empennage:
Although PVC was easy to work with, it was a
bit too heavy and caused problems when
balancing the aircraft.
More Design Efforts on the Landing Gear:
The landing gear was a point of weakness in the
design and as a result, failed upon landing.
Wing Design:
Although the wing design came out great, more
forethought should be put into the construction
process because taper and twist in the
configuration caused headaches when
construction began and the benefits in flight were
not quite worth the trouble.
Balance:
Pay more attention to balancing the aircraft when
designing it. An unbalanced aircraft is difficult to
fly and was evident in flight testing.
The team could have had better time
management and organizational structure
throughout the process to make for a more
successful final product but regardless, the
project did indeed meet the specifications
required and is considered a successful endeavor
by the team as a whole. With these
recommendations, future year’s SAE Aero team
will be even more successful. As for a first
iteration of this project, the team completed what
it set out to do and produced a competition
worthy aircraft.
- 45 -
Figure 49 - Take Off During Flight Testing
Figure 50 - Completing Aerial Circuit During Flight
Testing
- 46 -
15. References
Abbott, Ira H., and Albert E. Von. Doenhoff. Theory of Wing Sections: Including a Summary of Airfoil
Data. New York, NY: Dover Publ., 1982. Print.
Aerostudents. N.p., n.d. Web.
Anderson, David F., and Scott Eberhardt. Understanding Flight. New York: McGraw-Hill, 2001. Print.
Anderson, John D. Fundamentals of Aerodynamics. New York: McGraw-Hill, 2011. Print.
Anderson, John D. Introduction to Flight. New York: McGraw Hill, 2012. Print.
Ashley, Holt. Engineering Analysis of Flight Vehicles. New York: Dover, 1992. Print.
Barainca, B. (2013, May 6), Personal Interview
Basic Aircraft Design Rules, Unified Teaching Staff, Massachusetts Institute of Technology
http://ocw.mit.edu/courses/aeronautics-and-astronautics/16-01-unified-engineering-i-ii-iii-iv-fall-
2005-spring-2006/systems-labs-06/spl8.pdf
Bertin, John J., and Michael L. Smith. Aerodynamics for Engineers. Englewood Cliffs, NJ: Prentice-Hall,
1979. Print.
Bisplinghoff, Raymond L., Holt Ashley, and Robert L. Halfman. Aeroelasticity. New York: Dover
Publications, 1996. Print.
Bruhn, E. F. Analysis and Design of Flight Vehicle Structures. Cincinnati, OH: Tri-State Offset, 1965.
Print.
Etkin, Bernard, and Lloyd D. Reid. Dynamics of Flight: Stability and Control. New York: Wiley, 1996.
Print.
Fundamentals of Engineering Supplied-reference Handbook. Clemson, SC: National Council of Examiners
for Engineering and Surveying, 2011. Print.
Gere, James M., and Barry J. Goodno. Mechanics of Materials. Mason, OH: Cengage Learning, 2009.
Print.
Grant, Charles Hampson. Model Airplane Design and Theory of Flight; a Complete Exposition of the
Aerodynamics and Design of Flying Model Aircraft, with Fundamental Rules, Formulas and
Graphs. New York: Jay Corporation, 1941. Print.
- 47 -
Hartog, Jacob P. Den. Strength of Materials. New York: Dover Publ., 1977. Print.
Hepperle, Martin. "MainFrame." MainFrame. N.p., n.d. Web.
Hibbeler, R. C. Engineering Mechanics: Statics. Upper Saddle River, NJ: Prentice Hall, 2010. Print.
Hibbeler, Russell C. Engineering Mechanics: Dynamics. Upper Saddle River, NJ: Prentice Hall, 2010.
Print.
Hoerner, Sighard F., and Henry V. Borst. Fluid-dynamic Lift: Practical Information on Aerodynamic and
Hydrodynamic Lift. Brick Town, NJ: L.A. Hoerner, 1975. Print.
Hoerner, Sighard F. Fluid-dynamic Drag: Practical Information on Aerodynamic Drag and Hydrodynamic
Resistance. Alburqueque, NM: Db Hoerner Fluid Dynamics, 1965. Print.
rm n, Theodore Von. Aerodynamics: Selected Topics in the Light of Their Historical Development.
Mineola, NY: Dover Publications, 2004. Print.
Kermode, Alfred Cotterill., R. H. Barnard, and D. R. Philpott. Mechanics of Flight. Harlow, England:
Pearson Prentice Hall, 2006. Print.
Lennon, Andy. RC Model Airplane Design. Osceola, WI, USA: Motor International, 1986. Print.
Megson, T. H. G., and T. H. G. Megson. An Introduction to Aircraft Structural Analysis. Amsterdam:
Butterworth-Heinemann/Elsevier, 2010. Print.
Milne-Thomson, L. M. Theoretical Aerodynamics. New York: Dover Publications, 1973. Print.
Mises, Richard Von. Theory of Flight. New York: Dover Publications, 1959. Print.
Niu, Ch un-y n. Airframe Structural Design: Practical Design Information and Data on Aircraft
Structures. Hong Kong: Conmilit, 1988. Print.
Norton, Robert L. Machine Design: An Integrated Approach. Boston: Prentice Hall, 2011. Print.
Nyhoff, Larry R., and Sanford Leestma. FORTRAN 77 for Engineers and Scientists: With an Introduction
to FORTRAN 90. Upper Saddle River, NJ: Prentice Hall, 1996. Print.
P., Den Hartog J. Advanced Strength of Materials. New York: Dover Publications, 1987. Print.
P., Den Hartog J. Mechanics. New York: Dover Publications, 1961. Print.
Peery, David J. Aircraft Structures. New York: McGraw-Hill, 1950. Print.
ProAdvice 2: THE WING PLANFORM (n.d.): n. pag. Great Owl Publishing, 2010. Web.
ProAdvice 3: AILERON SIZING (n.d.): n. pag. Great Owl Publishing, 2010. Web.
- 48 -
Rao, Singiresu S. Mechanical Vibrations. Upper Saddle River, NJ: Prentice Hall, 2011. Print.
Raymer, Daniel P. Aircraft Design: A Conceptual Approach. Reston, VA: American Institute of
Aeronautics and Astronautics, 1999. Print.
Rivello, Robert M. Theory and Analysis of Flight Structures. New York: McGraw-Hill, 1969. Print.
Selig, Michael S., and James J. Guglielmo. "High-Lift Low Reynolds Number Airfoil Design." Journal of
Aircraft 34.1 (1997): 72-79. Web.
Selig, Michael S. Low Reynolds Number Airfoil Design Lecture Notes (n.d.): n. pag. Web.
Selig, Michael S. Summary of Low Speed Airfoil Data. Virginia Beach, VA: SoarTech Publications, 1995.
Print.
Shapiro, Ascher H. Shape and Flow; the Fluid Dynamics of Drag. Garden City, NY: Anchor, 1961. Print.
Shevell, Richard Shepherd. Fundamentals of Flight. Englewood Cliffs, NJ: Prentice Hall, 1989. Print.
