IFASD-2024-129

AEROELASTIC ANALYSIS OF DISTRIBUTED ELECTRIC PROPULSION

FLEXIBLE WINGS

Samet T. Dede1, Ali Tatar1,2, Djamel Rezgui1 and Jonathan E. Cooper1

1
University of Bristol, School of Civil, Aerospace and Design Engineering
Queen’s Building, University Walk
Bristol BS8 1TR, United Kingdom
[email protected]
[email protected]
[email protected]
[email protected]
2
Istanbul Technical University, Faculty of Aeronautics and Astronautics
Ayazaga Campus, 34469, Maslak/Istanbul, Turkey
[email protected]

Keywords: aeroelasticity, distributed electric propulsion, wing flutter, whirl flutter

Abstract: There is much current emphasis on the development of alternative aircraft propulsion
technologies to enable much reduced, and eventually, net-zero in-flight commercial aircraft
emissions. The success of this goal is not simply dependent upon the advancements in electric or
hydrogen-powered propulsion systems, but also on how to integrate them into the aircraft
structures. It is likely that distributed electric propulsion (DEP) configurations featuring small
engines spaced across the wing will be the most viable solution. However, the adoption of these
novel aircraft wing configurations might initiate the early onset of two potential aeroelastic
instabilities: wing flutter and whirl flutter, which must be addressed in the early design stages of
DEP aircraft wings. The main aim of this study is to create and evaluate a representative low-order
aeroelastic coupled wing-propeller model for parametric aeroelastic studies that can be used in the
flutter analysis at the early design stages. An aeroelastic numerical model was developed in
MATLAB to analyse the aeroelastic behaviour of a coupled flexible cantilever wing with a variable
number of flexibly mounted propellers/rotors. Reed’s model is employed to model the propeller
dynamics, with the structural model of the wing being derived through the assumed-mode
Rayleigh-Ritz method. The aerodynamic model of the wing was obtained from a combination of
the modified strip theory and Theodorsen’s unsteady aerodynamic theory. The proposed coupled
aeroelastic model can successfully estimate both wing and whirl flutter in DEP wings through
validations with results from the literature. The model was then used for several parametric
analyses investigating the effects of propeller spanwise position, advance ratio and rotor radius on
the stability of the integrated wing-propeller system. The parametric studies demonstrated that
advance ratio and rotor radius have a significant effect on the stability of the coupled wing-
propeller model. It was found that increasing the advance ratio has a destabilizing effect, whereas
increasing the rotor radius has a stabilizing effect.

1
 IFASD-2024-129

1 INTRODUCTION
To decrease in-flight emissions in commercial aviation, there has been a significant focus on the
idea of hybrid or fully electric aircraft in recent times [1, 2]. Unlike conventional aircraft powered
by gas turbine engines, electric aircraft generate the necessary lift and thrust using propellers
distributed along the wingspan, known as Distributed Electric Propulsion (DEP). Such a
configuration enables zero-emission propulsion using multiple electric motors. Recent examples
include the NASA X-57 Maxwell and DLR electric regional aircraft concepts, as seen in Figure 1.
Owing to the current constraints in battery technologies, DEP aircraft typically feature high aspect
ratio wings to enhance the aerodynamic efficiency and minimize the weight of the aircraft [3].

a) b)

Figure 1: Examples of Distributed Electric Propulsion Aircraft

a) NASA X-57 Maxwell [4], b) DLR EXACT [5]
These novel aircraft wing configurations bring about the possibility of aeroelastic instabilities.
Whirl flutter in particular occurs on aircraft with a flexibly attached propeller, and is defined by a
diverging spiral motion caused by the precession of the propeller. Unlike classical wing flutter,
which is caused by the interaction of aerodynamic, elastic, and inertial forces, whirl flutter is a
result of the interplay between aerodynamic and gyroscopic forces acting on the propeller [6]. The
onset of wing flutter or propeller whirl flutter can result in serious damage to aircraft structures
and even fatal accidents. Hence, it is crucial to account for these aeroelastic instabilities during the
initial design stages of DEP aircraft wings.

Dynamic aeroelastic instabilities have been extensively studied in the context of classical wing
flutter and propeller whirl flutter. For classical wing flutter, a structural model of a flexible wing
is employed, coupled with an aerodynamic model being represented by suitable unsteady
aerodynamics theories, such as the strip theory or Doublet-Lattice-Method (DLM). The classical
whirl flutter theory is presented for a flexibly mounted rigid propeller in pitch and yaw in [7-10].
Additionally, the aerodynamics of the propeller is established based on the blade geometry for
propellers operating under windmilling conditions in both pitch and yaw motions [11, 12]. For a
coupled wing-propeller model, Bennett and Bland [10] delivered an analytical approach and
compared it with experimental data.

However, a limited number of investigations on the aeroelastic behaviour of DEP wings have been
conducted, yet there is still potential for additional investigations. To understand the effect of
propellers on the aeroelastic behaviour of the wing, a general methodology is also described in

2
 IFASD-2024-129

[13]. Amoozgar et al. [14] devised a method to incorporate the influence of propellers on wing
aeroelastic characteristics by representing the electric propeller as a concentrated mass on the wing
and included the thrust force generated by the propeller as a follower force when formulating the
aeroelastic governing equations. Heeg et al. [15] extensively investigated the stability of propeller
whirl flutter during the development of the NASA X-57 Maxwell aircraft through multibody
dynamics simulation tools. Böhnisch et al. [16-18] introduced an aeroelastic model to examine the
whirl flutter of DEP wings and conducted various parametric studies on DEP wings. In their model,
the propellers are attached to a flexible wing through rigid pylons that can move in both pitch and
yaw directions. Recently, Tamer and Tatar [19] proposed a minimum complexity model for the
aeroelastic analysis of wing-propeller systems. Their study primarily shows how rapidly the model
can be updated during various parametric studies on DEP wings. Although the majority of these
studies are based on finite element and finite element volume methods, which have advantages
such as more accurate representation of the wing, lower-order models can still be useful for rapid
analysis, optimization purposes, or real-time simulations.

None of the aforementioned studies addressed the aeroelastic analysis of an electric aircraft wing
featuring a DEP configuration using the Rayleigh-Ritz approach, which can offer several
advantages due to its easy usage and reduced computational time. It maintains a large portion of
accuracy, especially if enough assumed modes are used. Therefore, this study aims to create and
evaluate a representative low-order aeroelastic coupled wing-propeller model in parametric
aeroelastic studies that can be used in flutter analysis at the early design stages. The analysis is
then used to examine the aeroelastic stability of a DEP aircraft wing. A systematic parameter
analysis is conducted by changing the propeller position and propeller parameters such as advance
ratio and rotor radius, identifying their impact on the stability of a representative DEP wing. The
propeller parameters above are some of the major parameters that can influence whirl flutter. Their
impact on an isolated propeller on a rigid wing has been well understood, whereas less research
has considered the case of a propeller flexibly attached to a flexible wing and the resulting
aeroelastic instabilities of DEP wings. This study presents a coupled aeroelastic model, which can
be used in the preliminary design stages of DEP wings and discusses the results of several
parametric analyses conducted on the stability of DEP wings.

