Autonomous Systems Laboratory

Dynamic Modeling of Fixed-Wing UAVs

(Fixed-Wing Unmanned Aerial Vehicles)

A. Noth, S. Bouabdallah and R. Siegwart

Version 2.0

12/2006
Dynamic Modeling of Fixed-Wing UAVs Dec 2006

1 Introduction
Dynamic modeling is an important step in the development and the control of a dynamic
system. In fact, the model allows the engineer to analyze the system, its possibilities and its
behavior depending on various conditions. Moreover, dynamic models are widely used in
control design.
This is especially important for aerial robots where the risk of damage is very high as a fall
from a few meters can seriously damage the platform. Thus, the possibility to simulate and
tune a controller before implementing it on the real machine is highly appreciable.
In this course, one will se the simplified example dynamic modeling of a lightweight model
airplane for which the Lagrange-Euler formalism will be used.

1.1 Objectives
The main objective of this course is to introduce the modeling of unmanned aerial vehicles
through the example of Sky-Sailor airplane used in research. The second objective is to make
the reader aware of the importance of modeling for engineers, especially for controlling
systems. The model presented below is derived for system’s analysis, control-law design and
simulation. Fig. 1 shows a situation where a model is used for control design.

PLANT’S
CONTROLLER
SCOPE MODEL
REFERENCE

Figure 1: Control simulation bloc diagram

2 Orientation of the UAV

Figure 2: General Coordinate System

Let’s consider Earth fixed frame E = { E X , EY , EZ } and body fixed frame B = { Ex , E y , Ez } on

the UAV. At each moment, we need to know the position and the orientation of B relative to
E.

A. Noth, S. Bouabdallah and R. Siegwart 1

Dynamic Modeling of Fixed-Wing UAVs Dec 2006

2.1.1 Rotation Parameterization

The rotation of a rigid body in space can be parameterized using several methods (Euler
angles, quaternion, Tait-Bryan angles …).
Tait-Bryan angles also called “Cardano angles” are extensively used in aerospace
engineering where they are called “Euler angles”. This conflicts with the real usage of “Euler
angles” which is a mathematical representation of three successive rotations about different
possible axes which are often confused in literature [11].

⎡1 0 0 ⎤ ⎡ cos β 0 sin β ⎤ ⎡cos γ − sin γ 0⎤

R ( x , α ) = ⎢0 cos α − sin α ⎥ R( y, β ) = ⎢ 0 1 0 ⎥ R ( z , γ ) = ⎢ sin γ cos γ 0⎥
⎢⎣0 sin α cos α ⎥⎦ ⎢⎣− sin β 0 cos β ⎥
⎦ ⎢⎣ 0 0 1⎥
⎦
Figure 3: Successive rotations in roll, pitch and yaw according to Tait-Bryan convention

In this document, we use Tait-Bryan angles to describe the orientation of our UAVs. It
consists of three successive rotations:
r
• Rotation of φ around x : roll (-π<φ<π)
r
• Rotation of θ around y : pitch (-π/2<θ<π/2)
r
• Rotation of ψ around z : yaw (-π<ψ<π)

The complete rotation matrix, called Direct Cosine Matrix is then [5]:

R (φ , θ ,ψ ) = R ( z ,ψ ) ⋅ R ( y , θ ) ⋅ R ( x, φ )

⎡ cosψ − sin ψ 0⎤⎡ cos θ 0 sin θ ⎤⎡1 0 0 ⎤

R (φ , θ ,ψ ) = ⎢ sin ψ cosψ 0⎥ ⋅ ⎢ 0 1 0 ⎥ ⋅ ⎢ 0 cos φ − sin φ ⎥
⎢ ⎥ ⎢ ⎥ ⎢ ⎥
⎢⎣ 0 0 1 ⎥⎦ ⎢⎣ − sin θ 0 cos θ ⎥⎦ ⎢⎣ 0 sin φ cos φ ⎥⎦

