Analysis and Verification of Cost-Effective Design ...

Iowa State University

From the SelectedWorks of A Ram (Bella) Kim

Winter January 8, 2018

Analysis and Verification of Cost-Effective DesignModifications to Commercially Available Fixed-Wing Unmanned Aerial Vehicle to ImprovePerformance, Stability and ControlCharacteristics, and Structural IntegrityA Ram (Bella) Kim, Iowa State University

Available at: https://works.bepress.com/arambella-kim/6/

http://www.iastate.edu

https://works.bepress.com/arambella-kim/

https://works.bepress.com/arambella-kim/6/

American Institute of Aeronautics and Astronautics

1

Analysis and Verification of Cost-Effective Design

Modifications to Commercially Available Fixed-Wing

Unmanned Aerial Vehicle to Improve Performance, Stability

and Control Characteristics, and Structural Integrity

Aaron Blevins1, Hady Benyamen 2, Grant Godfrey3, Daksh Shukla4, Bella Kim5,

University of Kansas, Aerospace Engineering, Lawrence, Kansas, 66045

In this work, several cost-effective design modifications were made to a commercially

available fixed-wing unmanned aerial vehicle in order to improve overall performance and

reliability. These modifications included the addition of 3D-printed winglets to improve

aerodynamic performance, applying fiberglass reinforcement to areas of structural concern,

the addition of rudder control surfaces for improved control authority, increasing the length

of the empennage booms for both stability and control improvements, and the addition of

dorsal fins to improve vertical fin stall characteristics. Design methodology and

manufacturing processes were outlined, a cost breakdown was presented, and results from

dynamic analysis were validated through extensive flight testing of the modified aircraft.

Nomenclature

ζph = Damping coefficient for phugoid dynamic mode

ζsp = Damping coefficient for short-period dynamic mode

ζspiral = Damping coefficient for spiral dynamic mode

α_trim = Steady State Angle of Attack (degrees)

b = Wingspan (ft)

CDi = Induced Drag Coefficient

Cl_p = Roll Moment Coefficient due to Roll Rate

Cmδe = Normalized Coefficient for Pitch Moment generated by Elevator Deflection

Cn_r = Yaw Moment Coefficient due to Yaw Rate

L/D = Lift-to-Drag Ratio

S = Wing Planform Surface Area (ft2)

Xac_Aircraft = Location of aircraft aerodynamic center aft of nose

I. Introduction

As unmanned aerial vehicles (UAV’s) are becoming more common in commercial, military, and research

applications, the demand is increasing for improved system reliability and flight characteristics. Additionally, as

avionic components are being developed at smaller sizes, it is now possible to conduct flight missions using a small

UAV that previously would have required a much larger-sized UAV. This is incredibly significant due to the reality

that the cost of UAVs scale exponentially with size. From a research point of view, unmanned aerial platforms are

used to mobilize an airborne sensor, and project funding has a much higher chance of being secured if the underlying

costs of the vehicle are decreased.

The SkyHunter is a low-cost commercially available UAV that is designed to carry aerial video equipment. At an

empty weight of 5.5 lbs, it is designed for a maximum takeoff weight of 10 lbs. This available payload weight ratio

makes it an ideal candidate to be used as a research platform. However, research instrumentation and hardware is

1 Ph.D. Student, Department of Aerospace Engineering, 2120 Learned Hall 1539 W. 15th st., Student member 2 Master’s Student, Department of Aerospace Engineering, 2120 Learned Hall 1539 W. 15th st., Student member 3 Undergraduate, Department of Aerospace Engineering, 2120 Learned Hall 1539 W. 15th st., Student member 4 Ph.D. Student, Department of Aerospace Engineering, 2120 Learned Hall 1539 W. 15th st., Student member 5 Ph.D. Student, Department of Aerospace Engineering, 2120 Learned Hall 1539 W. 15th st., Student member

Dow

nloa

ded

by I

OW

A S

TA

TE

UN

IVE

RSI

TY

on

Aug

ust 1

9, 2

018

| http

://ar

c.ai

aa.o

rg |

DO

I: 1

0.25

14/6

.201

8-22

61

2018 AIAA Information Systems-AIAA Infotech @ Aerospace

8–12 January 2018, Kissimmee, Florida

10.2514/6.2018-2261

Copyright © 2018 by the American Institute of Aeronautics and Astronautics, Inc. All rights reserved.