For this study, a baseline wing and propeller model are defined and used for creating a coupled
propeller-wing model. An aeroelastic numerical model is developed in MATLAB to analyse the
aeroelastic behaviour of the coupled system with no aerodynamic interference effects between the
wing and propeller for a freestream velocity range. Reed’s model is used to represent the propeller
dynamics, while the wing's structural model is derived using the assumed-mode Rayleigh-Ritz
method. The aerodynamic model of the wing is obtained by combining the modified strip theory
with Theodorsen’s unsteady aerodynamics theory. The developed coupled aeroelastic model is
subsequently used for several parametric analyses. The key findings are summarised in section 4.

2 MATHEMATICAL MODELLING
2.1 Derivation of Aeroelastic Equations of Motion of Standalone Wing
The simple rectangular cantilever wing with semi-span 𝑠 and chord 𝑐 considered to derive the
aeroelastic equations of motion has bending rigidity 𝐸𝐼 and torsional rigidity 𝐺𝐽. The equations of
motion of a wing are composed of the structural model and the aerodynamic model. In this study,
the structural model of the wing is represented by the Rayleigh-Ritz method, which is applied to

3
 IFASD-2024-129

model the deformation of the two-dimensional wing system. At any point along the wing, the
deformation 𝑧(𝑥,𝑦,𝑡) caused by bending and twisting is expressed by the following series [20]
𝑁

𝑧(𝑥, 𝑦, 𝑡) = ∑ 𝜓𝑖 (𝑥, 𝑦)𝑞𝑖 (𝑡) (1)

𝑖=1

In this series, 𝜓𝑖 (𝑥, 𝑦) represents one of N assumed deformation shapes, while 𝑞𝑖 (𝑡) denotes the
coefficients of generalized coordinates of the unknown magnitude. As the cantilever wing is
clamped at the root, the following fixed boundary conditions are imposed.
𝑧 = 𝑧̇ = 0 𝑎𝑡 𝑦 = 0 (2)
Two mode shapes (one bending and one torsion) are assumed. For bending and twisting about an
elastic axis,
𝑧(𝑥, 𝑦, 𝑡) = ℎ + (𝑥 − 𝑥𝑓 )𝛼 (3)
where 𝑥𝑓 is the chord-wise position of the elastic axis from the leading edge. The bending ℎ and
torsion 𝛼 deflections are expressed through the Rayleigh-Ritz in terms of mode shapes and
generalized coordinates by
𝑦 2 𝑦
ℎ(𝑦, 𝑡) = ( 𝑠 ) 𝑞ℎ 𝛼(𝑦, 𝑡) = ( 𝑠 ) 𝑞𝛼 (4)

Using these generalized coordinates 𝑞𝑖 (𝑡) and omitting the damping term, Lagrange's equations
can be then expressed as
𝑑 𝜕𝑇 𝜕𝑇 𝜕𝑈 𝜕𝐷
( )− + + = 𝑄𝑖 𝑖 = ℎ, 𝛼 (5)
𝑑𝑡 𝜕𝑞̇ 𝑖 𝜕𝑞𝑖 𝜕𝑞𝑖 𝜕𝑞̇ 𝑖
where 𝑄𝑖 denotes the generalized forces. T is the kinetic energy, U is the potential energy, and D
is the dissipation or damping function. The equations of motion of the wing can be derived through
Lagrange’s equations. The kinetic energy of the wing is

𝑚 𝑠 𝑐 ̇ 2
𝑇 = ∫ ∫ (𝑧) 𝑑𝑥 𝑑𝑦
2 0 0

2
𝑚 𝑠 𝑐 𝑦 2 𝑦
𝑇= ∫ ∫ (( ) 𝑞̇ ℎ + ( ) (𝑥 − 𝑥𝑓 )𝑞̇ 𝑎 ) 𝑑𝑥 𝑑𝑦
2 0 0 𝑠 𝑠

where 𝑚 is mass per unit area.

𝑠 𝑐
𝜕𝑇 𝑦 4 𝑦 3
( ) = 𝑚 ∫ ∫ (( ) 𝑞̇ ℎ + ( ) (𝑥 − 𝑥𝑓 )𝑞̇ 𝑎 ) 𝑑𝑥 𝑑𝑦 (6)
𝜕𝑞̇ ℎ 0 0 𝑠 𝑠

𝑑 𝜕𝑇 𝑠𝑐 𝑠 𝑐2
( ) = 𝑚 [ 𝑞̈ ℎ + ( − 𝑐𝑥𝑓 ) 𝑞̈ 𝑎 ]
𝑑𝑡 𝜕𝑞̇ ℎ 5 4 2
𝑠 𝑐
𝜕𝑇 𝑦 3 𝑦 2 2
( ) = 𝑚 ∫ ∫ (( ) (𝑥 − 𝑥𝑓 )𝑞̇ ℎ + ( ) (𝑥 − 𝑥𝑓 ) 𝑞̇ 𝑎 ) 𝑑𝑥 𝑑𝑦
𝜕𝑞̇ 𝑎 0 0 𝑠 𝑠

4
 IFASD-2024-129

𝑑 𝜕𝑇 𝑠 𝑐2 𝑠 𝑐3
( ) = 𝑚 [ ( − 𝑐𝑥𝑓 ) 𝑞̈ ℎ + ( − 𝑐 2 𝑥𝑓 + 𝑐𝑥𝑓 2 ) 𝑞̈ 𝑎 ]
𝑑𝑡 𝜕𝑞̇ 𝑎 4 2 3 3

The potential energy caused by the strain energy effect over the wing is expressed as
2
1 𝑠 𝜕 2𝑧 1 𝑠 𝜕𝑎 2
𝑈 = ∫ 𝐸𝐼 ( 2 ) 𝑑𝑦 + ∫ 𝐺𝐽 ( ) 𝑑𝑦
2 0 𝜕𝑦 2 0 𝜕𝑦

2 2
1 𝑠 2 1 𝑠 1
𝑈 = ∫ 𝐸𝐼 ( 2 𝑞ℎ ) 𝑑𝑦 + ∫ 𝐺𝐽 ( 𝑞𝑎 ) 𝑑𝑦 (7)
2 0 𝑠 2 0 𝑠

𝜕𝑈 4𝐸𝐼 𝜕𝑈 𝐺𝐽
( ) = 3 𝑞ℎ ( ) = 𝑞𝑎
𝜕𝑞ℎ 𝑠 𝜕𝑞𝑎 𝑠

Therefore, structural equations become

𝑠𝑐 𝑠 𝑐2 4𝐸𝐼
( − 𝑐𝑥𝑓 ) 0
5 4 2 𝑞̈ 3 𝑞ℎ 0
𝑚 2 3 [ ℎ] + [ 𝑠 ] [𝑞 ] = [ ] (8)
𝑠 𝑐 𝑠 𝑐 𝑞̈ 𝑎 𝐺𝐽 𝑎 0
( − 𝑐𝑥𝑓 ) ( − 𝑐 2 𝑥𝑓 + 𝑐𝑥𝑓 2 ) 0
[4 2 3 3 ] 𝑠
𝑡𝑡𝑡𝑡𝑡
In this study, the aerodynamic model is represented by a combination of the simplified Theodorsen
method with the aerodynamic strip theory, incorporating unsteady aerodynamic derivative 𝑀𝑎̇ to
account for pitch damping. 𝑀𝑎̇ = −1.2 is incorporated into the moment equation to enhance the
model's flutter behaviour predictions [20]. To determine the total lift and moment, the lift (𝑑𝐿)
and moment (𝑑𝑀) for each strip 𝑑𝑦 are provided in Equation (9). With the assumption of constant
𝑀𝑎̇ , this model remains unaffected by changes in the reduced frequency and thus