⎡ cosψ cos θ cosψ sin θ sin φ − sin ψ cos φ cosψ sin θ cos φ + sin ψ sin φ ⎤
R (φ , θ ,ψ ) = ⎢ sin ψ cos θ sin ψ sin θ sin φ + cosψ cos φ sin ψ sin θ cos φ − sin φ cosψ ⎥ (1)
⎢ ⎥
⎢⎣ − sin θ cos θ sin φ cos θ cos φ ⎥⎦

A. Noth, S. Bouabdallah and R. Siegwart 2

Dynamic Modeling of Fixed-Wing UAVs Dec 2006

2.1.2 Angular rates

. . .
The time variation of Tait-Bryan angles (φ ,θ ,ψ ) is a discontinuous function. Thus, it is
different from body angular rates ( p, q, r ) 1 which are physically measured with gyroscopes
for instance.
In general, an Inertial Measurement Unit (IMU) is used to measure the body rotations and
directly calculate for the Tait-Bryan angles.
We can get [8]:
⎡.⎤
⎡ p⎤ ⎢φ ⎥
⎢ q ⎥ = R ⎢θ. ⎥ (2)
⎢ ⎥ r ⎢ ⎥

⎢⎣ r ⎥⎦ ⎢.⎥
⎢ψ ⎥
⎣ ⎦
Where:
⎡1 0 − sin θ ⎤
⎢
Rr = ⎢ 0 cos φ sin φ cos θ ⎥⎥ (3)
⎢⎣ 0 − sin φ cos φ cos θ ⎥⎦

However, Tait-Bryan angles representation suffer from a singularity ( θ = ±π / 2 ) also known

as the “gimbal lock”. In practice, this limitation does not affect the UAV in normal flight
mode.

3 The Airplane example

Depending on the final applications, the configurations of an airplane, i.e its dimensions,
aerodynamics, propulsion system and controls, are very different from one model to another.
In fact, the design of a mono-place acrobatic aircraft will completely differ compared to a jet
airplane that can transport more than 400 passengers. Anyway, in both cases the principles of
flight are the same.
The example presented here is the model-scale solar airplane Sky-Sailor that is intended to
achieve continuous flight with the only energy of the sun. It consists of a glider with a
wingspan of 3.2 meters that is motorized by a DC motor connected to the propeller (a)
through a gearbox. The control surfaces are:
• Two ailerons (b) on the main wing that act mainly on the roll of the airplane
• The V-tail (c) at the rear side composed of two control surfaces that act mainly on the
pitch if they change symmetrically (elevator) and on the yaw if their deviation is not
symmetrical (rudder).

1
Notation used in aerospace engineering

A. Noth, S. Bouabdallah and R. Siegwart 3

Dynamic Modeling of Fixed-Wing UAVs Dec 2006

b
c

pitch

a B

roll x b

Y z
E

X yaw

Figure 9: The controls on Sky-Sailor are the propeller,

the ailerons and the V-tail that mixes rudder and elevator

3.1 Dynamic modeling

The forces acting on the airplane are mainly:
• the weight m·g located at the center of gravity
• the thrust of the propeller acting in the x direction. Modeling a propeller is a very
complex task, that is the reason why values measured by experiments will be taken.
• the aerodynamic forces of each part of the airplane, mainly the wing and the V-tail.
Here, the theory is simpler and will be topic of the next chapter.

3.2 Aerodynamic forces of an airfoil

The figure below shows the section of a wing also called airfoil. The chord of the wing is
the line between the leading and the trailing edge and the angle between the relative speed
and this chord is the angle of attack (Aoa).
As every other solid moving in a fluid at a certain speed, one can represent the sum of all
aerodynamic forces acting on the wing with two perpendicular forces: the lift force Fl and
the drag force Fd that are respectively perpendicular and parallel to the speed vector.

