Prof. Jean-Marc MoschettaProfessor of AerodynamicsDepartment of AerodynamicsEnergetics and PropulsionISAE BP54032, 31055 Toulouse CedexFrancejean-marc.moschetta@supaero.fr.

Hovering Capabilities of Fixed-WingMicro-Aerial Vehicles

Jean-Marc Moschetta1, Institut Supérieur de l’Aéronautique et de l’Espace, Université de Toulouse, Franceand

Sergey V. Shkarayev2, Department of Aerospace and Mechanical Engineering, University of Arizona, Tucson, Arizona

Fixed-wing micro air vehicles (MAV) are very attractive for outdoor surveillance missions since theygenerally offer better payload and endurance capabilities than rotorcraft or flapping-wing vehicles ofequal size. They are generally less challenging to control than rotorcraft in outdoor environment andallow for a dash capability to escape enemy attention. On the other hand, they usually fail miserablyto perform vertical take-off and landing (VTOL) and sustain stable hover flight which proves to becrucial for urban surveillance missions including building intrusion. The present paper investigatesthe possibility to improve the aerodynamic performance of fixed-wing MAV concepts so as to allowfor true hovering capabilities and still maintain high cruise speed for covertness.

Several combinations of rotors and fixed-lifting surfaces were tested, analyzed and compared. First,a tandem-rotor biplane MAV configuration was designed and tested as a result of different biplanepowered configurations. A low-speed autonomous fixed-wing MAV was fabricated and flight testedto perform multi-tasking outdoor surveillance missions. Secondly a side-by-side comparison of a tilt-wing and a tilt-body powered configuration with a pair of counter-rotating motors in tractor configurationwith MAV configurations were carried out. The tilt-body configuration was shown to be more suitablefor MAV applications with higher hovering performances. Second, a coaxial tilt-body concept basedon coaxial motors and contra-rotating propellers inspired from the Convair XFY1 “Pogo” experimentalaircraft was designed and tested. A wind tunnel test was carried out to fully characterize theaerodynamic performances of the coaxial tail-sitter configuration, named Vertigo. An autonomousversion was developed in order to autonomously perform transitions between horizontal and verticalflight. A smaller 300-mm span version, called mini-Vertigo, was designed and fabricated based on aseries of wind tunnel tests using miniaturized coaxial-rotor propulsion set. Autonomous altitude holdand attitude stability augmentation were then achieved using specific control laws adapted for thePaparazzi autopilot system. Thirdly, a new no-through-shaft coaxial-rotor configuration has beenproposed in order to enhance the prototype ruggedness through an embedded spherical structuremade of carbon rods. It is believed that such a crash-proof VTOL MAV, called Cyclope, can be veryattractive for the use of MAV systems in real operations and allows for further size reduction such asfor Nano Air Vehicles applications. Current prospects include both further wind tunnel tests usingnew high-precision micro sting-balances on coaxial-rotor tail-sitter MAVs and the development ofcontrol laws to autonomously perform transition flights.

1 Professor of Aerodynamics, Department of Aerodynamics, Energetics and Propulsion, ISAE BP54032, 31055 Toulouse Cedex, France, jean-

marc.moschetta@supaero.fr.2 Associate Professor, Department of Aerospace and Mechanical Engineering, The University of Arizona, PO Box 210119, Tucson AZ 85721-

00119,vs@u.arizona.edu.

Hovering Capabilities of Fixed-Wing Micro Air Vehicles

Jean-Marc MoschettaInstitut Supérieur de l’Aéronautique et de l’Espace , Toulouse, France

Sergey ShkarayevUniversity of Arizona , Tucson, Arizona

1st US-Asian Demonstration & Assessment of Micro-Aerial & Unmanned Ground Vehicle Technology, 10-15 March, Agra, India

2MAV’08March 10-15, 2008

Motivation

Launch

Loiter

Hover

Building intrusion

Dash

Phase 1Phase 1Phase 1Phase 1Phase 2Phase 2Phase 2Phase 2

Phase 3Phase 3Phase 3Phase 3

Phase 4Phase 4Phase 4Phase 4

Typical MAV mission profile

3MAV’08March 10-15, 2008

MAV design challenges

• Challenging to control

• Fast horizontal flight

Fixed Wing MAVs Rotorcraft MAVs

• VTOL Fixed-Wing MAV– High speed for covertness

– Low speed for clear images

– Limited electric consumption

• Poor image capturing

• Maneuverability in confined environment

MinusKiool (2003)

20 cm, 60 g

Br2C (2007)

36 cm, 490 g

4MAV’08March 10-15, 2008

Camber changes h = 3-9% with synchronous change of reflex hi = 1-3%

Test flights were conducted with Pitot tube sensor

With camber increases from 3% to 9% minimal speed decreases by about 100%

A low-speed in-flight adaptive wing MAV

5MAV’08March 10-15, 2008

A low-speed tandem wing MAV

TYTO250g - 40 cm

• Tandem-wing MAV

• Gimbaled 2-axis camera

Moschetta, Thipyopas, J Aircraft, vol. 44, 2007

6MAV’08March 10-15, 2008

New MAV concepts

Tilt-body

Tilt-rotor

Tilt-wing

7MAV’08March 10-15, 2008

• Wind tunnel

– Test section 45 x 45cm

– Velocity 2-20 m/s

• Measurements

– Improved 5-component balance :

• force 0.004 N

• moment 0.002 N.cm

Experimental setup

Ground

LC4

LC2

LC1

LC3Motor

LC6

LC5

Adjustableplate controlledby motor tochange AoA.