1 2 𝑧̇ 1 𝑧̇ 𝑎̇
𝑑𝐿 = 𝜌𝑉 𝑐𝑎𝑤 (𝑎 + ) 𝑑𝑀 = 𝜌𝑉 2 𝑐 2 [𝑒𝑎𝑤 (𝑎 + ) + 𝑀𝑎̇ 𝑐 ] (9)
2 𝑉 2 𝑉 4𝑉
where 𝑎𝑤 is the two-dimensional lift curve slope and 𝑒 denotes the eccentricity ratio between the
elastic axis and the aerodynamic centre. The incremental work (𝑊) done by these forces and
moments can be written as
𝑠
𝑦 2 𝑦
𝛿𝑊 = ∫ [𝑑𝐿 (− ( ) 𝛿𝑞ℎ ) + 𝑑𝑀 (( ) 𝛿𝑞𝑎 )] (10)
0 𝑠 𝑠

With the addition of the incremental work done (𝑊) by the aerodynamic forces, the generalised
forces are specified as
𝑠
𝑠
𝜕(𝛿𝑊) 1 𝑦 4 𝑞̇ ℎ 𝑦 3
𝑄𝑞ℎ = = − ∫ 𝑦 2 𝑑𝐿 = − 𝜌𝑉 2 𝑐𝑎𝑤 ∫ (( ) + ( ) 𝑞𝑎 ) 𝑑𝑦
𝜕(𝛿𝑞ℎ ) 2 0 𝑠 𝑉 𝑠
0

5
 IFASD-2024-129

1 𝑠 𝑠
𝑄𝑞ℎ = − 𝜌𝑉 2 𝑐𝑎𝑤 ( 𝑞̇ ℎ + 𝑞𝑎 ) (11)
2 5𝑉 4
𝑠
𝑠
𝜕(𝛿𝑊) 𝑦 1 𝑦 3 𝑞̇ ℎ 𝑦 2 𝑦 2 𝑞̇ 𝑎
𝑄𝑞𝑎 = = ∫ ( ) 𝑑𝑀 = 𝜌𝑉 2 𝑐 2 ∫ [𝑒𝑎𝑤 (( ) + ( ) 𝑞𝑎 ) + 𝑀𝑎̇ 𝑐 ( ) ] 𝑑𝑦
𝜕(𝛿𝑞𝑎 ) 𝑠 2 0 𝑠 𝑉 𝑠 𝑠 4𝑉
0

1 2 2 𝑠 𝑠 𝑠
𝑄𝑞𝑎 = 𝜌𝑉 𝑐 [𝑒𝑎𝑤 ( 𝑞̇ ℎ + 𝑞𝑎 ) + 𝑀𝑎̇ 𝑐 ( 𝑞̇ )]
2 4𝑉 3 12𝑉 𝑎

Hence, aerodynamic equations become

𝑐𝑠𝑎𝑤 𝑐𝑠𝑎𝑤
− 0 0 −
10 𝑞̇ 8 𝑞ℎ
𝜌𝑉 [ 2 3 ] [ ℎ ] + 𝜌𝑉 2 [ 2 ] [𝑞 ] (12)
𝑐 𝑠𝑒𝑎𝑤 𝑐 𝑠𝑀𝑎̇ 𝑞̇ 𝑎 𝑐 𝑠𝑒𝑎𝑤 𝑎
0
8 24 6

Implementing Lagrange’s equations to the expressions derived for the kinetic and potential energy
of the wing yields the aeroelastic equations of motion of the wing given in Equation (13).
𝑀𝑞̈ + (𝜌𝑉𝐶𝑎𝑒𝑟𝑜 + 𝐶𝑠𝑡𝑟𝑢𝑐𝑡𝑢𝑟𝑎𝑙 )𝑞̇ + (𝜌𝑉 2 𝐾𝑎𝑒𝑟𝑜 + 𝐾𝑠𝑡𝑟𝑢𝑐𝑡𝑢𝑟𝑎𝑙 )𝑞 = 0 (13)
where 𝑞 is the vector of generalized coordinates, 𝑀 is the mass matrix, 𝐶𝑎𝑒𝑟𝑜 is the aerodynamic
damping matrix, 𝐶𝑠𝑡𝑟𝑢𝑐𝑡𝑢𝑟𝑎𝑙 is the structural damping matrix, 𝐾𝑠𝑡𝑟𝑢𝑐𝑡𝑢𝑟𝑎𝑙 is the structural stiffness
matrix and 𝐾𝑎𝑒𝑟𝑜 is the aerodynamic stiffness matrix. Equation (13) can be written in matrix form
as
𝑠𝑐 𝑠 𝑐2 𝑐𝑠𝑎𝑤
( − 𝑐𝑥𝑓 ) 0
5 4 2 𝑞̈ ℎ 10 𝑞̇
𝑚 [ ] + 𝜌𝑉 [ 2 3 ] [ ℎ]
𝑠 𝑐 2
𝑠 𝑐 3 𝑞̈ 𝑎 𝑐 𝑠𝑒𝑎𝑤 𝑐 𝑠𝑀𝑎̇ 𝑞̇ 𝑎
( − 𝑐𝑥𝑓 ) ( − 𝑐 2 𝑥𝑓 + 𝑐𝑥𝑓 2 ) − −
[4 2 3 3 ] 8 24
𝑐𝑠𝑎𝑤 4𝐸𝐼
0 0
2 8 𝑠 3 𝑞ℎ 0
+ {𝜌𝑉 [ 2 ]+[ ]} [𝑞 ] = [ ] (14)
𝑐 𝑠𝑒𝑎𝑤 𝐺𝐽 𝑎 0
0 − 0
6 𝑠

Equation (14) is now in state-space form, which can be rearranged as

𝑞̇ 0 𝐼 𝑞
[ ]=[ −1 (𝐾 −1 ] [ ] (15)
𝑞̈ −𝑀 𝑠𝑡𝑟𝑢𝑐𝑡𝑢𝑟𝑎𝑙 − 𝐾𝑎𝑒𝑟𝑜 ) −𝑀 (𝐶𝑠𝑡𝑟𝑢𝑐𝑡𝑢𝑟𝑎𝑙 − 𝐶𝑎𝑒𝑟𝑜 ) 𝑞̇

where 𝟎 and 𝑰 denote two by two zero and identity matrices, respectively. The structural damping
of the wing is neglected since 𝐶𝑠𝑡𝑟𝑢𝑐𝑡𝑢𝑟𝑎𝑙 is set to zero.
The Jacobian matrix is then written as
0 𝐼
𝐽=[ ] (16)
−𝑀−1 (𝐾𝑠𝑡𝑟𝑢𝑐𝑢𝑡𝑎𝑙 − 𝐾𝑎𝑒𝑟𝑜 ) −𝑀−1 (𝐶𝑠𝑡𝑟𝑢𝑐𝑡𝑢𝑟𝑎𝑙 − 𝐶𝑎𝑒𝑟𝑜 )