Simons, Martin. Model Aircraft Aerodynamics. Swanley: Nexus Special Interests, 1999. Print.
"UIUC Airfoil Datasite." N.p., n.d. Web.
White, Frank M. Fluid Mechanics. New York, NY: McGraw Hill, 2011. Print.
- 49 -
16. Appendices
16.1 Appendix A - Plans and Specs
- 50 -
16.2 Appendix B - Wind Tunnel Modification and Testing
As part of the design testing, the group wanted to utilize the wind tunnel in Crosby Lab as a resource to
collect realistic data during wing design and configuration stages of the process.
16.2.1 Design Goals
Access to the wind tunnel in the Air Flow Lab in Crosby Hall gave the group an opportunity to test model
wing configurations and compare the results to our 2D XFLR5 data. This experimental data would provide
more realistic numbers for aerodynamic coefficients (such as lift coefficient, drag coefficient, etc.) than the
XFLR5 data which ignore the aerodynamic effects of finite wing span. The goals of the wind tunnel testing
are as follows:
Collect aerodynamic data for different wing configurations over the design range of Reynolds
numbers between 50,000 and 500,000.
Compare performance of different wing configurations and select the optimal configuration for
our craft.
16.2.2 Research
16.2.2.1 EXPLORING THE WIND TUNNEL FLOW
Once the group agreed that the wind tunnel experiments would be valuable, the flow in the tunnel was
probed with a hot film anemometer to determine the characteristics of the flow. The results made clear that
the current setup, shown in Figure 51, of the wind tunnel was inadequate for the following reasons:
Figure 51 - Current Wind Tunnel Flows
Airspeed was too slow to match design Reynolds numbers with reasonably sized models.
Flow just downstream of the fan swirled significantly and passed through a sudden enlargement,
causing a wasteful pressure loss and recirculating flow.
The velocity was very unsteady throughout the tunnel, meaning the quality experimental results
would be diminished by vibrational effects.
In order to achieve the flow characteristics desired, modifications would need to be made to the wind
tunnel.
16.2.2.2 DESIGNING WIND TUNNEL MODIFICATIONS
The flow needed to be conditioned in some way to provide the flow characteristics necessary for the
experiments. These modifications, shown in Figure 52, were proposed as additions to the wind tunnel to
improve flow quality and speed.
- 51 -
Figure 52 - Flow Straightening Proposal
Flow straighteners – to remove excessive swirling and free stream turbulence
A converging nozzle section – to accelerate flow to desired speed
A properly sized test section – to contain the test specimen
A diffuser – to efficiently decelerate flow
Figure 53 - Flow Mapping Rig
To get an estimate of what the velocity of the flow would be at any part of the modified tunnel, an estimate
of the mass flow rate throughout the tunnel was needed. This was obtained by mapping the flow in the 3ft
by 3ft square hole downstream from the fan.
A grid was made out of kite string and duct tape to indicate where the 25 equally-spaced sample points
would be to probe the flow. The intersections of grid strings indicate measurement points as shown in
Figure 53.
The results, shown in Figure 54, show the non-uniformity of the speed over the cross section. The average
speed of 91.3 ft/s was used for calculating the average mass flow rate through the tunnel.
The required size of the test section was dictated by several design constraints:
Mach number in test section must not exceed 0.3 to avoid compressibility effects.
Velocity in the test section must yield Reynolds numbers throughout the design range from 50k to
500k.
- 52 -
The size of the test section should be large enough to accommodate wings with chord lengths
sized to meet the Reynolds number criteria.
Width of the test section should be 150% larger than the wingspan (eight to nine times the max
chord) to avoid side wall interference.
Height of the test section should be five times the max chord to avoid top and bottom wall
interference.
Figure 54 - Velocity Contour at Outlet
16.2.3 Design Process
16.2.3.1 SIMPLE ANALYSIS
The cross section of the test section was kept rectangular for simplicity. An acceptable range for height and
width of the cross section was found by formulating the constraints on size (the five stated above) as
mathematical statements and solving for the dimensions. The dimensions of 36”x10” were within the
ranges and were chosen so that the nozzle only needed to converge vertically.
16.2.3.2 CFD ANALYSIS OF DUCT DESIGN
The duct was modeled in Solidworks and exported to ANSYS FLUENT for a detailed analysis of the flow
through the duct geometry. The inlet boundary condition was simply set to constant average velocity and
the flow direction and speed were visualized in CFD-Post as shown in Figure 55 and Figure 56.
The results of the CFD analysis indicated that speeds encountered in the test section might be higher than
expected and not as uniform over the entire cross section.
Figure 55 shows the streamlines through the duct and a velocity contour on the exit to show the
distribution of airspeed.
Figure 56 shows the contours of gage pressure on the duct walls. The pressure integrated over the bottom of
the converging duct created an aerodynamic force of nearly 1300 pounds, suggesting the need for a brace
structure underneath.
- 53 -
Figure 55 - Flow Streamlines Through Duct
Figure 56 - Pressure Contours Along Duct Wall
16.2.3.3 MORE SOPHISTICATED DESIGN
After an advisor meeting with Professor Boyle, where the wind tunnel modification ideas were discussed,
he suggested that the fan would perform differently than expected since it would be pushing air through a
different pressure drop. That is, the fan would move a larger or smaller mass flow rate through the duct
depending on the pressure drop it drives against. This consideration suggested the need for more fluid
mechanics related information including:
Fan performance data for the Joy Manufacturing Axivane Fan (used in the Crosby wind tunnel)
Minor loss coefficients for the modifications placed in the duct
And a resource for useful tips on low speed wind tunnel design.
16.2.4 Abandonment of Modification Idea
Unfortunately, the necessary resources that actually existed were unavailable until sometime in the spring
of 2013. This was determined to be too much of a set-back and the wind tunnel modification was
abandoned altogether. Modifications to the wind tunnel that would have cost hundreds of dollars and hours
of construction were too risky if they might not yield the flow needed for testing.
16.2.5 Final Results
The data from XFLR5 was settled on as what would lead to the decision of what wing configuration to use.
Moving forward with that data had less associated risk than the wind tunnel modification and testing idea,
so it was regrettably abandoned.