A. Noth, S. Bouabdallah and R. Siegwart 4

Dynamic Modeling of Fixed-Wing UAVs Dec 2006

Fl

Leading edge Chord length

Angle of attack
Trailing edge
Fd
Chord
Relative
Wind 25 % Chord Thickness

Figure 10: Section of the wing (airfoil) and the lift and drag forces

The application point of the lift and drag forces is very close to the 25% of the chord but
this can slightly change depending on the angle of attack. In order to simplify the problem,
the application point is considered as fixed and a moment is added to correct this
assumption.
ρ
Fl = Cl Sv 2 (18)
2
ρ
Fd = Cd Sv 2 (19)
2
ρ
M = Cm Sv 2 ⋅ chord (20)
2
with ρ : Density of fluid (air)
S : Wing area
v : Flight speed (relative to surrounding fluid)
CL : Lift coefficient
CD : Drag coefficient
CM : Moment coefficient

The lift, drag and moment coefficients depend on the airfoil, the angle of attack and a third
value that is the Reynolds number. It is representative of the viscosity of the fluid but will
not be explained more in details. Cl increases almost linearly with the angle of attack until
the stall angle is reached. The wing should never work in this zone where the Cl decreases
importantly which makes the airplane loosing altitude very rapidly. Cd has a quadratic
relation to the angle of attack.

A. Noth, S. Bouabdallah and R. Siegwart 5

Dynamic Modeling of Fixed-Wing UAVs Dec 2006

Figure 11: Lift and drag coefficients of Sky-Sailor airfoil

3.3 Modeling assumptions

The airplane is considered as a rigid body associated with the aerodynamic forces generated
by the propeller and the wing.
• The center of mass and the body fixed frame origin are assumed to coincide.
• The structure is supposed rigid and symmetric (diagonal inertia matrix).
• The drag force of the fuselage is neglected
• The wind speed in the Earth frame is set to zero so that the relative wind on the body
frame is only due to the airplane speed.
• The relative wind induced by the rotation of the airplane is neglected

3.4 Model derivation using Lagrange-Euler approach

The Lagrange-Euler approach is based on the concept of kinetic and potential energy:
d ⎛ δL ⎞ δL
Γi = ⎜ ⎟⎟ − (21)
dt ⎜⎝ δq& i ⎠ δqi
L = T −V (22)
avec qi : generalized coordinates
Γi : generalized forces given by non-conservatives forces
T : total kinetic energy
V : total potential energy
The kinetic energy due to the translation is immediately:
1 1 1
Ecin translation = mx& 2 + my& 2 + mz& 2 (23)
2 2 2

A. Noth, S. Bouabdallah and R. Siegwart 6

Dynamic Modeling of Fixed-Wing UAVs Dec 2006

As we state it in the hypothesis, we assume that the inertia matrix is diagonal and thus, that
the inertia products are null. The kinetic energy due to the rotation is:
1 1 1
Ecin rotation = I xxω x2 + I yyω y2 + I zzω z2 (24)
2 2 2
Where ω , ω , ω are the rotational speed that can be expressed as a function of the roll,
x y z
pitch and yaw rate (φ&,θ&,ψ& ) :

⎛ ωx ⎞ ⎛ φ& −ψ& ⋅ sin θ ⎞

⎜ ⎟ ⎜ & ⎟
⎜ ω y ⎟ = ⎜ θ ⋅ cos φ + ψ ⋅ sin φ cos θ ⎟
& (25)
⎜ ⎟ ⎜ & ⎟
⎝ ω z ⎠ ⎝ −θ ⋅ sin φ +ψ& ⋅ cos φ cos θ ⎠
This leads to the total kinetic energy:

T=
1
2
( mx& 2
+ my& + mz&
2 2
+ I xxω x2 + I yyω y2 + I zzω z2 ) (26)

The potential energy can be expressed by:

V = − m ⋅ g ⋅ Z = − m ⋅ g ⋅ ( − sin θ ⋅ x + sin φ cos θ ⋅ y + cos φ cos θ ⋅ z ) (27)

The Lagrangian is :
L = T −V (28)