6
 IFASD-2024-129

The frequency and damping of the system can be calculated from the eigenvalues of the Jacobian
matrix, while the mode shapes can be obtained from the corresponding eigenvectors. The
undamped natural frequency ω and damping ratio ζ for a given mode are derived through the real
and imaginary parts of its eigenvalue λ, as given in Equation (17).
−𝑅𝑒(𝜆)
𝜔 = √𝑅𝑒(𝜆)2 + 𝐼𝑚(𝜆)2 𝜁= (17)
𝜔
Increasing the number of terms incorporated into the model enhances the accuracy of the results.
For two bending and two torsion assumed modes, the deflection can be represented as
𝑦 2 𝑦 3 𝑦 𝑦 2
𝑧 = ( ) 𝑞ℎ1 + ( ) 𝑞ℎ2 + ( ) (𝑥 − 𝑥𝑓 )𝑞𝑎1 + ( ) (𝑥 − 𝑥𝑓 )𝑞𝑎2 (18)
𝑠 𝑠 𝑠 𝑠

2.2 Derivation of Aeroelastic Equations of Motion of Standalone Propeller

The references [12, 21, 22] describe a basic model of an isolated propeller with two degrees of
freedom. In this model, a rotor with radius 𝑅 and angular velocity Ω, possessing a moment of
inertia 𝐼𝑥 about its rotational axis, is allowed to move in both pitch 𝜃 and yaw 𝜓 about an effective
pivot point with a moment of inertia 𝐼𝑛 . The dynamics of the wing structure are represented by
structural stiffnesses (𝐾𝜃 , 𝐾𝜓 ) and structural damping (𝐶𝜃 , 𝐶𝜓 ) in both pitch and yaw directions at
the pivot point. The rotor is attached to the pivot point at a distance of 𝑎𝑅. The schematic of the
model is presented in Figure 2. Aerodynamic loads are calculated using the strip theory.

Figure 2: Schematic of Rotor-Nacelle Dynamic System [21]

The basic equations of motion for this rotor-nacelle dynamic system, as presented by Bielawa [21],
can be expressed as
𝐼𝑛 𝜃̈ + 𝐶𝜃 𝜃̇ − 𝐼𝑥 Ω𝜓̇ + 𝐾𝜃 𝜃 = 𝑀𝜃
𝐼𝑛 𝜓̈ + 𝐶𝜓 𝜓̇ + 𝐼𝑥 Ω𝜃̇ + 𝐾𝜓 𝜓 = 𝑀𝜓 (19)

7
 IFASD-2024-129

where 𝑀𝜃 and 𝑀𝜓 denote aerodynamic moments in pitch and yaw directions, respectively. The
aeromechanical model of this dynamic system in matrix form can thus be defined as
𝐼 0 𝜃̈ 𝐶𝜃 −𝐼𝑥 Ω 𝜃̇ 𝐾𝜃 0 𝜃 𝑀𝜃
[𝑛 ] [ ̈ ] + [𝐼 Ω ] [ ] + [ ] [ ] = [ 𝑀𝜓 ]
(20)
0 𝐼𝑛 𝜓 𝑥 𝐶𝜓 𝜓̇ 0 𝐾𝜓 𝜓
There are two aerodynamic forces and moments acting on the propeller hub, which are formulated
as
𝑁𝐵 𝜓̇ 𝜃̇
𝐿𝑦 = 𝐾𝛼 (𝐴1′ 𝜓 − 𝑎𝐴1 + 𝐴2 )
2 Ω Ω
𝑁𝐵 𝜃̇ 𝜓̇
𝐿𝑧 = 𝐾𝛼 (𝐴1′ 𝜃 − 𝑎𝐴1 − 𝐴2 ) (21)
2 Ω Ω
𝑁𝐵 𝜓̇ 𝜃̇
𝑀𝑦 = 𝐾𝛼 𝑅 (𝐴′2 𝜓 − 𝑎𝐴2 + 𝐴3 )
2 Ω Ω
𝑁𝐵 𝜃̇ 𝜓̇
𝑀𝑧 = 𝐾𝛼 𝑅 (𝐴′2 𝜃 − 𝑎𝐴2 − 𝐴3 )
2 Ω Ω
where 𝑁𝐵 and 𝑎 are the number of blades and pivot length to rotor radius ratio.
1
𝐾𝑎 = 𝜌𝐶 𝑅 4 Ω2 (22)
2 𝑙𝛼
where 𝜌 and 𝐶𝑙𝛼 denote the density of air and blade lift slope, respectively. The 𝐴𝑖 terms are
written as
1𝑐 𝜇2
𝐴1 = ∫0 𝑑𝜂 𝐴1′ = 𝜇𝐴1
𝑅 √𝜇 2 +𝜂 2

1𝑐 𝜇𝜂 2
𝐴2 = ∫0 𝑑𝜂 𝐴′2 = 𝜇𝐴2 (23)
𝑅 √𝜇 2 +𝜂 2

1𝑐 𝜂4
𝐴3 = ∫0 𝑑𝜂 𝐴′3 = 𝜇𝐴3
𝑅 √𝜇 2 +𝜂 2
𝐽 𝑉 𝑟
where 𝜇 = 𝜋 = Ω𝑅 is the advance ratio and 𝜂 = 𝑅 . 𝑐 denotes the blade chord. The propeller
pitching and yawing moments around the pivot point can be defined such that
𝑀𝜃 = −𝑀𝑦 + 𝑎𝑅𝐿𝑧
𝑀𝜓 = 𝑀𝑧 + 𝑎𝑅𝐿𝑦 (24)
which leads to
𝑁𝐵 𝜃̇
𝑀𝜃 = 𝐾𝛼 𝑅 [−(𝐴3 + 𝑎2 𝐴1 ) − 𝐴′2 𝜓 + 𝑎𝐴1′ 𝜃]
2 Ω
𝑁𝐵 𝜓̇
𝑀𝜓 = 𝐾𝛼 𝑅 [−(𝐴3 + 𝑎2 𝐴1 ) + 𝐴′2 𝜃 + 𝑎𝐴1′ 𝜓] (25)
2 Ω

8
 IFASD-2024-129

The aerodynamic damping (𝐶𝑎𝑒𝑟𝑜 ) and aerodynamic stiffness (𝐾𝑎𝑒𝑟𝑜 ) matrices derived from the
above two aerodynamic moments about the pivot point are given as
𝑁𝐵 (𝐴3 + 𝑎2 𝐴1 )
𝐶𝑎𝑒𝑟𝑜1 = − 𝐾𝛼 𝑅
2 Ω
𝑁𝐵 𝑁𝐵
𝐾𝑎𝑒𝑟𝑜1 = 𝐾𝛼 𝑅𝑎𝐴1′ 𝐾𝑎𝑒𝑟𝑜2 = 𝐾𝛼 𝑅𝐴′2 (26)
2 2

and hence the general aerodynamic model of the rotor-nacelle dynamic system is
𝐶𝑎𝑒𝑟𝑜1 0 𝜃̇ 𝐾 −𝐾𝑎𝑒𝑟𝑜2 𝜃 𝑀𝜃
[ ] [ ̇ ] + [ 𝑎𝑒𝑟𝑜1 ] [ ] = [𝑀 ] (27)
0 𝐶𝑎𝑒𝑟𝑜1 𝜓 𝐾𝑎𝑒𝑟𝑜2 𝐾𝑎𝑒𝑟𝑜1 𝜓 𝜓