- 54 -
16.3 Appendix C - Budget and Costs
Preliminary Research and Design Expenses
Date Approved QTY Description Unit $ Total $
Purch
Req #
9/28/2012 1 Magnum XLS-61 $99.99 $99.99 1
9/28/2012 1 Magnum Muffler XLS-61 $25.00 $25.00 1
9/28/2012 1 Dubro Fuel Tank 14oz. $5.99 $5.99 1
9/28/2012 1 Windsor 13x8 K-Series Propeller $4.79 $4.79 1
9/28/2012 1 APC 12x6 Propeller $4.25 $4.25 1
9/28/2012 1 APC 11x7 Propeller $2.99 $2.99 1
9/28/2012 1 APC 11x7 3-blade K-Series Propeller $9.49 $9.49 1
9/28/2012 1 Tower Power 15% Airplane Fuel $7.29 $7.29 2
10/17/2012 1 12V 4.5A Starter Battery $40.46 $40.46 3
10/17/2012 1 AA Slim Glow Starter $10.49 $10.49 3
10/17/2012 1 Torq Master 180 Heavy Duty 12 Volt $44.99 $44.99 3
10/17/2012 1 After Run Engine Oil 2 Fl Oz $10.11 $10.11 3
10/17/2012 1 Ultra Precision Fuel Filter $3.79 $3.79 3
10/17/2012 1 Flite Power Point Propeller Balancer $19.99 $19.99 3
10/17/2012 1 Spinner Set w/ Prop Adapter Switch $6.29 $6.29 3
10/17/2012 8 Great Planes Nylon Control Horns Large $1.19 $9.52 3
10/17/2012 1 Sullivan Push Cable $5.59 $5.59 3
10/17/2012 1 Dubro Quik-Fill Fuel Pump $21.70 $21.70 3
11/27/2012 1 Assorted Balsa $56.49 $56.49 4
11/27/2012 1 Model Aircraft Epoxy $12.79 $12.79 4
1/14/2013 1 Engine Mount $4.49 $4.49 5
1/14/2013 1 4.8V Battery $19.99 $19.99 5
1/14/2013 1 Mid Cure Epoxy $7.19 $7.19 5
1/14/2013 1 CA Glue $6.99 $6.99 5
1/14/2013 6 Clevis/Push Rod $6.90 $41.40 5
1/14/2013 6 Servos $13.99 $83.94 5
1/14/2013 1 Spektrum Dxi6 Transmitter & Reciever $194.99 $194.99 5
2/12/2013 1 Monokote $11.99 $11.99 6
2/12/2013 2 Monokote $12.99 $25.98 6
Total Spent Less Shipping: $798.96
Shipping Cost: $22.98
Total Spent on Design and Research: $821.94
- 55 -
Build Cost
Date Appoved QTY Description Unit $ Total $
Purch
Req #
2/12/2013 1 Asorted Balsa Wood $114.60 $114.60 6
2/12/2013 2 Epoxy $7.19 $14.38 6
2/12/2013 15 1/4" x 1/4" Squre Dowel $0.61 $9.15 6
2/12/2013 2 6 Pack Hinges $3.53 $7.06 6
2/12/2013 1 15 Pack Hinges $6.12 $6.12 6
2/12/2013 1 6 oz Gas Tank $4.35 $4.35 6
2/12/2013 1 Sealing Iron $19.99 $19.99 6
2/12/2013 1 Servo $13.99 $13.99 6
2/12/2013 2 Y Servo Wire Splitter $6.29 $12.58 6
2/12/2013 2 CA Glue $6.00 $12 6
2/12/2013 3 36" Servo Wire Extension $4.79 $14.37 6
2/12/2013 2 18" Servo Wire Extension $3.49 $6.98 6
2/12/2013 1 12" Servo Wire Extension $2.99 $2.99 6
2/12/2013 1 Servo $13.99 $13.99 6
2/12/2013 1 36" Servo Wire Extension $4.79 $4.79 6
2/12/2013 2 Flat Hinge Packs $5.00 $10 6
2/12/2013 4 18" Servo Wire Extensions $3.00 $12 6
2/12/2013 2 Servo Horn Extensions $3.00 $6 6
2/12/2013 1 Assorted Balsa $25.40 $25.40 6
3/28/2013 2 3/4" x 3/4" Aluminum Square Tube $2.95 $5.90 7
3/28/2013 2 1" x 1" Aluminum Square Tube $3.28 $6.56 7
4/17/2013 5 Monokote, Black, LXHV26 $13.99 $69.95 8
4/17/2013 1 Monokote, Trans Blue, LXHW63 $59.99 $59.99 8
Total Spent Less Shipping $453.14
Shipping Costs: $13.51
Total Spent Build Cost: $466.65
Total Spent Less Shipping $1,252.10
Total Shipping Costs: $36.49
SAE Aero Design Total Spent: $1,288.59
- 56 -
16.4 Appendix D - Mech. Lab III (MEE 443) Report
Static Thrust and Fuel Consumption Testing
Of A Magnum XLS .61A Engine with Several Propellers
Final Experiment Report
Crosby Laboratory
University of Maine, Orono
SAE Aero Design Group
David Chandpen
Travis Cushman
Matthew Maberry
Joseph Travaglini
Zachary Veilleux
Benjamin Waller
MEE 443
April / May 2013
- 57 -
Table of Contents
Table of Contents………………………………………………………………………………….. 57
Introduction………………………………………………………………………………………… 58
Experimental Objectives…………………………………………………………………………. 59
Apparatus, Equipment, and Instrumentation………………………………………………….. 60
Experimental Theory……………………………………………………………………………… 62
Experimental Procedure…………………………………………………………………………. 65
Experimental Results and Conclusions……………………………………………………..…. 67
Appendix A: Uncertainty Level Buildup…………………………………………………..…… 73
Appendix B: Beam Calibration……………………………………………………………..…… 77
- 58 -
16.4.1 Introduction
16.4.1.1 PROJECT DESCRIPTION
The MEE capstone project group authoring this report is the SAE Aero Design Team. The members
comprising this team are Ben Waller, David Chandpen, Joe Travaglini, Matt Maberry, Travis Cushman,
and Zach Veilleux. The goals of the group with respect to the project undertaken are guided by the rules
and restrictions of the competition involved. Specifically, the SAE Aero Design competition held annually
as an international event for college groups. This competition requires the design, testing, construction and
flight of small scale radio-controlled aerial vehicle. This is essentially a heavy lift competition, given that
points are awarded for the amount of mass carried around an aerial circuit with restraints on take off and
landing distances. The maximum weight of any vehicle is 55 pounds, and thus designing the lightest
possible aircraft capable of takeoff with desired payload is necessary. The powerplant of the aircraft must
be a designated 0.61 cubic inch displacement two stroke model aircraft engine running on nitromethane
fuel. Size restrictions on the vehicle are only a maximum combined length-width-height measurement of
225 inches. The main restriction on materials for construction is the outlaw of fiber reinforced plastics. The
vehicle must obtain lift under its own power within 200 feet, and must safely land again in a space of 400
feet. Accurate prediction of the aircraft capabilities are essential, as additional points are given for closely
matching the limits of theoretical lifting ability. These limitations and demands focus the design work of
the team, and have required a range of engineering considerations, from aerodynamic characteristics and
loads, to structural design planning for component integration. It is decided at this time that entering the
competition is not feasible for several reasons. However, the goal of the group remains to produce a
competition-worthy vessel of decent performance. Figure 1 shows an example of the type of aircraft
typically designed for the competition.
