Flutter Analysis of Supersonic Low

Aspect Ratio Composite Wings

Using FSI Methodology

T. Babin and N. Sangeetha

Abstract A two-dimensional plate model for the aeroelastic flutter of low aspect
ratio composite wing is taken based on the classical plate theory and fluid–structure
interaction methodology and is adapted for further studies. The domain model and
the corresponding supersonic loads are defined in accordance with Hamilton
principle and first-order piston theories, respectively. Natural frequencies and flutter
velocity are analysed and verified using the differential quadrature method
(DQM) literatures, in comparison with the fluid–structure interaction method.
Based on the parametric study carried out, the results show that the flutter char-
acteristics are strongly dependent on cross-ply laminates and these kinds of
vibration in the low aspect ratio composite wings can occur only in the high
circumferential frequencies. The effects of the aspect ratio on supersonic flutter
characteristics of laminated wings are investigated, with constant surface area and
constant thickness, and all the layers are of equal thickness.

Keywords Flutter Low aspect ratio wings
Fluid–structure interaction


Composite wings Ply angle

1 Introduction

Fibre-reinforced composite materials are being mostly used in the design of aero-
space vehicles, since they have high stiffness, high strength-to-weight ratio and
anisotropy composite in the heat exchange. When these thin-walled structural
components are subjected to the action of a high-speed airflow on the outer surface,
they may become dynamically unstable at a certain critical speed of the vehicle.

T. Babin (&)
N. Sangeetha

Department of Mechanical Engineering, Kumaraguru College of Technology,
Coimbatore, India
e-mail: [email protected]
N. Sangeetha
e-mail: [email protected]

© Springer Nature Singapore Pte Ltd. 2019 361

K. S. Vijay Sekar et al. (eds.), Advances in Manufacturing Processes, Lecture Notes
in Mechanical Engineering, https://doi.org/10.1007/978-981-13-1724-8_35
362 T. Babin and N. Sangeetha

This phenomenon is known as flutter. The flutter behaviour of composite laminated

sheets has attracted many researchers because of its continuous growth towards the
problem. Karpouzian [1, 2] was the first to give the detailed flutter analysis on
anisotropic composite cantilevered wing structures based on the Laplace transform
method, while the Giles [3, 4] studied the flutter analysis with equivalent plate
models of planform geometry. Later on, Meirovitch [5] derived the equations
governing the static and dynamic aeroelastic displacement equations of cantilevered
wing structures based on the plate theories.
One of the proven methods to solve the flutter problem is the differential
quadrature method (DQM). Shu [6, 7] in his book gave the detailed application of
DQM in various engineering applications including compressible fluid flows and
other aeroelastic applications. And Liu and Wu [8] were the first to derive the DQM
for boundary value problems to obtain the solution. Fluid–structure interaction is a
recently developed method for analysing the effect of fluid forces on structural
components. Out of some of the ways to couple both the mechanics such as
one-way and two-way coupling strategies, two-way coupling methodology is
adapted where both the fluid and structural analysis take place at same time suc-
cessively. From the recent researches in bridge and piping vibration analysis [9–11]
made through FSI, showed the way to solve the flutter problem by the same
technique. FSI technique has shown the improved accuracy and reliability over
compressible flows. Kamakoti [12] was the first to use the FSI in aeroelastic
application; he predicted the flutter for a wing at different Mach numbers based on
moving grid technique. In addition, Kwon et al. [13] studied the impact on com-
posite plates with FSI by conducting a real-time experiment and showed the
application of FSI on composite plates. This paper is majorly concerned with the
study on aeroelastic flutter characteristics of low aspect ratio composite wings at
compressible flow condition using FSI technique. The results are compared with
DQM method based on the literatures [6, 7]. At last, a detailed parametric study has
been conducted to study the influences of aspect ratios, ply angle and stacking
sequence on the supersonic flutter behaviour of composite laminate panel.