Following the determination of the structural and aerodynamic models separately, the aeroelastic
model of this dynamic system can be written as
𝐼𝑛 0 𝜃̈ 𝐶𝜃 −𝐼𝑥 Ω 𝜃̇ 𝐾𝜃 0 𝜃 𝐶𝑎𝑒𝑟𝑜1 0 𝜃̇ 𝐾 −𝐾𝑎𝑒𝑟𝑜2 𝜃
[ ] [ ] + [𝐼 Ω 𝐶𝜓 ] [𝜓̇] + [ 0 𝐾𝜓 ] [𝜓] = [ 0 ] [ ] + [ 𝑎𝑒𝑟𝑜1 ][ ] (28)
0 𝐼𝑛 𝜓̈ 𝑥 𝐶𝑎𝑒𝑟𝑜1 𝜓̇ 𝐾𝑎𝑒𝑟𝑜2 𝐾𝑎𝑒𝑟𝑜1 𝜓

The stability of the linear system is evaluated by calculating its frequency and damping through
eigenvalue analysis. For a mechanical system with damping and gyroscopic terms, solving the
eigenvalue problem is more sophisticated because these terms bring the first derivatives of the
generalized coordinates to the equations of motion. Therefore, the eigenvalue problem converts
into a quadratic eigenvalue problem, which can be reduced to a first-order state-space model.
2.3 Derivation of Aeroelastic Equations of Motion of Wing with Single Propeller
The propeller-wing model integrates the previous two models by including relevant structural and
aerodynamic coupling terms. The rotor and nacelle system that is discussed in subsection 2.2
assumes flexibly mounted engines on rigid wings. This assumption can be useful when
investigating the influences of particular parameters on the whirl flutter. However, in general, the
flexibility of the wing influences the dynamic characteristics of the entire system, which
consequently influences the whirl flutter characteristics. Therefore, the effect of the wing
flexibility is included in the equations of motion.

Lagrange’s equations for the coupled wing-propeller model are provided in Equation (29). This
model considers propeller pitch 𝜃, propeller yaw 𝜓, propeller-wing attachment point displacement
ℎ𝑝 and propeller-wing attachment point twist 𝑎𝑝 which becomes
𝑑 𝜕𝑇 𝜕𝑇 𝜕𝑈 𝜕𝐷
( ) − 𝜕𝑞 + 𝜕𝑞 + 𝜕𝑞̇ = 𝑄𝑖 𝑖 = 𝜃, 𝜓, ℎ𝑝 , 𝑎𝑝 (29)
𝑑𝑡 𝜕𝑞̇ 𝑖 𝑖 𝑖 𝑖

The equations of motion of the coupled wing-propeller model obtained through Lagrange’s
equations are given in Equation (30) and the terms in the equations of motion of the coupled model
[23] are explained in [13, 23].
𝐼𝜃,𝑝 𝜃̈ + 𝐾𝜃 𝜃 + 𝐶𝜃 𝜃̇ + 𝐼Ω Ω𝜓̇ + 𝑆𝜃,𝑝 ℎ̈𝑝 + 𝐼𝜃𝛼,𝑝 𝛼̈ 𝑝 = 𝑄𝜃,𝑝
−𝐼Ω Ω𝜃̇ + 𝐼𝜓,𝑝 𝜓̈ + 𝐾𝜓 𝜓 + 𝐶𝜓 𝜓̇ − 𝐼Ω Ω𝛼̇ 𝑝 = 𝑄𝜓,𝑝 (30)
𝑀𝑤 ℎ̈ + 𝑆𝜃,𝑝 𝜃̈ + 𝑀𝑝 ℎ̈𝑝 + 𝑆𝛼,𝑝 𝛼̈ 𝑝 + 𝑆𝛼,𝑤 𝛼̈ = 𝑄ℎ,𝑤 + 𝑄ℎ,𝑝
𝐼𝛼,𝑤 𝛼̈ + 𝑆𝛼,𝑤 ℎ̈ + 𝐼𝜃𝛼,𝑝 𝜃̈ + 𝐼Ω Ω𝜓̇ + 𝑆𝛼,𝑝 ℎ̈𝑝 + 𝐼𝛼,𝑝 𝛼̈ 𝑝 = 𝑄𝛼,𝑤 + 𝑄𝛼,𝑝

9
 IFASD-2024-129

The propeller aerodynamic moments about the pivot point are given as
𝑄𝜃,𝑃 = 𝑀𝑌 + 𝑒𝜃 𝐹𝑍
𝑄𝜓,𝑃 = 𝑀𝑍 − 𝑒𝜓 𝐹𝑌 (31)
where 𝑒𝜃 and 𝑒𝜓 are the distance between the propeller hub and elastic centres in pitch and yaw,
respectively. The aerodynamic effect of propellers on the wing is considered by applying propeller
aerodynamic loads onto the wing structural model so that
𝑄ℎ,𝑃 = 𝐹𝑍
𝑄𝛼,𝑃 = 𝑀𝑌 + 𝑒𝛼 𝐹𝑍 (32)
where 𝑒𝛼 denotes the distance between the propeller hub and the propeller pivoting point and 𝐹𝑌 ,
𝐹𝑍 , 𝑀𝑌 and 𝑀𝑍 are the aerodynamic forces and moments produced by the propeller.

1 2 ′ 𝜃̇𝑅 𝜓̇𝑅
𝐹𝑌 = 𝜌𝑉 𝑆 (𝐶𝑌𝜃 𝜃 + 𝐶𝑌𝜓 𝜓 + 𝐶𝑌𝑞 + 𝐶𝑌𝑟 )
2 𝑉 𝑉

1 2 ′ 𝜃̇𝑅 𝜓̇𝑅
𝐹𝑍 = 𝜌𝑉 𝑆 (𝐶𝑍𝜃 𝜃 + 𝐶𝑍𝜓 𝜓 + 𝐶𝑍𝑞 + 𝐶𝑍𝑟 ) (33)
2 𝑉 𝑉

2 ′
𝜃̇𝑅 𝜓̇𝑅
𝑀𝑌 = 𝜌𝑉 𝑆 𝑅 (𝐶𝑚𝜃 𝜃 + 𝐶𝑚𝜓 𝜓 + 𝐶𝑚𝑞 + 𝐶𝑚𝑟 )
𝑉 𝑉

2 ′
𝜃̇ 𝑅 𝜓̇𝑅
𝑀𝑍 = 𝜌𝑉 𝑆 𝑅 (𝐶𝑛𝜃 𝜃 + 𝐶𝑛𝜓 𝜓 + 𝐶𝑛𝑞 + 𝐶𝑛𝑟 )
𝑉 𝑉
The propeller forces and moments given above are determined employing the sixteen propeller
stability derivatives, provided in [8, 10, 11], which are applicable under windmilling conditions.
These derivatives explain how the forces and moments vary with the effective pitch and yaw
angles. Due to the interaction with the wing, these effective pitch and yaw angles for the coupled
model are specified as
𝑒𝜃 𝑒𝛼 1 𝑒𝜓
𝜃 = 𝜃 + 𝛼𝑝 + 𝜃̇ + 𝛼̇ 𝑝 + ℎ̇𝑝 𝜓 = 𝜓 + 𝜓̇ (34)
𝑉 𝑉 𝑉 𝑉
and the generalized aerodynamic forces of the coupled model can be defined as
𝑄𝜃 = 𝑄𝜃,𝑃
𝑄𝜓 = 𝑄𝜓,𝑃 (35)
𝑄ℎ = 𝑄ℎ,𝑃 + 𝑄ℎ,𝑊
𝑄𝛼 = 𝑄𝛼,𝑃 + 𝑄𝛼,𝑊
With the combination of the expressions in the obtained generalized aerodynamic forces, a coupled
aeroelastic wing-propeller model can be derived in matrix form, as provided in [13].