The motion equations are then given by:

d ⎛ ∂L ⎞ ∂L d ⎛ ∂L ⎞ ∂L d ⎛ ∂L ⎞ ∂L
⎜ ⎟ − = Fx ⎜ ⎟ − = Fy ⎜ ⎟ − = Fz (29)
dt ⎝ ∂x& ⎠ ∂x dt ⎝ ∂y& ⎠ ∂y dt ⎝ ∂z& ⎠ ∂z
d ⎛ ∂L ⎞ ∂L d ⎛ ∂L ⎞ ∂L d ⎛ ∂L ⎞ ∂L
⎜ ⎟− = τψ ⎜ ⎟ − = τφ ⎜ ⎟ − = τθ (30)
dt ⎝ ∂ψ& ⎠ ∂ψ dt ⎝ ∂φ& ⎠ ∂φ dt ⎝ ∂θ& ⎠ ∂θ

After calculation, we obtain the left parts of equations above:

d ⎛ ∂L ⎞ ∂L
⎜ ⎟ − = mx&& + mg sin θ (31)
dt ⎝ ∂x& ⎠ ∂x
d ⎛ ∂L ⎞ ∂L
⎜ ⎟ − = my&& − mg sin φ cos θ
dt ⎝ ∂y& ⎠ ∂y
(32)

d ⎛ ∂L ⎞ ∂L
⎜ ⎟ − = mz&& − mg cos φ cos θ (33)
dt ⎝ ∂z& ⎠ ∂z
d ⎛ ∂L ⎞ ∂L
⎜ ⎟− = I xxω& x − ( I yy − I zz )ω yω z (34)
dt ⎝ ∂φ& ⎠ ∂φ
d ⎛ ∂L ⎞ ∂L
⎜ ⎟− = − sin φ (ω& z I zz − ω xω y ( I xx − I yy ))
dt ⎝ ∂θ& ⎠ ∂θ (35)
+ cos φ (ω& y ⋅ I yy − ω xω z ( I zz − I xx ))

A. Noth, S. Bouabdallah and R. Siegwart 7

Dynamic Modeling of Fixed-Wing UAVs Dec 2006

d ⎛ ∂L ⎞ ∂L
⎜ ⎟− = − sin θ ⋅ (ω& x I xx − ω yω z ( I yy − I zz ))
dt ⎝ ∂ψ& ⎠ ∂ψ
+ sin φ cos θ ⋅ (ω& y I yy − ω xω z ( I zz − I xx )) (36)
+ cos φ cos θ ⋅ (ω& z I zz − ω xω y ( I xx − I yy ))

The non-conservative forces and moments come from the aerodynamics. On the airplane,
seven parts are considered as depicted on the figure below where the right and left side of
are divided into a portion with and without control surface.

Fl5
M7 Fl7
M5 Fl6
Fd5
Fl4 Fd7

M4 Fd6
Fd4 M6
Fl3
M3
Fd3
Fl2

M2 Fl1
Fd2
Fprop M1
Fd1

Figure 12: Forces and moments on the airplane

7
Ftot = Fprop + ∑ FLi + Fdi
i =1
7
M tot = ∑ M i + FLi × ri + Fdi × ri
i =1

Fprop = f ( x&, U1 )
ρ
Fli = Cli Si v 2
2
ρ (37)
Fdi = Cdi Si v 2
2
ρ
M i = Cmi Si v 2 ⋅ chordi
2

A. Noth, S. Bouabdallah and R. Siegwart 8

Dynamic Modeling of Fixed-Wing UAVs Dec 2006

[Cl1 Cd 1 Cm1 ] = f ( Aoai ,U 2 )

[Cli Cdi Cmi ] = f ( Aoai ) for i=2,3,4
[Cl 5 Cd 5 Cm5 ] = f ( Aoai ,U 3 ) (38)

[Cl 6 Cd 6 Cm 6 ] = f ( Aoai ,U 4 )
[Cl 7 Cd 7 Cm7 ] = f ( Aoai ,U 5 )

U1 to U5 are the control inputs: U1 is the voltage on the motor and U2, U3, U4, U5 are
respectively the deflection of the left aileron, the right aileron, the left V-tail and the right
V-tail.