Left Strut

Right Strut

Aft Strutalwaysperpendicular tothe model

longitudinal axis

Pivot Point

LC1 – 4 are fixed onto the ground,LC5 - 6 are supported by adjustable plate and alwaysrotate with the model

rl

Free Stream Velocity

Kochersberger, AIAA paper 2005-4759

Suharyono, Aerospace Sc. & Tech., 2006

8MAV’08March 10-15, 2008

Future MAV wind tunnel (2009)

Length 19 mWidth 4.5 mHeight 7 m

Test section 120 x 80 cmContraction ratio 10:1Turbulence 0.3%Speed 0 - 25 m/s

Variable-pitch fan

Test section length 2.5 m

6-component sting micro-balance: 8-mm

diameter, 100g max load,

9MAV’08March 10-15, 2008

Tilt-wing vs. tilt-body

• Speed 0-15 m/s

• tilt angles 0°, 30°, 60°, 90°

• PWM signal 40, 60, 80, 100%

• Struts corrections (AoA, speed)

30°

60°

Tilt-body (0°)

10MAV’08March 10-15, 2008

Λ,S2

,2

ΛS

( )02ααπ −Λ×= qSLTB ( ) TBTW L

SqL ×=−Λ×=

2

1

422 0ααπ

( ) ( )204

1 ααπ −×Λ= qSDi TB ( ) ( ) ( )TBTW DiqSDi ×=−×Λ=2

1

8

1 20ααπ

Tilt-body Tilt-wingααα == TBTW

11MAV’08March 10-15, 2008

Λ,S2

,2

ΛS

( ) WqSL TBTB =−Λ×= 02ααπ ( ) W

SqL TWTW =−Λ×= 042

2 ααπ

02 ααα −= TBTW

( ) ( )204

1 ααπ −×Λ= TBTB qSDi ( ) ( ) ( )TBTWTW DiqSDi ×=−×Λ= 28

1 20ααπ

WLL TBTW ==Tilt-body Tilt-wing

12MAV’08March 10-15, 2008

Tilt-wing vs. tilt-body

0

10

20

30

40

50

60

70

80

90

0 5 10 15 20

Velocity (m/s)

Tilt

ang

le o

r an

gle

of a

ttack

(de

g)

tilt wing

tilt body

0

20

40

60

80

100

0 5 10 15 20

Velocity (m/s)P

WM

Sig

nal (

%)

tilt wing

tilt body

Lift = Weight

Drag = Thrust

Tilt-wing may benefitfrom vortical lift effects

Tilt-wing requires more power near hover

Downwardforce

13MAV’08March 10-15, 2008

Tilt-wing vs. tilt-body

-0,009

-0,008

-0,007

-0,006

-0,005

-0,004

-0,003

-0,002

-0,001

0

0,001

0 5 10 15 20

Velocity (m/s)

dM(C

G)/d

AoA

tilt wing

tilt body

• « CG » located at 10% chord ahead of the AC at 16 m/s

• Each gradient calculated aroundequilibrium points

• Longitudinal stabilitygreatly reduced at lowspeeds

• Tilt-wing / tilt-bodyMAV configurations very comparable

14MAV’08March 10-15, 2008

A coaxial tail-sitter mini-UAV

• Tilt-body

• Coaxial rotor in tractor configuration

• Autonomoustransition flight

• rotor 500 mm

• weight 1.6 kg (incl. IMU, GPS, Modem…)

Convair « Pogo » XFY-1

(1954)

ISAE « Vertigo » mini-UAV

(2006)

15MAV’08March 10-15, 2008

Wind tunnel measurements

-0,6

-0,4

-0,2

0

0,2

0,4

0,6

0,8

1

1,2

-20 0 20 40 60 80 100

AoA°C

L

• Sting-mounted model

• Speed 0, 5, 10, 15, 20 m/s

• AoA -10°to 90°, sideslip

• flap efficiency, throttle

• strong vortical effects on wing-alone tests

• CD0 = 0.054, L/Dmax = 4

• CL > 0.8 at AoA 15°-55°

Stallangle

16MAV’08March 10-15, 2008

Coaxial tail-sitter

-0,3

-0,25

-0,2

-0,15

-0,1

-0,05

0

0,05

0,1

-20 0 20 40 60 80 100

AoA°

CM

@C

G

stallangle

• Stabilizing vortical effects (low aspect-ratio)