10
 IFASD-2024-129

3 NUMERICAL ANALYSIS
3.1 Baseline Wing
The schematic of a baseline wing from [13] is illustrated in Figure 2 with the main parameters of
the tapered wing outlined in Table 1. The elastic axis (EA) and centre of gravity (CG) are
coincident and located at the wing mid-chord.

Figure 3: Schematic of baseline wing

Table 1: Baseline wing parameters

Wing Parameter Value
Wing span [m] 11.4
Root chord [m] 1.25
Tip chord [m] 0.8
Mass per unit length [kg/m] 25
Aerodynamic centre [-] 25% chord
Elastic axis [-] 50% chord
Centre of gravity [-] 50% chord
Radius of Gyration about CG [-] 25% chord
Bending rigidity [Nm2] 7E5
Torsional rigidity [Nm2] 2E5
Lift curve slope [-] 2π
Density of air [kg∕m3] 0.96287

The structural model of the wing is based on the methodology presented in subsection 2.1. Table
2 compares the natural frequencies of the baseline wing. It is seen that there is a good agreement
between the present study and the results achieved in the reference paper [13] in terms of natural
frequencies.

11
 IFASD-2024-129

Table 2: Comparison of natural frequencies of baseline wing

Mode Present Study Reference Study [13] Deviation
[Hz] [Hz] [%]
First Bending 2.88 2.88 0
First Torsion 16.68 16.68 0
Second Bending 18.06 18.06 0
Second Torsion 46.49 46.59 0.21
Third Bending 50.58 50.57 0.04

The first four wind-off mode shapes of the wing are also given in Figure 4. Since the mass axis
and the elastic axis are coincident at the wing mid-chord axis, the wind-off mode shapes are
composed of pure bending and pure torsion without coupling.

Figure 4: Wind-off mode shapes of baseline wing

The flutter speed and frequency of the wing are predicted by the methodology provided in
subsection 2.1 and are tabulated in Table 3. The velocity-frequency and velocity-damping curves
illustrated in Figure 5 demonstrate that the first torsional mode (flutter mode) is unstable due to
interaction between the first torsion mode and the first bending mode, whereas divergence mode
is the first bending mode.

12
 IFASD-2024-129

Table 3: Comparison of aeroelastic results of baseline wing

Parameter Present Study Reference Study [13] Deviation [%]
Flutter Speed 151.4 m/s 152 m/s 0.91
Flutter Frequency 8.41 Hz 8.08 Hz 4.08

Figure 5: Variation of modal frequency and damping with freestream velocity of baseline wing

3.2 Baseline Propeller

A baseline propeller from [13] is represented by two concentrated masses: Rotor mass and motor-
nacelle mass. The main parameters of this cruise propeller used in the NASA X-57 Maxwell
electric aircraft are summarized in Table 4. It is assumed that the baseline propeller analysed in
this study have fixed-pitch blades and operate under windmilling conditions. Hence, the propeller
advance ratio remains constant and is determined based on the specified geometric collective pitch
angle. The rotational speed of the propeller changes with the freestream velocity to sustain this
advance ratio. The pitch and yaw motions of the isolated propeller are considered to be symmetric.
In the current 2-DOF propeller model, the frequency and damping coefficient for propeller pitch
and yaw motions are used for analysing the propeller’s structural dynamics. The uncoupled pitch
and yaw frequencies ensure to determine the pitch stiffness (𝐾𝜃 = 7𝐸4 𝑁𝑚𝑟𝑎𝑑 −1 ) and yaw
stiffness (𝐾𝜓 = 7𝐸4 𝑁𝑚𝑟𝑎𝑑 −1 ). Additionally, the pitch and yaw damping coefficients are utilized
to assess the structural damping of the nacelle.

13
 IFASD-2024-129

Table 4: Baseline propeller parameters

Propeller Parameter Value
Number of blades [-] 3
Rotor radius [m] 0.762
Rotor mass [kg] 8
Rotor position [m] 1.16
Motor & nacelle mass [kg] 35
Motor position [m] 0.86
Blade chord [m] 0.094
Advance ratio [-] 1.96
Pitch and yaw stiffness [Hz] 7
Pitch and yaw damping coefficient [-] 0.005

With the assumption that the baseline propeller is flexibly attached to a rigid wing, whirl flutter
analysis is conducted using Reed’s model in line with the methodology given in subsection 2.2.
The whirl flutter speed and frequency of the baseline propeller are tabulated and compared with
the results of the reference paper in Table 5.

Table 5: Comparison of aeroelastic results of baseline propeller

Parameter Present Study Reference Study [13] Deviation [%]
Whirl Flutter Speed 223 m/s 221 m/s 0.91
Whirl Flutter Frequency 4.92 Hz 5 Hz 1.63

Figure 6: Variation of modal damping ratio and frequency with freestream velocity

14
 IFASD-2024-129

The corresponding whirl flutter curves illustrated in Figure 6 indicate similar trend with the results
gained in [13]. The corresponding whirl flutter curves illustrated in Figure 6 indicate similar trend
with the results gained in [13]. The propeller’s motion is characterized by two modes: a backward
whirling mode and a forward whirling mode. Since the rigid propeller blades and rigid wings are
assumed, the forward whirling mode is stable and becomes even more stable with growing
airspeeds. Whirl flutter only develops in the backward whirling mode. The forward whirling mode
has higher frequencies compared to the backward whirling mode. Since the advance ratio remains
constant in this whirl flutter analysis, increased incoming airspeeds result in higher propeller
rotational speeds. As the propeller’s rotational speed rises, the frequency of the forward whirling
mode increases, while the frequency of the backward whirling mode falls. This relationship is due
to the difference in the sign of the gyroscopic term presented in Equation (19).

For further validation of the baseline propeller, a typical stability boundary curve was derived, as
depicted in Figure 7. The attachment stiffnesses of the propeller pivoting point in pitch (𝐾𝜃 ) and
yaw (𝐾𝜓 ) were changed and the corresponding whirl flutter stability boundary curve was achieved
for the whirl flutter speed of 223 m/s. As anticipated, the resulting curve, which shows the
necessary stiffness to maintain stable dynamic behaviour, aligns with existing literature [23].

Figure 7: Stability boundary for baseline propeller model in the structural pitch stiffness 𝐾𝜃 and
yaw stiffness 𝐾𝜓 plane for a whirl flutter speed of 223 m/s
To determine the influence of variations in propeller parameters on stability boundary in the
structural pitch stiffness 𝐾𝜃 and yaw stiffness 𝐾𝜓 plane, parametric studies on the developed whirl
flutter model of the baseline propeller are conducted. As can be seen from Figure 8, raising the
freestream velocity V, the rotor’s moment of inertia 𝐼𝑥 , the nacelle’s moment of inertia 𝐼𝑛 and the
rotor radius 𝑅 expands the unstable region symmetrically. On the other hand, increasing the
damping in both pitch and yaw (𝐶𝜃 , 𝐶𝜓 ) and the advance ratio J reduces the unstable region
symmetrically.