Final model with small angle approximation

Isolating the acceleration and applying the small angle approximation, where the rotational
speed in the solid basis are equal to Euler’s angles rates, we obtain:
⎧ F ⎧ Ftot , X
⎪ &&x = tot , x − g sin θ ⎪ X&& =
⎪ m ⎪ m
⎪ && Ftot , y ⎪Y&& = tot ,Y
F
⎪ y = m + g sin φ cos θ ⎪ m
⎪ F ⎪ Ftot , Z
⎪ &&z = tot , z + g cos φ cos θ ⎪ Z&& = m + g
m
⎪⎪ ⎪⎪
⎨φ&& = I yy − I zz ψθ M tot , x or in Earth frame ⎨φ&& = I yy − I zz ψθ M tot , x (39)
&&+ &&+
⎪ I xx I xx ⎪ I xx I xx
⎪ ⎪
⎪ && I zz − I xx & M tot , y ⎪ && I zz − I xx & M tot , y
⎪θ = I ψφ
& +
I yy
⎪θ = I ψφ
& +
I yy
⎪ yy ⎪ yy
⎪ I −I M ⎪ I −I M
⎪ψ&& = xx yy θφ & & + tot , z ⎪ψ&& = xx yy θφ & & + tot , z
⎪⎩ I zz I zz ⎪⎩ I zz I zz

Note that only the angular dynamics were developed using the Lagrangian approach.

3.5 Open-Loop behavior

The simulation of the airplane shows a good stability in open loop, this means if the motor is
turned off and all the deflection of the ailerons are set to zero. This behavior of natural
stabilization was also tested on the real airplane during experiments.
The following figure shows the evolution of the airplane that start with a pitch angle of 0.2
rad (11.5°). One can see that after 40 seconds, the airplane is quite stable. With horizontal
and vertical speeds of 8.2, respectively 0.32 [m/s], we obtain a glide slope of 25.6 which is
very close to the reality.

A. Noth, S. Bouabdallah and R. Siegwart 9

Dynamic Modeling of Fixed-Wing UAVs Dec 2006

Figure 13: Open-loop behavior with an initial pitch of 0.2 rad

4 Model verification
Like all models, the airplane model developed above contains physical parameters which
must be known with good precision.
These parameters are also m, Ixx, Iyy, Izz and all the aerodynamic coefficients Cli, Cdi, Cmi that
depends on the speed and the angle of attack. There exist software that allow a calculation of
these values but the better way is for sure a direct measurement of the forces and the torques
on a piece of the wing placed in a wind tunnel.

A. Noth, S. Bouabdallah and R. Siegwart 10

Dynamic Modeling of Fixed-Wing UAVs Dec 2006

5 References

[1] http://www.asl.ethz.ch/
[2] R.M. Murray, Z. Li, and S.S. Sastry, A mathematical introduction to robotic manipulation. CRC
Press, 1994.
[3] S. Boubdallah, P.Murrieri and R.Siegwart, Design and Control of an Indoor Micro Quadrotor.
ICRA 2004, New Orleans (USA), April 2004.
[4] S. Boubdallah, A.Noth and R.Siegwart, PID vs LQ Control Techniques Applied to an Indoor
Micro Quadrotor. IROS 2004, Sendai (Japan), October 2004.
[5] R. Olfati-Saber, Nonlinear Control of Underactuated Mechanical Systems with Application to
Robotics and Aerospace Vehicles. Phd thesis, Department of Electrical Engineering and Computer
Science, MIT, 2001.
[6] H. Baruh, Analytical Dynamics.McGraw-Hill, 1999.
[7] A. Noth, Synthèse et Implémentation d’un Contrôleur pour Micro Hélicoptère à 4 Rotors, Diploma
Project 2004.
[8] Robert F. Stengel, Flight dynamics. Princeton University Press, 2004.
[9] A. Noth, W. Engel, R. Siegwart, Design of an Ultra-Lightweight Autonomous Solar Airplane for
Continuous Flight. In Proceedings of Field and Service Robotics, Port Douglas, Australia, 2005.
[10] http://www.mh-aerotools.de/airfoils/javafoil.htm
[11] http://en.wikipedia.org/wiki/Tait-Bryan_angles

A. Noth, S. Bouabdallah and R. Siegwart 11