2 Theoretical Formulations

2.1 Low Aspect Ratio Wing Model

For simplifying the problem, 2D shear deformable plate model is chosen for the
study. Wing structure is idealized as a fibre-reinforced laminated composite plate
whose constituent laminates are located by different angles in cross-ply and
stacking sequence with constant surface area and constant layer thickness.
Two-dimensional plate model is developed such that x and y are the coordinates
representing the chord-wise and span-wise coordinates, respectively. The point O is
the origin located at the intersection of the mid-plane chord with the leading edge of
Flutter Analysis of Supersonic Low Aspect Ratio … 363

Fig. 1 2D plate model with

displacement

wing structure. The 2D plate model representing low aspect ratio composite wing is
assumed to act as an elastic plate which is clamped at the root chord and set to free
at other three boundaries as shown in Fig. 1.
According to Kirchoff’s hypothesis, the displacements (u, v, w) are given by [2]:

@w
uðx; y; z; tÞ ¼ u0 ðx; y; tÞz ;
@x
@w ð1Þ
vðx; y; z; tÞ ¼ v0 ðx; y; tÞ  z ;
@y
wðx; y; z; tÞ ¼ wðx; y; tÞ:

Since the displacement along w(x, y, z, t) is independent of transverse coordinate

z and assumption made that root chord is fixed, the transverse normal and shear
strains due to plunge and pitch are zero. u0 and v0 are the mid-plane elastic
deflections in the x and y directions, respectively. Partial derivatives refer to the
corresponding angular displacements.

2.2 Supersonic Aerodynamic Load

In the compressible supersonic flight region, the aerodynamic load is unsteady in

nature which can be evaluated by the first-order theory. These aerodynamic loads
on the faces of the composite wing can be expressed by:

 

2qV @w @w @H 4q @w 1 @w @H
DF ¼ V þ þV ¼ þ þ :
Ma @x @t @x Ma @x V @t @x
2
The dynamic pressure is given by q ¼ qV2 .
364 T. Babin and N. Sangeetha

Where q is the air density, Ma is the Mach number, V is the velocity of the wing,
w(x, y, t) is the transverse displacement of the wing, q is the dynamic pressure, and
H(x, y) is the wing thickness function.

2.3 Energy Expression as Governing System

The dynamical problem of flutter can be formulated as strain energy expression of

plate segment, and it can be expressed as [3]:

Zt1
DE ¼ ðdP  dT  dW Þ:dt ¼ 0 ð3Þ
t0

where P is the potential energy, T is the kinetic energy, and the W is the virtual
work performed by the wing due to non-conservative forces. Potential energy of the
wing is the total strain energy stored in the wing along the respective directions.

@u @u0 @2w @2w @v @v0 @2w @2w

ex ¼ ¼  z 2 ¼ z 2 ; ey ¼ ¼  z 2 ¼ z 2 ;
@x @x @x @x @y @y @y @y
ð4Þ
@u @v @u0 @v0 @ w
2
@ w
2
cxy ¼ þ ¼ þ  2z ¼ 2z :
@y @x @y @x @x:@y @x:@y

In order to get the simple form, the mid-plane elastic deflections are assumed
zero, such that the partial derivatives containing u0 and v0 were eliminated. ex and ey
are the normal strains in their respective directions, and cxy is the engineering shear
strain.
For the nth general orthotropic layer of composite wing with principle axes, the
constitutive equation of stresses takes the form,
8 n 9 2 ! ! ! 38 9
< rx = Q11 Q12 Q16 < enx =
6 ! ! ! 7
rny ¼ 4Q Q22
n
Q26 5: ey ; ð5Þ
: sn ; 12
! ! ! c n
xy Q16 Q26 Q66 xy

The values of the elements [Q]n in the above-discussed matrix are related to
material properties of the nth layer by [6],