15
 IFASD-2024-129

Figure 8: Effect of changes in some propeller parameters on stability boundary in the structural
pitch stiffness 𝐾𝜃 and yaw stiffness 𝐾𝜓 plane
The effect of the mounting stiffness on the whirl flutter velocity and frequency of the standalone
propeller is investigated. It can be observed from Figure 9 that a higher mounting stiffness leads
to a higher whirl flutter speed and frequency. As expected, this behaviour is in line with the
literature [23].

Figure 9: Effect of mounting stiffness of propeller on whirl flutter velocity and frequency

16
 IFASD-2024-129

3.3 Coupled Wing-Propeller Model

The coupled wing-propeller model considered in this study consists of the baseline wing and
propeller described in sections 3.1 and 3.2, respectively. The schematic of this coupled model is
shown in Figure 10. The propeller is installed at a spanwise location of 31% from the wing root.
The pivoting point of the propeller is located exactly on the elastic axis of the wing (the wing mid-
chord axis). The propeller hub is situated 0.6 m from the leading edge, while the electric motor is
positioned 0.3 m behind the propeller hub.

Figure 10: Schematic of the wing-propeller model

In the first setup, the effect of adding a non-rotating baseline propeller to the wing on the instability
of the system is investigated. The non-rotating propeller model serves as a pylon and is rigidly
attached to the wing, which means that the propeller mounts are assumed to be rigid. Hence, the
added propeller changes the mass and inertial distributions of the wing. The second setup features
a rotating propeller flexibly attached to the wing, enabling pitch and yaw movements and thus
gyroscopic effects are included. Table 6 provides the natural frequencies of the rigidly-mounted
propeller wing and the flexibly-mounted propeller wing models. It suggests that the whirling
modes of the propeller are coupled with wing modes and this coupling largely depends on the
uncoupled pitch and yaw frequencies of the propeller. Moreover, the pitch frequency of the
propeller decreases as a consequence of the coupling between propeller pitch movement and wing
torsion. However, the yaw frequency stays constant since the in-plane motions of the wing were
not taken into account in this study. Additionally, it can be monitored that incorporating flexible
mounts resulted in higher natural frequencies of the wing modes compared to the application of
rigid mounts.
A rotating propeller introduces gyroscopic effects due to the presence of flexible mounts. To focus
on the gyroscopic effects, propeller aerodynamics are ignored. Propeller rotational speed affects
the propeller structural damping matrices and the propeller-wing coupling structural matrices.
Figure 11 illustrates how increasing propeller rotational speed affects the natural frequencies of
the propeller-wing system. The forward whirling mode frequency increases and the backward
whirling mode reduces with increasing propeller rotational speed. These findings emphasize the
importance of considering propeller whirling modes in the aeroelastic analysis of the DEP wings.

17
 IFASD-2024-129

Table 6: Comparison of natural frequencies of the rigidly-mounted propeller wing and the
flexibly-mounted propeller wing

Rigid Mounts Flexible Mounts

and Non-rotating [Hz] and Non-rotating [Hz]
Mode
Present Reference Study Present Reference Study
Study [13 ] Study [13 ]
First Bending 2.85 2.85 2.85 2.85
Propeller Pitch Mode - - 5.52 5.48
Propeller Yaw Mode - - 7 7
First Torsion 8.32 8.19 17.75 17.75
Second Bending 17.84 17.84 19.54 19.48
Second Torsion 26.31 26.23 49.29 49.25
Third Bending 49.98 49.98 51.34 51.37

Figure 11: Influence of gyroscopic effects on natural frequencies of propeller-wing model

18
 IFASD-2024-129

Table 7 outlines the aeroelastic results of the propeller-wing model with a rigidly attached
propeller and a flexibly attached propeller. On the one hand, in the coupled propeller-wing
configuration with a non-rotating propeller and rigid mounts, the propeller, acting as an inertial
mass, decreased the flutter velocity compared to the baseline wing. On the other hand, in the case
of the coupled propeller-wing configuration with a non-rotating propeller and rigid mounts, the
propeller increased the flutter velocity compared to the baseline wing. Also, whirl flutter occurs in
the forward whirling mode.
Table 7: Aeroelastic results of the rigidly and the flexibly-mounted propeller wings
Parameter Rigid Mounts Flexible Mounts
Wing Flutter Speed 125 m/s 202 m/s
Wing Flutter Frequency 6.40 Hz 2.79 Hz
Whirl Flutter Speed - 302 m/s
Whirl Flutter Frequency - 7.41 Hz

The velocity-damping and velocity-frequency curves of the propeller-wing model with both a
rigidly mounted propeller and a flexibly mounted propeller are depicted in Figures 12 and 13.

Figure 12: Variations of modal frequency and damping with freestream velocity for the coupled
wing-propeller model with rigid mounts

19
 IFASD-2024-129

Figure 13: Variations of modal frequency and damping with freestream velocity for the coupled
wing-propeller model with flexible mounts

3.4 Parametric Analyses

Several parametric analyses are carried out by changing the propeller’s spanwise position, advance
ratio and rotor radius, determining their impact on the stability of the coupled wing-propeller
model defined in subsection 3.3. The first parametric analysis is undertaken by attaching the
propeller to three different spanwise positions (a spanwise position of %31, a spanwise position of
%70, and a spanwise position of %100). For this analysis, the same baseline wing and propeller
are employed. The effect of each spanwise position of the propeller is evaluated on the stability of
the coupled wing-propeller model. As shown in Figure 14, installing the propeller at the wingtip
slightly affects the stability of the coupled model, decreasing the unstable region.
The second parametric analysis focuses on investigating the effect of three different advance ratios
of the propeller on the stability of the coupled model. This study uses fixed-pitch propellers, where,
under windmilling conditions, the propeller's rotational speed changes according to the incoming
airspeed to maintain the advance ratio determined by the collective pitch angle. For the baseline
isolated propeller, increasing the advance ratio has a stabilizing effect. This trend is illustrated in
Figure 8 for the isolated propeller. However, for the coupled propeller-wing model, the variation
of the advance ratio primarily affects the aerodynamics of the propeller and its gyroscopic

20
 IFASD-2024-129

coupling. Increasing the advance ratio is destabilizing. This effect is depicted in Figure 15 for the
coupled propeller-wing model.
The last parametric study is undertaken by varying the rotor radius in the coupled wing-propeller
model. On the one hand, for the baseline isolated propeller, raising the rotor radius has a
destabilizing effect, as depicted in Figure 8. On the other hand, it is observed from Figure 16 that
raising the rotor radius has a stabilizing effect on the coupled model.