AIAA SciTech Forum

http://crossmark.crossref.org/dialog/?doi=10.2514%2F6.2018-2261&domain=pdf&date_stamp=2018-01-07


2

incredibly expensive, and as such, should not be exposed to unnecessary risks. For this purpose, cost-effective design

modifications to the aircraft are proposed to improve overall flight performance and reliability.

For research purposes, the SkyHunter platform used in this research is equipped with a commercially-off-the-shelf

(COTS) autopilot hardware, Pixhawk, which is utilized only as the data acquisition or the sensor interface board,

connecting servos and using its onboard IMU sensor, with externally interfaced U-Blox GPS and a pressure sensor.

This board is interfaced with a computational platform, Odroid XU4 that runs the Guidance, Navigation and Control

(GNC) software and communicates to Pixhawk for retrieving sensor values and sending servo commands, via

MAVROS software.

II. Winglets

Winglets were designed in order to improve the aerodynamic performance of the aircraft (Ref. [1], [2]). The

winglets reduce the strength of the wingtip vortices, and therefore cause a reduction in the induced drag of the aircraft

(Ref. [3], [4], [5]). Based upon work done using CFD analysis in Ref. [6], winglet designs were chosen and sized

accordingly. The two winglet configurations chosen for this project were a blended winglet and a split-scimitar

winglet. The winglets were fabricated using a 3D printer, and attached to the aircraft. The winglets were attached by

simply slipping them on the end of the wings, as they were designed to closely match the GOE-438 airfoil. The effects

of winglets on small unmanned aerial vehicles was investigated in Ref. [7], [8], [9]. In this research, the effects of

these winglet designs on the SkyHunter aircraft were investigated via two separate dynamic modeling softwares, in

which both results indicate that the addition of the winglets had a positive effect on the performance, stability, and

handling qualities of the aircraft. Both modeling software used a baseline case of no winglets to compare with the

blended winglets and with the split scimitar winglet design. The aircraft geometric reference parameters typically

affected by the winglet choices were taken into account and are shown below in Table 1.

Table 1: Winglet Effect on Aircraft Reference Geometry Parameters

Parameter No Winglets Blended Winglet Split Scimitar

Wingspan (ft) 5.9 6.9 6.5

Total Wing Surface Area (ft2) 4.41 4.82 4.70

Total Aircraft Weight, lbs 7.81 8.27 8.43

A. AVL Results for each Winglet type

Athena Vortex Lattice (AVL) is a software developed by MIT that applies extended vortex lattice method to lifting

surfaces to determine forces and moments via a full linearization of the aerodynamic model about the flight condition.

Because non-lifting surfaces have a negligible effect on this method, fuselage and other components are typically

ignored. AVL is very useful to the authors due to the ease of setting flow conditions for a numerical analysis on the

aircraft geometry, without all of the complexity of a full CFD analysis. The AVL geometry for the aircraft with various

winglet conditions are shown below in Figure 1, Figure 2, and Figure 3.

Figure 1: AVL No-Winglets Model

Dow

nloa

ded

by I

OW

A S

TA

TE

UN

IVE

RSI

TY

on

Aug

ust 1

9, 2

018

| http

://ar

c.ai

aa.o

rg |

DO

I: 1

0.25

14/6

.201

8-22

61


3

Figure 2: AVL Blended-Winglet Model

Figure 3: AVL Split-Scimitar Winglet Model

The results from AVL for the key parameters of this analysis are shown below in Table 2.