16.4.1.2 INFORMATION DESIRED
The experimental work constituting the MEE 443 project which has been performed involved measurement
of the capabilities of the aircraft power plant. It was desirable to obtain an accurate and real knowledge of
the thrust performance provided by the aircraft’s engine, which was already acquired, when mated with
several different propellers of slight variation. The amount of force with which the engine and propeller
combinations can pull is critical information for the project. While the aircraft has been kept light in
weight, there is undoubtedly be friction and inertia to be overcome on upon run up and take off.
Aerodynamic data regarding the chosen wing section and designed wing are known, and a target therefore
exists in terms of the speed required for takeoff. The thrust of the engine is the key parameter determining
whether this can be accomplished inside of the limited distance afforded by the competition rules. The
engine thrust also plays a large role in the performance of the aircraft once it is aloft and relying on the
energy transferred from the engine to the air as a motive force. Design of the aircraft proceeded with
technical estimates and predictions of the engine performance, but experimental data is desired.
Specifically, the group has produced a number of data curves, plotting pulling force of the engine/propeller
combination versus RPM, for several different propellers. The competition rules do not designate any
specific propeller for usage. That is, no propeller is mandated. The group acquired several propellers of
varying pitch and diameter, including one with three blades as opposed to the standard two-blade jobs. The
group has prepared visualization of the thrust provided by each of the propellers over a range RPM
conceivable for the engine to operate at. This data is comparable to the estimates made and allows for
assurance that the aircraft will be able to obtain appropriate groundspeed to generate sufficient airflow over
the wings for takeoff with the current wing design.
16.4.1.3 VARIABLES MEASURED
The independent variables measured are the forward pulling force generated by the apparatus (propellers
coupled to the engine), and the rotational speed of the propeller, in revolutions per minute.
- 59 -
16.4.1.4 NATURE OF EXPERIMENT
In addressing the nature of the experiment, the following can be said:
The group designed for the engine to be mounted to and securely fixed to a level platform. This setup
utilized an existing engine mount for the engine. The mounting platform was designed to translate freely
and with minimal frictional effects in a forward manner, and was restricted in reverse or rotational motion.
Coupled to the rear end of this platform was a beam scale measuring the magnitude of the pulling force.
The platform allowed for several additional conditions. It was necessary for there to be enough space for
auxiliary equipment needed for engine operation to be mounted. This included the fuel tank and tubing.
The platform was designed remain mobile in one direction, but provide enough clearance from the surface
on which it rests for the propeller to freely rotate. The largest propeller is thirteen inches in diameter. To
measure the rotational speed of the propeller, a tachometer arrangement was designed. This was done with
preexisting equipment and knowledge. The propeller surface was subjected to a concentrated light source
on one side. On the other side of the propeller was a photosensitive component which picked up the light
sourced intermittently as the propeller interrupted it at some defined rate, which was measured by software
rigged to the photosensor. A high test fishing line was used to link the sliding platform to the anchor point.
The experiment required several people to operate successfully. One person was needed to control the light
source, one person was be required to operate the throttle of the engine, and one person was needed to
monitor the recording of the data.
The details of the assembly and the experimental process are described in more detail below.
16.4.2 Experimental Objectives
The objectives for the thrust test experiment of the Magnum engine and propeller combinations were
broken into two categories as follows:
16.4.2.1 MEASUREMENT OBJECTIVES:
These are the independent variables for which data is desired.
- Record thrust of the engine in lbf
- Record the revolutions per minute, RPM of the engine, or some other parameter leading directly
to it through computation, at each associated measurement of thrust
- Record the average amount of fuel consumed throughout the experimental data collection
- Sample the data over the course of the experiment with a data acquisition system
- Record the total amount of time of engine operation
16.4.2.2 2. RESULTS OBJECTIVES:
The following results are desired for each of the four propellers currently in the group’s possession.
- Determine the maximum static thrust capability
- Produce data curves for visualization of pulling force performance with RPM
- Determine the average rate of fuel consumption of the engine
The above measurement and results have been sufficient to allow the group to compare the propellers
directly, as well as predict takeoff capabilities of the aircraft by utilizing the data. Overall, meeting the
objectives has produced a more accurate idea of the aircraft performance.
- 60 -
16.4.3 Apparatus, Equipment, and Instrumentation
In order to fulfill the requirements designated by the objectives portion of the proposal documentation, the
experimental setup and apparatus touched on above were developed. The arrangement, as it was utilized, is
described as follows:
16.4.3.1 GENERAL DESCRIPTION
The whole system was arranged and contained on a relatively large, level surface Namely, a laboratory
workbench desktop outside Crosby Hall. Clamped securely to one end of the lab bench was be an
aluminum beam mounted with strain gauges and wired to accept leads to an ADC. This arrangement was
very similar to that of a laboratory experiment performed in one of the Mechanical Laboratory classes, and
as such was advantageous to the group members. This aluminum beam was mounted longitudinally parallel
to the desk top, such that the direction weakest to bending was subjected to a force from the engine thrust.
The beam was calibrated before the test began.
The force on the beam resulted from the aircraft engine, linked to the beam by the fishing line. Some of the
aluminum beams used in the laboratory have small holes drilled near the ends, through which a cable could
be passed and tied to a pin on the other side, preventing it from passing back through, and modeling the
contact as a point load. This appeared to be the most convenient method for fastening to the beam, and was
indeed employed. On the other end was a stand made from square steel tubing cut to accept the linkage in a
similar manner. This was anchored to the bench using C-clamps. Also fixed to this platform was the engine
and propeller combination. This setup was suited so the propeller tip had necessary clearance over the
surface of the bench. The platform to which the engine attached was only able move perpendicular to the
aluminum beam, such that no component of the thrust acted at an angle to the measurement device, thereby
diminishing the accuracy of the reading. This was assured by utilizing a pair of conventional drawer slides
mounted to the platform, and contained on either side by a section of 2x4 holding the corresponding tracks
for the slides.
At the other end, the aluminum beam arrangement was be wired to an ADC connected to the DAQ device
for automated storage of data.
Data to be taken in addition to the thrust was the RPM of the propeller as well as the fuel consumption of
the engine. The fuel tank was placed on the same platform as the engine due to length restrictions on the
fuel lines. The RPM was measured by a simple voltage divider circuit with one resistor as a reference
resistor and the other as a light-sensitive photoresistor. A light source was set on the opposite side of the
propeller from the photoresistor circuit so that the blades interrupted the light every time they passed
between them. The output from the voltage divider was a time-varying, periodic, analog signal sent to the
ADC. The frequency of the signal was computed by a LabView VI from which the RPM was calculated
based on the number of blades on the propeller.
Engine fuel consumption was be measured by weighing the full fuel tank before the run and the partially
depleted tank after the run. This determined the weight of fuel consumed. The specific weight of the fuel
was measured by weighing a known volume of fuel and dividing that weight by the volume. Then, the
volume of fuel consumed in the test was calculated based on the difference in fuel tank weight before and
after the experiment and the specific weight of the fuel.