E1 #12 E2 E2
Q11 ¼    ; Q12 ¼    ; Q22 ¼    ;
1  #212 EE21 1  #212 EE21 1  #212 EE21 ð6Þ
Q66 ¼ G12 :
Flutter Analysis of Supersonic Low Aspect Ratio … 365

where mij, Ei and Gij are Poisson’s ratio, Young’s modulus and shear modulus,
respectively. The relationship between local material expressions discussed above
and the global ones can be related as follows:
8 ! 9 2 3
>
> Q11 >> m4 n4 2m2 n2 4m2 n2
> ! >
>
> Q22 >
> >
> 6 n4 m4 2m2 n2 4m2 n2 78 Q11 9
< ! >
> = 6 7>< >
=
Q12 6 m2 n2 m2 n2 m 4 þ n4 4m2 n2 7
! 6
¼6 2 2 7 Q22
ð7Þ
ðm2  n2 Þ 7
2
>
> Q66 >> 6m n m 2 n2 2m2 n2 7>: Q12 >;
>
> >
! > 4 m3 n mn3 mn3  m3 n 2ðmn3  m3 nÞ 5 Q66
>
> Q16 >>
>
: ! >
;
Q26 mn3 m3 n m3 n  mn3 2ðm3 n  mn3 Þ

where m = cosh and n = sinh.

It is assumed that the laminate consists of 2k symmetrical layers of constant
thickness and it is a continuous function of x and y. The laminate strain energy for a
single layer n is given by:
Ztn n o
Pn ¼ rnx enx þ rny eny þ rnxy enxy :dz ð8Þ
tn1

The laminate strain energy for all the layers is given by integrating the above
equation over the domain of the plate extended in x-direction as wing chord and in
y-direction as wingspan.
8 t 9
Za Zb Xk < Zn h i =
dP ¼ 2 rnx enx þ rny eny þ rnxy enxy :dz ð9Þ
n¼1
: ;
0 0 tn1

where a is the wing chord breadth and b is the wingspan length.

To simplify the above strain energy expression, the material properties along
with thickness are related to a single equation of total laminate bending stiffness
coefficient Dmn which is defined as:

2X k
!
Dmn ¼ t3  tn1
3
Qnij ð10Þ
3 n¼1 n

The non-conservative virtual work is mainly due to fluid forces on the face of the
wing. The unsteady aerodynamic load multiplied by corresponding virtual dis-
placement yields non-conservative virtual work density. By integrating it over the
entire domain of the wing. The obtained non-conservative virtual work is given by:

Za Zb Za Zb

4q @w 1 @w @H
dW ¼ DF:dw:dy:dx ¼ þ þ dw:dy:dx: ð11Þ
Ma @x V @t @x
0 0 0 0
366 T. Babin and N. Sangeetha

Also, the kinetic energy of the wing takes the following equation:

Za Zb
dT ¼ qh€
wdw:dy:dx: ð12Þ
0 0

By substituting (9), (11) and (12) in the extended Hamilton’s principle Eq. (3),
the governing system equation based on the energy conservation can be obtained as:

D11 wxxxx þ 4D16 wxxxy þ 2ðD12 þ 2D66 Þwxxyy þ 4D26 wxyyy þ D22 wyyyy þ qh€
w


4q @w 1 @w @H ð13Þ
þ þ þ ¼0
Ma @x V @t @x

For the flutter problem, the boundary condition of the low aspect ratio composite
wing is:
(a) Fixed support at the root:

@w
wðx; y; tÞ ¼ 0; ¼ 0;
@y

(b) Free at the other ends:

Mn ¼ 0; Pn ¼ 0;

Here,

@Ms  
Pn ¼ Qn þ ; where Ms ¼ nx ny My  Mx þ n2x  n2y Mxy ;
@s
Mn ¼ n2x Mx þ 2ny Mxy;y nx þ n2y My ;

where the subscripts n and the s represent the normal and the tangential directions
of the edge, respectively. And the Mn and the Ms represent the normal bending
moment and twisting moment, respectively. Also, nx and ny denote the direction
cosines of unit normal vector at a point.

3 FSI for Flutter Analysis

3.1 Fluid Solver

A semi-spherical domain of radius 5 m is chosen as fluid zone, and the origin is

located at the intersection of trailing edge with root chord. Hexagonal block mesh is
created using ICEM CFD with 445,864 elements and 82,398 nodes. Fluent 18.2 is
used as fluid solver with density-based Navier–Stokes equation and k–e turbulence
Flutter Analysis of Supersonic Low Aspect Ratio … 367

model. The pressure far field is the condition given to outer walls in an assumption
that the wing is a cantilevered structure at open atmosphere. Implicit scheme and
second-order upwind scheme are used to solve the governing equations.