Table 2: AVL Results


CD_induced 0.0049 0.0037 0.0035

L/D Ratio 9.38 9.69 10.1

α_trim (deg) 3.06 2.45 2.34

Cl_p -0.5433 -0.9379 -0.9794

Cn_r -0.1262 -0.1261 -0.1258

From these results, it can be seen that the addition of the winglets leads to a reduction in induced drag, an increase

in Lift-to-Drag ratio, a lower trim angle of attack, as well as increased damping in both the lateral and directional axes.

These trends are very positive, but it should be noted that while the addition of winglets obtains these desired effects,

the difference in the winglet choices is not significant, with the split-scimitar outperforming the blended winglet in

this AVL analysis.

B. AAA Result for each Winglet type

Advanced Aircraft Analysis (AAA) is a dynamic modeling software developed by DARCorporation that employs

high fidelity physics based methods combined with semi-empirical methods to determine aircraft flight characteristics.

A vast historical database of aircraft information is drawn from to create the empirical relations used in AAA.

However, the software has been known to be less accurate for small Reynold’s numbers, which are experienced by

small unmanned aerial systems.

Dow

nloa

ded

by I

OW

A S

TA

TE

UN

IVE

RSI

TY

on

Aug

ust 1

9, 2

018

| http

://ar

c.ai

aa.o

rg |

DO

I: 1

0.25

14/6

.201

8-22

61


4

The 3-View of the aircraft geometry with the various winglet conditions are shown below in Figure 4, Figure 5,

and Figure 6.

Figure 4: AAA No-Winglets Model

Figure 5: AAA Blended-Winglets Model

Figure 6: AAA Split-Scimitar Model

Dow

nloa

ded

by I

OW

A S

TA

TE

UN

IVE

RSI

TY

on

Aug

ust 1

9, 2

018

| http

://ar

c.ai

aa.o

rg |

DO

I: 1

0.25

14/6

.201

8-22

61


5

The results from AAA for the key parameters of this analysis are shown below in Table 3.

Table 3: AAA Results


L/D Ratio 11.5 11.87 11.3

α_trim (deg) 3.39 2.61 3.09

Cl_p -0.4980 -0.5565 -0.5293

Cn_r -0.1744 -0.1231 -0.1345

From the AAA analysis, it is seen that the blended winglets significantly outperform the split-scimitar in terms of

longitudinal characteristics, and that the split-scimitar actually has a detrimental effect on the L/D ratio of the aircraft.

However, the blended winglets was found to reduce the directional stability of the aircraft.

C. Winglet Manufacturing

Using the CAD models for the winglets, they were then manufactured using a MakerGear M2 3D printer with

standard PLA filament. A honeycomb infill method was used to optimize structural integrity with respect to the overall

weight of the winglets. Though the printing process took 20+ hours to finish, the final products were satisfactory with

regard to proper shape, smoothness, and overall weight. Figure 7 through Figure 12 document the manufacturing

process for the winglets.

Figure 7: Print Preview in Simplify3D Software for Split-Scimitar Winglet

Dow

nloa

ded

by I

OW

A S

TA

TE

UN

IVE

RSI

TY

on

Aug

ust 1

9, 2

018

| http

://ar

c.ai

aa.o

rg |

DO

I: 1

0.25

14/6

.201

8-22

61


6

Figure 8: 3D Printed Blended Winglet

Figure 9: Aircraft Without Winglets

Figure 10: Aircraft with Blended Winglets

Dow

nloa

ded

by I

OW

A S

TA

TE

UN

IVE

RSI

TY

on

Aug

ust 1

9, 2

018

| http

://ar

c.ai

aa.o

rg |

DO

I: 1

0.25

14/6

.201

8-22

61


7

Figure 11: Aircraft with Split-Scimitar Winglets

Figure 12: Close-Up of Split-Scimitar Winglets

III. Addition of Rudders

Rudder control surfaces were added to the aircraft in order to improve directional control authority. Rudders were

cut from the existing vertical tails of the aircraft and secured using hinges. Rudder servos were mounted into the

vertical tail to actuate the control surfaces, and these servos were mounted into wood blocks to reinforce the structure.