- 61 -
16.4.3.2 INSTRUMENTATION
The following instrumentation was utilized in the experiment:
- Aluminum beam scale with mounted strain gages
- Mass Scale
- Volumetric measurement container
- Power supply
- Computer with LabView software
- NI cDAQ-9174 Data Acquisition Chassis
- NI cDAQ-9174 AC Power Adapter
- NI 9219 Universal Analog Input Module
- CdS Photoresistor
- 1 kΩ resistor
- Light Source
16.4.3.3 SCHEMATIC OF EXPERIMENTAL SETUP
Figure 57 - Elevation View of Experimental Setup
(Cont’d)
- 62 -
Figure 58 - Plan View of Experimental Setup
16.4.4 Experimental Theory
The idea behind this experiment was essentially to mirror a simple cantilever beam experiment, the type of
which has been included in the Mechanical Laboratory curriculum here at the University. The goal was to
use the cantilever beam as a scale to measure unknown thrust forces produced by the engine. By first
applying known loads to the beam and measuring the response, unknown loads may later be determined by
examining the behavior of the beam when loaded.
The theory and derivation of using a rectangular beam’s strain as a scale is as follows:
It can be shown through basic strengths of materials type analysis that by calibration, Equation 6 can be
used to relate the loads P and T on a rectangular beam to the strain at some location.
εA = k1P
Equation 6
Where P represents the load force. Note that for the experimental setup proposed, the load path from the
engine to the beam is such that at the attachment to the beam, where the load is applied, pure bending is
approximated, and negligible torsion of the aluminum will occur. No element of the structure undergoes a
torsional load and thus torsional effects are ignored.
Next consider Hooke’s Law:
Equation 7
Or a form that is more useful for Equation 6:
Equation 8
- 63 -
Where:
= Stress(lb/in2)
E = Young’s Modulus for the material (lb/in2).
= Strain (dimensionless)
And for bending stress on beam:
Equation 9
Where:
M = Bending Moment (lb-in)
z* = distance from surface to neutral axis(in)
I = Moment of inertia(in4)
For a beam with rectangular cross-section:
I = (bh3)/12
Equation 10
Where:
b = width of cross-section(in)
h = height of cross-section, its thickness(in)
*Note: For a rectangular, uniform beam:
z = h/2
The bending moment, M, as mentioned earlier is determined as follows:
Equation 11
Where:
P is the applied load (lb)
L is the moment arm (in)
Substituting this equation for the bending moment into Equation 9 Equation 9 the
result is:
Equation 12
Or in a different form:
Equation 13
- 64 -
Substituting Equation 13 into Equation 8 yields the strain at a given point along the beam:
ɛ = Equation 14
Thus:
ɛ = k1P
Equation 15
But also from Equation 14:
ɛ =
So we have:
k1P
And thus the value of k1 is found:
k1 =
Before the experiment, as noted in the introduction, the beam was calibrated. The process in which the
beam is calibrated will be explained later on in greater detail in the procedure section. However, it is
important to note the significance of the calibration process. By placing a variety of known loads on a point
of the beam, a series of strains is induced. Each different load will have a matching strain that can be read
and recorded for later use.
Repeating this loading process yields a series of data that can be adapted into a graphical format. This
graph, theoretically, will be a straight line of constant slope, as described above in Eq. 7. Using the
equation of the line that can be generated easily by programs such as Excel, one will see that there will be
some error in the data collected from calibration, a “y-intercept”.
The equation of the line or calibration curve is of the general form:
ɛA = k1P + b
Equation 16
where ɛA is the induced strain from the applied load, P, k1 is the slope of the line and b is the y-axis
intercept(error).
During the thrust testing, a series of strain values was collected from unknown thrusts. And as the equation
of the line is known from calibration, on top of those strain, ɛ values collected from the experiment, the
unknown thrust values, P, for each strain reading have been determined.
16.4.5 Experimental Procedure
- 65 -
16.4.5.1 EXPERIMENTAL METHOD
The first step in the experimental procedure was to inspect the setup and ensure that all components were
secure. The aluminum beam was calibrated to be accurate and useful for experimentation per the above
discussion. To do this, it was clamped in the same position where it was used while the engine arrangement
was linked to it. A reading of strain with zero load was recorded while in this position. Following this, the
beam was turned so that the short cross sectional dimension was perpendicular to the load direction. A
mass hanger was placed near the end of the beam (the same point at which it was pulled later). Several data
points will be taken by placing increasing amounts of mass on the hangar and recording the stain indicated
by LabView. These data points were used to find the calibration slope as described in the previous section.
The zero load strain taken first was be used as the curve intercept. It was assumed that the slope of the
calibration curve will independent of load orientation During the experiment, the beam was again situated
as it was for the zero-load reading of strain.
The fuel was be measured out, pumped into the fuel tank, and then weighed. The tank was then secured to
the cart and fed to the engine.
The light source was turned on and directed at the photoresistor, with the propeller arc between the two.
The computer was started and LabView opened. Any conversions or computations which the program
needed to do while recording data were set up in the interface beforehand. Data acquisition was set up to
record information for the propeller attached to the engine at that time. In all, three different propellers
were tested. The ability to measure the frequency and strain was checked before proceeding.
The National Instruments data acquisition chassis (NI cDAQ-9174) was located and set near the in-house
built setup. The two analog signal input modules, the NI 9205 and NI 9129, are placed into slots 1 and 2 on
the cDAQ chassis, respectively. The power adapter needed to power the chassis was then connected
between a standard 120 VAC outlet and the appropriate power receptacle located on the front face of the
cDAQ chassis. After connecting the power adapter, the green "POWER" and orange "ACTIVE" lights
appear, indicating the cDAQ chassis has power and is operational. The USB cable was then connected
between the workstation computer and the appropriate USB port, also located on the front face of the
chassis.
The next step was to connect the tachometer cables to the NI 9205 Analog Module. Using alligator clips,
connections were made between the output and ground pins on the tachometer circuit board to the
appropriate terminals on the NI9205. Using the two banana plugs with one end stripped to the wire, output
to Pin 1 on the NI 9205 Module was connected. Another connection was made in the same manner for the
grounded signal to Pin 19 on the NI 9205 Module. These connections ensure the measurements are being
recorded on the differential channel configuration; the output signal voltage is referenced to negative signal
voltage, which then mathematically relates to a differential analog signal.
All that's left was to connect the strain gage setup and the Elenco Power Supply to the NI 9219 Analog
Module. The male end of the banana plug, with the other end stripped to the wire, was connected to the
Variable 2-20V output on the Elenco Power Supply. The connection from the red variable output was
connected to the stripped end to Pin 2 on the NI 9219 Module. The connection from the black "Com"
terminal of the variable output was connected this to Pin 20 on the NI 9219 Module. Taking the stripped
wires that were connected to the appropriate strain gage slots on the end of the strain bridge, connections
were made for the red positive wire to Pin 4, the black negative wire to Pin 5, and the white output wire to
pin 6 on Channel 0 on the NI 9219 Module.
Once all connections were made, the thrust test setup was nearly complete; the next step was to create the
virtual interface which will run the programmable experiment and record the desirable responses. This was
achieved by utilizing the data acquisition software LabView previously installed on the workstation
computer.