Figure 14: Effect of propeller spanwise position on the stability of coupled model

Figure 15: Effect of advance ratio on the stability of coupled propeller-wing model

21
 IFASD-2024-129

Figure 16: Effect of rotor radius on stability of coupled propeller-wing model

4 CONCLUSIONS
This paper focuses on developing a representative low-order aeroelastic coupled wing-propeller
model that can be used during the preliminary design stages of DEP wings. A numerical aeroelastic
model was created in MATLAB to investigate the aeroelastic behavior of a coupled flexible
cantilever wing with a varying number of flexibly mounted propellers/rotors. Propeller dynamics
are modeled using Reed’s model, while the wing's structural model is derived using the assumed-
mode Rayleigh-Ritz method. The wing's aerodynamic model is obtained by combining the
modified strip theory and Theodorsen’s unsteady aerodynamic theory. The proposed coupled
aeroelastic model successfully predicted both wing and whirl flutter in DEP wings.
The baseline propeller used in this study involves fixed-pitch blades operating under windmilling
conditions. Therefore, the propeller advance ratio remains constant and is determined by the
specified geometric collective pitch angle. To maintain this advance ratio, the propeller's rotational
speed varies with the freestream velocity. As expected, the gyroscopic effects cause the propeller
hub to exhibit backward whirling motion at lower frequency modes and forward whirling motion
at higher frequency modes. In the coupled propeller-wing configuration with a non-rotating
propeller and rigid mounts, incorporating a non-rotating propeller decreased the critical speed in
comparison to the baseline wing. In the case of involving a rotating propeller and flexible mounts,
wing flutter increased compared to the baseline wing. The forward whirling mode led to a critical
speed. Compared to the baseline isolated propeller, the backward whirling mode remains stable.
However, the forward whirling mode, which is always stable in the baseline isolated propeller,
becomes unstable due to wing aerodynamics.

22
 IFASD-2024-129

The developed coupled model was subsequently applied onto parametric analyses identifying the
impact of propeller spanwise position, advance ratio and rotor radius on the stability of the coupled
wing-propeller system. The results of parametric studies indicated that advance ratio and rotor
radius have a significant impact on the stability of the coupled wing-propeller model, whereas the
spanwise position of the attached propeller has a slight effect. As the advance ratio of the attached
propeller rises, a destabilizing effect on the stability of the coupled model is monitored. It was also
observed that increasing rotor radius has a stabilizing effect on the stability of the coupled wing-
propeller model.
The presented study shows how rapidly the proposed coupled aeroelastic model can be modified
in various parametric studies that can be performed during the preliminary design stages of the
DEP wings. The current investigations are limited to a wing-propeller model with one inboard
propeller. Future studies should focus on a DEP aircraft wing configuration with multiple
propellers and additional parameters, such as their chordwise position and mounting stiffness.
Furthermore, the physical model should be enhanced to account for aerodynamic interference
effects between the propeller and wing aerodynamics.

REFERENCES
[1] Nguyen, N.T., Reynolds, K., Ting, E., et al. (2018). Distributed propulsion aircraft with
aeroelastic wing shaping control for improved aerodynamic efficiency. Journal of Aircraft,
55, 1122-1140.
[2] Brelje, B. J. and Martins, J. R. R. A. (2019). Electric, hybrid, and turboelectric fixed-wing
aircraft: A review of concepts, models, and design approaches. Progress in Aerospace
Sciences, 104, 1–19.
[3] Moore, J. B. and Cutright, S. (2017). Structural design exploration of an electric powered
multi-propulsor wing configuration. In 58th AIAA/ASCE/AHS/ASC Structures, Structural
Dynamics, and Materials Conference. p. 203.
[4] NASA. https://www.nasa.gov/image-detail/x-57_maxwell_city_0/, (accessed April 25,
2024).
[5] DLR. https://www.dlr.de/en/latest/news/2021/04/20211215_towards-zero-emission-
aviation, (accessed April 25, 2024).
[6] Rotundo, C., Jinks, E., and Öngüt, E., (2023). Whirl flutter analysis across multiple
configurations: Preprocessor implementation and verification. In Royal Aeronautical
Society 8th Aircraft Structural Design Conference.
[7] Reed, W.H. and Bland, S.R. (1961). An analytical treatment of aircraft propeller precession
instability. NASA Technical Note, p. 659.
[8] Houbolt, J. C. and Reed, W. H. (1962). Propeller-nacelle whirl flutter. Journal of the
Aerospace Sciences, 29(3), 333–346.
[9] Bland, S. R., and Bennett, R. M. (1963). Wind-tunnel measurement of propeller whirl-
flutter speeds and static-stability derivatives and comparison with theory. Langley Research
Center, Washington, USA Technical Note D-1807.
[10] Bennett, R. M., and Bland, S. R. (1964 ). Experimental and analytical investigation of
propeller whirl flutter of a power plant on a flexible wing. NASA Technical Note D-2399.
[11] Rodden, W. P. and Rose, T. L. (1989). Propeller/nacelle whirl flutter addition to

23
 IFASD-2024-129

MSC/NASTRAN.
[12] Reed III, W.H. (1966). Propeller-rotor whirl flutter: A state-of-the-art review. Journal of
Sound and Vibration, 4(3), 526–544.
[13] Liu Xu, V. Q. (2020). Propeller-wing whirl flutter: An analytical approach. Master's
Thesis, Delft University of Technology.
[14] Amoozgar, M., Friswell, M. I., Fazelzadeh, S. A., et al. (2021). Aeroelastic stability analysis
of electric aircraft wings with distributed electric propulsors. Aerospace, 8(4),100.
[15] Heeg, J., Stanford, B. K., Kreshock, A., et al. (2019). Whirl flutter and the development of
the NASA X-57 Maxwell. In International Forum on Aeroelasticity and Structural
Dynamics, NF1676L-31615.
[16] Böhnisch, N., Braun, C., Koschel, S., et al. (2022). Whirl flutter for distributed propulsion
systems on a flexible wing. In AIAA SCITECH 2022 Forum. p. 1755.
[17] Böhnisch, N., Braun, C., Koschel, S., et al. (2022). Dynamic aeroelasticity of wings with
distributed propulsion system featuring a large tip propeller. In International Forum on
Aeroelasticity and Structural Dynamics, Madrid, Spain.
[18] Böhnisch, N., Braun, C., Muscarello, V., et al. (2023). A sensitivity study on aeroelastic
instabilities of slender wings with a large propeller. AIAA SCITECH 2023 Forum. p. 1755.
[19] Tamer, A. and Tatar, A. (2023). A minimum complexity model for the aeroelastic analysis
of distributed propulsion aircraft. In the European Rotorcraft Forum.
[20] Wright, J. R. and Cooper, J. E. (2007). Introduction to Aircraft Aeroelasticity and Loads.
CBS Publishers & Distributers.
[21] Bielawa, R.L. (2006). Rotary Wing Structural and Aeroelasticity. American Institute of
Aeronautics & Astronautics, Washington.
[22] Mair, C., Rezgui, D., and Titurus, B. (2018). Nonlinear stability analysis of whirl flutter in
a rotor-nacelle system. Nonlinear Dynamics, 94(3), p. 2013-2032.
[23] Cecrdle, J. (2015). Whirl Flutter of Turboprop Aircraft Structures. United Kingdom:
Woodhead Publishing.

ACKNOWLEDGEMENT
This study was funded by the Ministry of National Education of the Republic of Turkey under
YLSY grant.

COPYRIGHT STATEMENT
The authors confirm that they, and/or their company or organisation, hold copyright on all of the
original material included in this paper. The authors also confirm that they have obtained
permission from the copyright holder of any third-party material included in this paper to publish
it as part of their paper. The authors confirm that they give permission, or have obtained permission
from the copyright holder of this paper, for the publication and public distribution of this paper as
part of the IFASD 2024 proceedings or as individual off-prints from the proceedings.

24