Improvement of aircraft control and stability improvements will be assessed through flight testing. The design

schematic for the rudders is shown below in Figure 13 and Figure 14, and the final rudder assemblies are shown in

Figure 15.

Figure 13: Rudder Creation Schematic

Dow

nloa

ded

by I

OW

A S

TA

TE

UN

IVE

RSI

TY

on

Aug

ust 1

9, 2

018

| http

://ar

c.ai

aa.o

rg |

DO

I: 1

0.25

14/6

.201

8-22

61


8

Figure 14: Top View of Rudder Implementation

Figure 15: Rudder Mounted on Vertical Tail

IV. Increased Length of Empennage Booms

The importance of an aircraft’s empennage is to create forces that generate moments with respect to the aircraft’s

center of gravity. Because of this, the moment arm for the empennage plays an important role in the flying qualities

of the aircraft. For example, smaller empennage forces can be used to create similar moments if the empennage is

distanced further from the center of gravity. The booms that connect the wings to the empennage on the SkyHunter

are made of half-inch diameter carbon tubes, which are commercially available for a variety of lengths. This allows

for easy adjustment of the empennage moment arm. In this work, dynamic analysis was conducted in the AAA

software, and the booms were increased from 33 inches to 36 inches. The 3-View of the AAA model is shown below

in Figure 16.

Dow

nloa

ded

by I

OW

A S

TA

TE

UN

IVE

RSI

TY

on

Aug

ust 1

9, 2

018

| http

://ar

c.ai

aa.o

rg |

DO

I: 1

0.25

14/6

.201

8-22

61


9

Figure 16: AAA Geometric Model with Increased Empennage Boom Length

Results of the dynamic analysis are shown below in Table 4. From the results, the 3” increase in boom length, the

aircraft experiences an increase in pitch control authority from the elevator, as well as an aft-shift of the aerodynamic

center of the aircraft. This shift increases the static margin of the aircraft, which in turn improves longitudinal stability

characteristics. The damping coefficient for both of the longitudinal dynamic modes (short period and phugoid) were

increased, as well as the damping coefficient for the spiral mode.

Table 4: AAA Dynamic Results from increasing Boom Length

Parameter Before Boom Length Increase After Boom Length Increase

Cmδe (per degree) -0.965 -1.027

Xac_Aircraft (inches) 17.46 17.62

ζsp 0.782 0.795

ζph 0.075 0.079

ζspiral 0.296 0.319

V. Structural Reinforcement

To improve the reliability and durability of the aircraft, fiberglass reinforcement was added to areas of structural

concern. These areas include the lower fuselage, the wing-boom junction, and the empennage junctions. Applying

fiberglass layers over the foam to stiffen critical areas of the aircraft can be seen as a cost-effective way to improve

system reliability. The patches add a near-negligible amount of weight to the aircraft, while providing a larger margin

of safety on the design loads and enlarge the flight envelope. Each fiberglass patch was comprised of 3 three layers,

in a 45-0-45 pattern. The total weight of the fiberglass reinforcements and epoxy are shown below in Table 5.

Table 5: Fiberglass Reinforcement Weights

Section Weight (lbs)

Fuselage 0.33

Empennage 0.24

Wings 0.19

Total 0.75

The lower fuselage was reinforced for protection against hard landings, as well as to enable the aircraft to operate

without landing gear on grass runways if needed. The reinforced lower fuselage is shown below in Figure 17.