- 66 -
Once LabView has been started, a new Virtual Interface (VI) can be created to begin programming and
acquiring data for any given experiment. Using the Functions Palette, the DAQ Assistant function can be
used to setup the input signals from the two sensors which is read and processed by LabView once the
program is running. This function allows for unit configuration, signal input ranges, terminal configuration,
max and min voltage settings, number of samples to read, the rate at which data is recorded, and many
other parameters.
Once the DAQ Assistant function block is setup, the frequency detected by the tachometer and the thrust
force acquired through the strain calibration can be graphed as functions of time that are approximately
real-time readings. However, the output signal reading from the tachometer needs to be converted from Hz
to RPM by a factor of 60 and a factor of 1665 is applied to accurately convert between the strain
experienced and the force generated by the engine.
Once the Waveform Graph function shows accurate readings for frequency and force responses, the last
step before running the experiment was to use the Write To Measurement File function from the Functions
Palette. This allows for recorded data to be saved to a file or drive for easy access to analyze data. A
completed version of the VI Bock Diagram for recording the frequency and force transients is shown below
in Figure 59.
Figure 59 - LabView VI Schematic
From the figure above, the DAQ Assistant is setup to read the strain and appropriately convert it to thrust
force through proper calibration techniques. The DAQ Assistant2 is setup to read from the analog
tachometer reading and channel it through 14 different filters to eliminate 60 Hz noise, set high and low
frequency filters, and eliminate any offset disturbances. The resulting frequency is then extracted and
converted to RPM in order to determine which prop can be used to generate the most thrust throughout the
throttle's range of operation.
Remote control of the throttle was ensured, and the needle valve of the engine was adjusted for startup. The
glow plug igniter was taken from its charger, and here used to warm the glow plug
- 67 -
The engine was then primed and started, given that all group members were in safe positions. Good data
being recorded, the engine was run at speed for some recorded time. The engine was then stopped, and the
fuel tank weighed again.
The next step was to start the engine again. The throttle was slowly opened to increase the rotational speed
of the engine and propeller, and thus the thrust. This continued until the engine reached the highest
obtainable RPM, at which point the throttle was closed, all while recording data.
The propeller was then be interchanged with the next, and the previous steps were repeated for each as
necessary.
Repetition of the tests was not necessary as for each throttle position, multiple sets of data are recorded for
that one position. As the throttle is fixed, the DAQ system records multiple different readings in small time
intervals. The repetition is done within the experiment itself, so there was no need to run through the entire
throttle range multiple times.
16.4.6 Experimental Results and Conclusions
16.4.6.1 PREDICTION OF RESULTS
In the general interest of the group, and due to the fact that basic predictions were able to be made with
relative ease, the theoretical static thrust of one of the available propellers was calculated prior to
completion of the experiment. The goal of this calculation was to obtain a quantitative feel for the validity
of the results garnered in the laboratory experiment afterward. In the absence of another location in which
to discuss these predictions, a brief description of how they were made, and the theoretical results follows,
before the laboratory results.
An excellent source of aerodynamic data for low Reynolds number airfoils is the UIUC Data site,
maintained by University of Illinois professor Michael Selig. It was found that this data site also contains
performance data for select common “small” scale aircraft propellers. One of these is in fact, in possession
of the group. The propellers are listed by brand, diameter, and pitch.
The data given on the site graphically displays the propeller thrust coefficient versus RPM. Due to the
limited number of thrust coefficient-RPM data pairs, these were plotted and fitted with a trendline.The
specific propeller is the APC brand 11x7.
(Cont’d)
- 68 -
APC 11x7 Static Thrust Coefficient vs. RPM
y = 8E-11x2 + 4E-06x + 0.1361
0.14
0.145
0.15
0.155
0.16
0.165
0 1000 2000 3000 4000 5000 6000
RPM
C_to
Ct0 vs. RPM Poly. (Ct0 vs. RPM)
RPM C_to
1800 0.144
2100 0.145
2400 0.1465
2700 0.148
3000 0.1495
3300 0.151
3600 0.1525
3900 0.1535
4200 0.1545
4500 0.1565
4800 0.158
5100 0.16
5400 0.161
5700 0.1625 Figure 60 - Data Table and Plot of Static Thrust Coefficient v. RPM for APC 11x7
As can be seen from the trendline equation on the plot, the trendline fitted to the data is:
y = (8*10^-11)x^2 + (4*10^-6)x + 0.1361
Equation 17
This was used to acquire data points for a wider range of RPM and with increased resolution, due to the
fact that predictions were desired at values of RPM between those given with the thrust coefficients. With
virtually any thrust coefficient in hand, all was ready to utilize the thrust equation:
T = ((0.5)*(rho)*((omega)^2)*(pi)*((D)^4)*(C_t))
Equation 18
where:
T = thrust
Rho = air density
Omega = propeller rotational rate, RPM
D = propeller diameter, in inches
C_t = thrust coefficient
Calculations of thrust were made with this formula over a range of 0 to 10,000 RPM, with standard sea
level density. The results are plotted below:
- 69 -
Theoretical Propeller STATIC Thrust vs. RPM (APC 11x7)
0
2
4
6
8
10
12
14
16
18
0 2000 4000 6000 8000 10000 12000
Angular Rate (RPM)
Th
rust
(lb
f)
Propeller Thrust vs. RPM
Figure 61 - Theoretical Propeller Thrust vs. RPM (APC 11x7)
As can be seen above, the thrust seems to vary parabolically with RPM as might be expected from the
(omega)^2 term in the thrust equation.
It should be noted that the APC 11x7 propeller was not actually tested in the experiment due to the fact that
is was found to be designed for electric motors, and also had several chips. However, another 11x7 was
tested.
16.4.6.2 EXPERIMENTAL RESULTS
The experiment was completed over the course of a single day. Three propellers were tested. These are as
follows:
1. Evo Brand 11x7
2. Evo Brand 12x6
3. K Series Master Airscrew
First, the calibration of the beam scale was completed. The results of the calibration are included in the
appendix of the report.
A description of the LabView VI and the experimental setup for data acquisition was not fully included in
the preliminary sections written before this point due to the fact that it could they could not be known with
certainty at that point, beyond the hardware needed. These will be included in the “Apparatus, Theory and
Procedure” Section of the final report, rather than here in the results, where such information would be out
of place.
The VI was used to acquire data in real time at a high sampling rate, and so a large amount of data points
were generated. For this reason, all of the data obtained in the thrust testing will not be included in the
report. The data was written to a spreadsheet. The data is presented graphically below.