• Constant pitching moment gradient after stall (CoP)

17MAV’08March 10-15, 2008

Powered tests

0

2

4

6

8

1 0

1 2

1 4

0 1 0 2 0 3 0 4 0 5 0 6 0 7 0 8 0 9 0 1 0 0

Ao A °

V(m

/s)

0,0

0,2

0,4

0,6

0,8

1,0

1,2

1,4

1,6

1,8

0 10 20 30 40 50 60 70 80 90 100

For

ces

(daN

)

0,00

0,02

0,04

0,06

0,08

0,10

0,12

0,14

0,16

0,18

Pitc

hing

mom

ent (

daN

.m)

Aircraft horizontal force

Aircraft vertical force

Propeller horizontal force

Propeller vertical force

Aircraft pitching moment

Propeller pitching moment

• Balanced lift and drag (V, AoA, throttle)

• Propeller-alone tests carried out to evaluate propellers contribution during transition

• Equilibrium found throughout transition

• Propeller-induced nose-up pitching moment (side force)

• Aircraft pitching moment almost constant in transition (flap efficiency)

18MAV’08March 10-15, 2008

Propeller-wing interaction

10 20 30 40 50 60 70 80 900

2

4

6

8

10

12

14

AoA, deg

V, m

/s

1cos22

0

2

00

3

0

4

0

=

+

+

w

V

w

w

w

V

w

w

w

w α

2

0

2

000

sin2cos2

+

+= αα

w

w

w

w

w

V

w

VR

= − αα sin2sin 0

0

1

RS V

w

w

w

T

2w V

α

αS VR

Acknowledgement: Boris Bataillé & Damien Poinsot, ISAE & ONERA, Toulouse

McCormick

exp.

+ model

19MAV’08March 10-15, 2008

Simulation of transition flight

10 20 30 40 50 60 70 80 900

5

10

15

20

AoA, deg

AoA

w, d

eg

10 20 30 40 50 60 70 80 900

5

10

15

AoA, deg

Del

ta m

, deg

10 20 30 40 50 60 70 80 900

2

4

6

8

10

12

14

AoA

Vel

ocity

, m/s

VV

r

Effective wing AoA Relative wind speed Flap deflection

• Unstalled wing (stall angle 23°)

• Flap efficiency throughout transitionPoinsot et al., VTOL conference, RAeS 2008

20MAV’08March 10-15, 2008

MiniVertigo

200

Speedcontroller

Servo

Battery cell

2 Bolts, D2

Carbon tube 6 mm

Fiberglass 2 plies/side

Fiberglass, 1 ply/side Plastic insert

46

Motor mount

RC Receiver

Carbon tube 3 mm

Elevon300

Gyro-stabilized prototype

(2007)

Dual motor-propeller 50 g

3-cell Li-Poly batteries 60 g

Speed controller & receiver 16 g

Gyros & servos 22 g

Airframe 44 g-------------------------------------------------Total weight 192 g

21MAV’08March 10-15, 2008

Scaling effects

4 x 104

16 x 104

Rec

30

65

Size

(cm)

3216051.91Mini-Vertigo

671600241.80Vertigo

m/S (g/dm²)m (g)S (dm²)ΛMAV

8.229.0451011.55140Mini-Vertigo

9010.51658015.7500Vertigo

Pind(W)

PL (W/N)

P (W)DL (N/m2)T (N)Drotor(mm)

MAV

• Mini-Vertigo used high-pitch counter-rotating propellers (low RPM)

• Power loading (PL) includes motor efficiency

Shkarayev et al., J Aircraft (submitted), 2007

22MAV’08March 10-15, 2008

Conclusions

• Tilt-body MAV configurations combining tandem/coaxial rotors and fixed wings are promising candidates for complex outdoor/indoor surveillance missions in autonomous mode

• Tilt-body better than tilt-wing MAV

• Coaxial tail-sitter allows for balanced transition

23MAV’08March 10-15, 2008

Current prospects: nanorotors

• Weight decreases with size as

• FoM and power loading decrease with size (XROTOR)

535.1884.17 dW = Acknowledgements

Liu Zhen, PhD candidate, ISAE, Feb. 2008

24MAV’08March 10-15, 2008

Current prospects: autonomy

• Step 1: altitude/attitude hold

• Step 2: autonomous transitionAcknowledgements

Roman Krashanitsa, Univ. Arizona, 2007

25MAV’08March 10-15, 2008

Current prospects : improved airframe

• Embedded coaxial fixed-wing

• No through-shaft system

• Roll & fly

Vision (35 cm, 450 g)

Frankfurt, Germany, oct. 2007

MAV’07, Toulouse, France, sept. 2007

Acknowledgements

Dominique Bernard, ISAE, 2007

26MAV’08March 10-15, 2008

Thank you

Cyclope (20 cm, 100g)