Dow

nloa

ded

by I

OW

A S

TA

TE

UN

IVE

RSI

TY

on

Aug

ust 1

9, 2

018

| http

://ar

c.ai

aa.o

rg |

DO

I: 1

0.25

14/6

.201

8-22

61


10

Figure 17: Aircraft Reinforced Fuselage

Fiberglass patches were also placed over the wing-boom junctions to reinforce that area, as stress in this area as a

result of empennage moments was increased due to the increased boom length. The fiberglass was applied over the

original foam structure, as shown below in Figure 18.

Figure 18: Reinforced Wing-Boom Junctions

Dow

nloa

ded

by I

OW

A S

TA

TE

UN

IVE

RSI

TY

on

Aug

ust 1

9, 2

018

| http

://ar

c.ai

aa.o

rg |

DO

I: 1

0.25

14/6

.201

8-22

61


11

Fiberglass reinforcements were also applied at the junction of the horizontal and vertical tails on the empennage,

as this area was relatively flimsy in the original design. The fiberglass patches were placed both on the inside and the

outside of the junctions in order to increase stiffness. The additional load of the rudder also necessitates the stiffening

of the vertical tail and its junction with the horizontal tail. The reinforced empennage is shown below in Figure 19.

Figure 19: Reinforced Empennage Junctions

VI. Dorsal Fins

Dorsal fins are thin surfaces placed in front of the vertical tail that act as a vortex generator for the vertical tail. In

Ref. [10], extensive discussion, CFD results, and results of wind tunnel testing was presented on the effect of dorsal

fins as vortex generators for the vertical tail. A method was then proposed for sizing a dorsal fin based upon aircraft

geometry characteristics, in order to expedite the design process. From this analysis, it was shown that a dorsal fin

could prolong the stall on the vertical tail from 16 degrees to 25 degrees of sideslip. Methods proposed from that work

were used to size the dorsal fin, which was manufactured from sheet aluminum, and is shown below in Figure 20. The

addition of the dorsal fins are expected to have a significant effect in high crosswind landings, and can be evaluated

by the pilot during flight testing.

Dow

nloa

ded

by I

OW

A S

TA

TE

UN

IVE

RSI

TY

on

Aug

ust 1

9, 2

018

| http

://ar

c.ai

aa.o

rg |

DO

I: 1

0.25

14/6

.201

8-22

61


12

Figure 20: Dorsal Fin

VII. Cost-Breakdown

A full cost breakdown is presented in this section. The cost breakdown for the aircraft before design modifications

is given below in Table 6. The total price of the SkyHunter aircraft before modifications is $820. This includes the

empty aircraft, servos, engine assembly and propeller, electric speed controller, communication antenna, flight

batteries, and the RC receiver. This cost breakdown does not contain the flight controller unit or the offboard computer

needed to perform autonomous flight.

Table 6: SkyHunter Aircraft Costs

Component Price

Airframe $120

Servos $160

Electric Engine and Propeller $200

Electronic Speed Controller $90

Receivers and Satellites $90

Communication Antenna $60

Batteries $100

Total: $820

The cost breakdown for the design modifications to the aircraft is presented below in Table 7. This breakdown

assumes access to a 3D printer and CAD design software, as well as a clean-room where composites can be properly

laid up. This breakdown does not account for labor hours, though the only significantly tasking process in these

modifications is the CAD design for the winglets and rudders. Though the 3D printing process for each winglet takes

approximately 20 hours, once it is started it does not have to be monitored by a human operator.

Dow

nloa

ded

by I

OW

A S

TA

TE

UN

IVE

RSI

TY

on

Aug

ust 1

9, 2

018

| http

://ar

c.ai

aa.o

rg |

DO

I: 1

0.25

14/6

.201

8-22

61


13

Table 7: Aircraft Modification Costs

Component Price

PLA filament for Winglets $30

Fiberglass and Epoxy $20

Hinges for Rudders $5

Rudder Servos $45

Carbon Booms for Empennage $15

Aluminum Plates for Dorsal Fins $10

Total $125

Table 8 shows the cost comparison of the aircraft before modifications with the aircraft after modifications.