- 70 -
1. Evo Brand 11x7:
E 11x7 Static Thrust vs. RPM
y = 7E-08x2 - 0.0009x + 3.8333
0
1
2
3
4
5
6
7
6500 13000
RPM
Thru
st (
lbf)
thrsut vs rpm Poly. (thrsut vs rpm)
Figure 62 - EVO 11x7 Static Thrust vs. RPM
2. Evo Brand 12x6
E 12x6 Static Thrust vs. RPM
y = 5E-08x2 - 0.0002x + 1.0416
0
1
2
3
4
5
6
5500 6500 7500 8500 9500 10500 11500 12500
RPM
Th
rust
(lb
f)
Thrust vs. RPM Poly. (Thrust vs. RPM)
Figure 63 - EVO 12x6 Static Thrust vs. RPM
- 71 -
3. K-Series Master Airscrew
K Series Master Aiscrew Thrust vs. RPM
y = 1E-07x2 - 0.002x + 8.4639
0
1
2
3
4
5
6
6500 7500 8500 9500 10500 11500 12500
RPM
Th
rust
(lb
f)
Thrust vs. RPM Poly. (Thrust vs. RPM)
Figure 64 - K Series Master Airscrew Static Thrust vs. RPM
Initially examining the results of the testing for each propeller, one can make a number of observations.
First, the pattern of each plot is relatively successful in matching the parabolic behavior predicted by the
initial calculations based on thrust coefficient data. This lends a measure of validity to the thrust
experimentation.
Second, there is some significant scatter in the data. This in reality was to be expected, as in the dynamic
system of even a restrained propeller / engine combination, some motion and play in the system is evident.
One cannot expect a clean pattern of curves to result. This is one reason the group aimed to use data
acquisition to sample a large number of thrust/RPM pairs over time. Trend lines (second order quadratic)
have been fitted to each plot to model the behavior of each propeller. These trend lines are used later for a
direct comparison of the propellers. The curve fit equations are presented on the plots above.
Third, one will note that for any given value of RPM, the values of thrust obtained in the experiment are
much lower than those predicted theoretically. It should be noted that a direct comparison of the theoretical
results and the experimental results is not made because the propeller used for theoretical predictions was
unable to be tested. Reliably sourced data is as yet unknown for making prediction of thrust with the
propellers which were used in experimentation. However, qualitatively, one can see that in the upper range
of RPM, experimental results only yield approximately 25-30% of the predicted magnitude. There are
many potential reasons for this. The rolling track system utilized in the experiment is not by any means
friction less, although good effort was made in minimizing the resistance to motion. The test was
performed outdoor, and so the propeller was subject to crosswind conditions etc. as it would be in actual
flight. The non-rigid linkage between the beam scale and the rolling cart allowed for small backward
motion in some instances and the line would perhaps have a tendency to pull taught and then slacken
repeatedly, which may account for some scatter in the data.
The beam scale was properly calibrated before the test began. However, during calibration it was noted that
once loaded and unloaded, the beam would respond with different output if loaded again. During
calibration there was significant scatter in the output measurements which made fitting a curve for
calibration relatively difficult. This most like contributed to some inaccuracy when comparing the
experimental and theoretical results. Again, the beam calibration data can be seen in the appendix rather
- 72 -
than here in the results section. However, results remain in a reasonable range for competition standard and
are expected to be completely usable values of thrust for the aircraft configuration.
One of the goals of the experiment was to determine the maximum value of thrust afforded by the engine /
propeller system. Listed below are the maximum values of thrust given by each of the tested propellers, and
the corresponding RPM value.
Propeller Maximum Thrust Obtained @ RPM
Evo 11x7 5.87916 lbf 13338.2
Evo 12x6 5.1496 lbf 11050.4
K-Series Master Airscrew 5.1824 lbf 11098.1 Table 2 - Comparative Results of Max Thrust vs. RPM
Based on the above table, the Evo 11x7 clearly gave the highest value of thrust recorded at any point in the
experimentation, and achieved this thrust at a higher value of RPM than the other two in the testing.
A second goal of the experiment was to determine the most adequate propeller for use in flight. Although
the data in Table 2 is taken into consideration when determining this, the highest value of thrust seen is not
the only parameter looked at. The three trend lines generated above are plotted below for a better
simultaneous comparison of the propellers over a typical RPM range.
Thrust vs. RPM Propeller Comparison
0
1
2
3
4
5
6
7
8
6500 7500 8500 9500 10500 11500 12500 13500
RPM
Th
rust
(lb
f)
E 11x7 E 12x6 K Series
Figure 65 - Propeller Trendline Comparison
Cleary, while the Evo 11x7 generated the highest overall single value of thrust, if a curve is fit to its scatter
data points, an overall lower trend than the other two propellers emerges. The other two propellers contend
for highest overall trended thrust value.
In the application of the project, it is intended to optimize maximum pulling force, due to the fact that in
takeoff this is desirable. For very heavy scale aircraft, such as the one to which the engine will be fixed, it
is likely that a large percentage of flight time will be spent with the throttle a good deal of the way open
and the prop blades turning in the higher range of RPM. Thus, it is a good idea to prioritize performance in
the higher half of the x-axis of Figure 6. Throughout essentially all of the tested and useful range of RPM,
the Evo 12x6 propeller produces the greatest pulling force. Although the K-Series Master Airscrew begins
to overtake as the rate continues to grow, the choice of the group for actual implementation is the Evo 12x6
- 73 -
The level of uncertainty in the thrust measurements was found using the data taken throughout the
calibration of the beam scale used. A linear curve was fit to the plot of load vs. voltage ratio to model the
behavior of the beam for later reference. Values from this calibration curve equation were used in
conjunction with the methods presented in the uncertainty level portion of this document, above. The
uncertainty for any given measurement of thrust was found to be +/- 0.527 lbf. This is very close to the
uncertainty level desired by the group during initial formulation of the experiment. A plot of the calibration
curve created for the experiment can be seen in the appendices of the report.
A third goal of the experimentation was to measure the fuel consumption of the engine. To do this, the
experimental procedure above was followed. The fuel consumption was measured after preliminary
examination of the other experimental data.
Initially the tank was filled with 50 mL of nitromethane fuel. The engine was fitted with the Evo 12x6
propeller the group plans to use during flight, and was left running at half throttle for a timed 5 minutes.
The amount of fuel left in the tank was then measured again and found to be 13.5 mL. Thus the average
rate of fuel consumption was calculated as 7.3 mL/min.
The tank size implemented in the fuselage, when completely full, will hold 6 fluid ounces, or 177.4 mL of
fuel. Calculating the total time which can be spent in the air with average rate of fuel consumption, we have
a total of 24 minutes of flying time. This is more than sufficient for the goals of the competition, and would
certainly allow for more aggressively throttled flight for some duration as well.
16.4.7 Mech. Lab Appendix A: Uncertainty Level Buildup
16.4.7.1 GENERAL COMMENTS
A typical build-up of uncertainty for an experiment of this nature would start at the base level of
measurement, with the uncertainties of the instruments and tools used in the procedure. The resolution or
certainty of measurement achievable with said devices will lead to the magnitude of uncertainty in the data
and parameters desired, and this propagates to the results. Obviously, the results’ uncertainties must be
evaluated as desirable or not, and subsequent adjustments may need to be made.
However, the scenario for the experiment proposed here is somewhat different. While the measurement of
the pulling force exerted by the propeller depends on the measurement of the strain, and the strain
measurements can be arrived at using beam bending theory, the uncertainties of parameters such as the
Young’s Modulus and beam dimensions need not be directly reconciled. This is due to the fact that the
beam will be utilized to generate a calibration curve by applying known loads and measuring the response.
Creating this curve and using it as a basis for the results effectively takes into account the aforementioned
uncertainties. Instead, the uncertainty desired is the uncertainty of using the calibration curve, which results
from the uncertainty of the strain gauge response when subjected to load. When the experiment is done, the
reading taken by the data acquisition system will be the voltage across the resistance of the strain gauges.
16.4.7.2 THRUST MEASUREMENT
The aluminum beam scale to be used as a transducer is to be calibrated so that the relationship between
output voltage and force applied will be known to sufficient precision. The relationship that is expected
between force and voltage ratio (output voltage divided by excitation voltage) is a linear one, assuming the
following conditions are satisfied:
Gages on top and bottom have negligible differences in characteristics.
The misalignment of the gages from the beam axis is negligible.
The force has negligible components in any directions other than intended.
- 74 -
That is, the curve formed from the data collected during calibration should take on the following form:
where:
VOUT is the output voltage (measured across the bottom strain gage)
VEX is the excitation voltage
P is the force applied to the beam and
a and b are constants.
Data is collected during the calibration process with a known load applied and several measurements of
voltage ratio taken. Scattering of data about an average value of voltage ratio is expected since the signal,
in general, will contain noise.
The constants, a and b come from the equation of the linear regression computed from the data. The
equation is then inverted so that the force is a function of voltage ratio like so:
The norm of residuals computed from the data is then taken to represent the uncertainty in the force, Wp.
where:
n is the number of data points collected and
Pi is the force at data point number i.
The value of this uncertainty in force will not be known until data is collected. Very small uncertainties,
while attractive, are not necessary to achieve the goal of comparing different propeller performances unless
their performance curves are very close. Furthermore, if two propellers have very nearly the same thrust
performance over the throttle range, the choice of propeller would not make much of a difference anyway.
In conclusion, the precisions and resolutions of the devices used in the experiment (see Theory and
Apparatus section) will be more than adequate to yield results with uncertainties low enough for the
purpose of propeller selection and estimate of maximum static thrust capability. Due to the supposed
suitability of the current proposed devices, the procedure could remain unaltered.
Fuel Consumption Measurement
The measurement of fuel consumption rate will consist of weighing the fuel tank before and after a run at a
fixed throttle level with the selected propeller mounted. The fuel weight consumed divided by the time
elapsed over the run will be the resulting average fuel consumption rate. Since only weights and times are
measured, the uncertainty in fuel consumption rate is only a function of precision of two devices:
the scale upon which fuel and tank weight are measured and
the stopwatch (along with its operator) used to measure start and stop times for the run
- 75 -
Equation 19
where:
is the average fuel consumption rate
Wf is the final weight of fuel plus tank
Wi is the initial weight of fuel plus tank
tf is the time at the end of the run
ti is the time at the beginning of the run
It might seem strange at first that ti, the time at the start of the run, is not just taken to be exactly zero and
ignored in the uncertainty analysis. Keep in mind that there is error due to the stopwatch and the stopwatch
operator that prevents the start time of the stopwatch from being exactly equal to the start time of the
engine. Indeed, this will add to the total uncertainty in the average fuel consumption rate.
The average fuel consumption rate, as a function of four variables shown above, will have an uncertainty
dependent on the precision of the measurement devices as well as the sensitivities of the consumption rate
to each independent measured variable. This relationship, from a Klein-McClintock analysis, is shown
below.
Equation 20
where:
δW is the uncertainty in measurement of weight (0.1 oz)
δt is the uncertainty in measurement of time (0.5 s)
Which are the magnitudes of the data parameter uncertainties which have been selected.
The partial derivatives represent the sensitivity of average fuel consumption rate to each independent
variable. These partial derivatives may be found by differentiating Equation 19 symbolically and plugging
in nominal values for the independent variables to obtain a number for the sensitivities. This will be
demonstrated here for the sensitivity to Wf only, with an expected run time of 15 minutes and final and
initial weights of 5oz and 1oz, respectively.
After all uncertainties and sensitivities are computed, Equation 20 will yield the overall uncertainty in the
measurement of average fuel consumption rate.
Taking the estimate of average fuel consumption rate to be 4oz per 900 seconds, the relative uncertainty
can be computed.
- 76 -
This estimate of uncertainty of 3.5% is adequate as it is under the 5% upper limit imposed. Thus, one may
conclude that the scale accurate to the tenth-ounce and the time-keeping accurate to the half-second will
yield results of sufficiently low uncertainty for average rate of fuel consumption.
Given the above quality of the results uncertainty, the bounds on the data parameters selected above do not
need to be tightened and the procedure can remain as suggested previously.
- 77 -
16.4.8 Mech. Lab Appendix B: Beam Calibration
When the beam scale used in the experiment was calibrated in the manner described throughout the report
above, the following calibration curve was produced:
Figure 66 - Calibration Data Plot
The equation representing this curve was used in the VI setup to compute values of thrust produced by the
engine in real time. Numerical tabulation of calibration data is avoided in the report as it is seen as
unnecessary to the accurate description of the group’s methods.
- 78 -
16.5 Appendix E - Flight Simulation Code
clear all
clc
format short g
%Take off distance (ft)
Dis = 200;
%Aspect Ratio
AR=9;
mu=3.848*10^-7;
Rho=0.0023769; %Slugs/ft^3
gravity=32.1740;
CL=1.48;
CD=0.01541;
count=0;
Count=0;
for T=4.88:7.88
for weight=35:5:50
for I=9:18
count=count+1;
Chord=I/12;
Wingspan=.85*Chord*AR;
% Wingspan=Chord*AR;
%Surface area of airfoil (ft^2)
% Area = Chord*Wingspan;
Area=Wingspan^2/AR;
Mass=weight/gravity;
%Drag Calculations + Motion Simulation
Drag=.5*Rho*CD*Area;
%Lift Calculations
LiftCo=.5*Rho*CL*Area;
sim('MotionEq2')
rows = length(Pos_x(:,1));
Lift=Lift2(rows,2);
if Lift>weight
Count=Count+1;
K(Count,1)=T;
K(Count,2)=weight;
- 79 -
K(Count,3)=Chord;
K(Count,4)=Wingspan;
K(Count,5)=Lift; %Lift at takeoff distance
K(Count,6)=V_x(rows,2);
end
end
end
end
K
end
- 80 -
16.6 Appendix F - Servo Performance MathCAD Worksheet
- 81 -
- 82 -
16.7 Appendix G - Engine Specs
- 83 -
- 84 -
- 85 -
16.8 Appendix H – Project Timeline
- 86 -
16.9 Appendix I – Team Photograph
[L to R: Travis Cushman, Ben Waller, Zach Veilleux, David Chandpen, Joe Travaglini, Matt Maberry]