Table 8: Aircraft Before and After Modifications Cost Breakdown

Configuration Price

Aircraft Before Modifications $820

Modifications $125

Aircraft After Modifications $945

VIII. Flight Test Validation

Flight tests of the modified aircraft were conducted at the Clinton International Model Airport in Lawrence,

Kansas, where the KU Flight Systems Lab has conducted over 100 flight tests on research UAV’s in the last several

years. The aircraft was flown using both sets of winglets, as well as the incorporation of rudder control surfaces and

the previously mentioned structural reinforcements. The different winglets are shown on the flying aircraft in Figure

21 and the aircraft using the split-scimitar winglets is shown during takeoff in Figure 22.

Figure 21: Flight Test Winglet Comparison

Dow

nloa

ded

by I

OW

A S

TA

TE

UN

IVE

RSI

TY

on

Aug

ust 1

9, 2

018

| http

://ar

c.ai

aa.o

rg |

DO

I: 1

0.25

14/6

.201

8-22

61


14

Figure 22: Aircraft Takeoff

Using the offboard computer and the Pixhawk unit for data acquisition, the system was able to use custom

guidance, navigation, and control schemes to fly autonomously around predefined waypoints at the flight field. The

aircraft was able to do this with either set of winglets attached. The results for the autonomous flight is shown below

in the following figures.

Figure 23: Autonomous Flight Tracking

Dow

nloa

ded

by I

OW

A S

TA

TE

UN

IVE

RSI

TY

on

Aug

ust 1

9, 2

018

| http

://ar

c.ai

aa.o

rg |

DO

I: 1

0.25

14/6

.201

8-22

61


15

Figure 24: Autonomous Flight Longitudinal States

Figure 25: Autonomous Flight Lateral-Directional States

From this data, it can be seen that the aircraft states and control commands collected during the flight test

compare similarly to those generated in the simulation environment. Differences in signal frequencies may be due to

Dow

nloa

ded

by I

OW

A S

TA

TE

UN

IVE

RSI

TY

on

Aug

ust 1

9, 2

018

| http

://ar

c.ai

aa.o

rg |

DO

I: 1

0.25

14/6

.201

8-22

61


16

noise from unsteady wind disturbances, and the noticeable difference between the throttle command between

simulation and actual flight test may be due to inaccurate engine modeling. 2-D tracking of the aircraft tends to

overshoot the desired waypoint box, which is due to the nature of the guidance algorithms which highly penalize

control input rates. Though the aircraft flew well autonomously and was able to track the desired altitude, airspeed,

and waypoints, much work is needed to improve tracking performance. However, the aircraft displays satisfactory

flight characteristics and has proven to be a reliable platform for the KU Flight Systems Lab to improve its guidance,

navigation, and control algorithms on. The flight verification of the effect of extending the empennage booms and the

addition of dorsal fins is left for future work.

Additionally, the SkyHunter aircraft was chosen for KU’s AE 245 course, which is an introductory to

Aerospace Engineering course offered to freshmen. The class was split into 8 groups, which each built a SkyHunter

aircraft from the kits, applied fiberglass reinforcements as outlined in this project, and modified the vertical tails to

incorporate rudders following the outlined procedure in this project. All 8 groups were able to perform 3 flights

each, where they first performed a velocity calibration for the pressure sensor, then performed an auto-tuning flight

for the PixHawk flight controller, and finally performed autonomous flight around user-defined waypoints. All

aircraft flew incredibly well, with help from the added rudder servos, and all aircraft maintained structural integrity

throughout the flights, which is a testament to the structural reinforcements added to the aircraft. Some of the

aircraft used for this course are shown below in Figure 26.

Figure 26: SkyHunters Flown for AE 245 Course

IX. Conclusion

In this work, several cost-effective modifications were made to a commercially available UAV to improve the

flight characteristics and overall system reliability. Winglets were designed in a CAD software and manufactured

using a 3D printer, and analysis of the dynamic modeling showed satisfactory improvements in induced drag reduction

and aircraft stability. Rudder control surfaces were also designed and manufactured using the existing vertical tail

structure. The booms that connect the empennage to the wings were increased to improve empennage effectiveness,

improving both stability and control characteristics of the aircraft, and this was done through purchasing low-cost

commercially available carbon spars to replace the original booms. Areas of structural concern on the aircraft due to

inherent design and from the modifications in this work were reinforced using fiberglass patches, which added

structural stiffness with negligible associated weight penalties. Dorsal fins were manufactured from sheet aluminum,

and installed on the aircraft to improve the stall characteristics of the vertical tails, which is very advantageous on high

Dow

nloa

ded

by I

OW

A S

TA

TE

UN

IVE

RSI

TY

on

Aug

ust 1

9, 2

018

| http

://ar

c.ai

aa.o

rg |

DO

I: 1

0.25

14/6

.201

8-22

61


17

crosswind landings. A full cost breakdown of the aircraft and the modifications was conducted, and flight tests were

conducted to validated improvements in flight characteristics. Autonomous flight was conducted on the platform, and

flight test data was used for validating of the aircraft dynamic model. Extensive testing of the airframe with rudders

and structural modifications was conducted via flight testing 8 separate aircraft built by a freshmen course at the

University of Kansas.

Acknowledgments

The authors would like to acknowledge our remarkable advisor Dr. Shawn Keshmiri, as well as the support of the

Aerospace Engineering Department at the University of Kansas, and our outstanding RC pilot, Matt Tener. Funding

for aircraft, avionics components, and flight testing were provided by NASA Learn project #NNX15AN94A, NASA

CAN project #NNX15AN04A, and the Paul G. Allen Family Foundation.

References

1R. Eppler, “Induced drag and winglets”, Aerospace Science and Technology, 1997. 2P. Gerontakos and T. Lee, “Effects of Winglet Dihedral on a Tip Vortex”, Journal of Aircraft, 2006. 3Ali Elhama, Michel J.L.van Tooren, “Winglet multi-objective shape optimization”, Aerospace Science and Technology, 2014. 4George Russell Hicks, “The Design and Optimization of a Winglet Using a Vortex Lattice Method and Improved Drag

Calculations”, Thesis, Mississippi State University, 1997. 5Antoine Moreau, “Implementation and Validation of a Method for Computing the Induced Drag of Multiple Lifting Surfaces

Aircrafts”, Thesis, École De Technologie Supérieure Université Du Québec, 2011. 6Narayan, G., and John, M., “Effect of winglets induced tip vortex structure on the performance of subsonic wings”, Aerospace

Sciences and Technology, 2016. 7Blevins, A., Fritz, L., Weaver, J., and Williams, C., “Effects of Winglets on Small Unmanned Aerial Systems”, AIAA

Modeling and Simulations Technologies Conference, Kissimmee, Florida, 2015. 8P.Panagiotou, P.Kaparos, K.Yakinthos, “Winglet design and optimization for a MALE UAV using CFD”, Aerospace Science

and Technology, 2014. 9Jacob Weierman and Jamey Jacob, “Winglet Design and Optimization for UAVs”, AIAA Applied Aerodynamics Conference,

2010. 10W.A. Anemaat, A. Karwas, W. Liu, S. Johnson, "Dorsal Fin Design Method: A Low Cost Aerodynamic Solution to Prevent

Loss-of-Control", AIAA-2017-0694, presented at the 55th AIAA Aerospace Sciences Meeting, AIAA Science and Technology

Forum and Exposition, January 9-13, 2017, Grapevine, Texas.

Dow

nloa

ded

by I

OW

A S

TA

TE

UN

IVE

RSI

TY

on

Aug

ust 1

9, 2

018

| http

://ar

c.ai

aa.o

rg |

DO

I: 1

0.25

14/6

.201

8-22

61

Analysis and Verification of Cost-Effective Design ...

Documents