3.2 Structural Solver

Finite element method is used to solve the structural equations. ACP (pre) from
ANSYS 18.2 models the composite laminates. Figures 2 and 3 show the finite
element model and the laminate model of the composite wing, respectively. The
geometrical specifications and material properties are given in Table 1. Closed
coupling of fluid forces into structural faces results in mechanical deformations,
which is further directed to modal analysis to obtain the excitation frequencies,
so-called flutter frequencies.

4 Results and Discussion

4.1 Verification with DQM

In this section, the aeroelastic flutter characteristics of the cross-ply [h/−h/h/−h]s

laminated low aspect ratio composite wing are analysed using FSI method. The

Fig. 2 Finite element model

of the composite wing
368 T. Babin and N. Sangeetha

Fig. 3 Second ply of the stacking sequence—45° oriented ply

Table 1 Baseline parameters Parameters Value

of composite wing
a 0.4 m
b 1m
t 0.002 m
q 1490 kg/m3
Ex 4  1010 N/m2
Ey 8.6  109 N/m2
Poisson ratio XY 0.25
Strain Z 0.0032

comparison between the excited frequencies of FSI method and the DQM method is
discussed in Table 2, and the graph shows the deviation between the two methods.
The convergence of the FSI and DQM results is represented in the graph form in
Fig. 4. The values plotted in each case are the excited frequencies obtained at speed
of 600 m/s. For several structural models and the flutter velocity with variation in
ply angle, the listed table and the graph show that the results obtained using fluid–
structure interaction method have good solution in comparison with the results
obtained by DQM [9]. These results show that the flutter analysis can be carried out
further for parametric analysis such as stacking sequence and other aspect ratios.

4.2 Aeroelastic Characteristics

Aeroelastic characteristics of composite wings are highly dependent upon the

structural parameters such as aspect ratios, ply angle and stacking sequence. Based
on the dynamic stability, the flutter coalescence occurs at critical dynamic pressure
Table 2 Excited frequencies of the cross-ply [h/−h/h/−h]s laminated wings at Mach number 1.8
h FSI DQM
15 30 45 60 15 30 45 60
Frequency (Hz) Mode 1 5.435 8.81 16.318 23.8 6.012 9.132 16.32 25.33
Mode 2 31.42 50.412 62.845 58.989 32.20 51.61 62.86 61.01
Flutter Analysis of Supersonic Low Aspect Ratio …

Mode 3 35.62 52.189 92.762 114.65 37.42 54.04 92.45 117.8

Mode 4 104.78 153.24 160.15 165.42 101.6 152.8 161.0 162.1
369
370 T. Babin and N. Sangeetha

Fig. 4 Excited frequencies of composite wing versus ply angle

Table 3 Coalescence of frequencies for various stacking sequence

Stacking sequence [−45/45/90/0] [−45/90/0/45] [45/90/−45/0] [90/−45/45/0]
Flutter velocity (m/s) 986 826 916 687

Table 4 Flutter velocities of composite panel at different ply angles with different aspect ratios
Flutter velocity [15/−15/15/−15] [30/−30/30/−30] [45/−45/45−45] [60/−60/60−60]
(m/s)
k = 1.25 615 765 859 958
k = 1.50 584 721 784 809
k = 1.75 569 648 704 749

by frequency coincidence theory. However, flutter characteristics of supersonic low

aspect ratio wings can occur with higher circumferential frequencies. Hence, the
frequency coincidence should be assured to accurately describe certain higher
frequencies for flutter prediction of cantilevered wings using fluid–structure inter-
action method. Frequency coalescence is calculated based on the coincidence of
bending and torsion mode at different flight speeds. The excited frequency coa-
lescence of frequencies for different stacking sequences is discussed in Table 3.
Also, the flutter velocity of the composite panel with varied aspect ratios and ply
angle was analysed and the results are plotted for discussion in Table 4.
The influence of aspect ratios on supersonic aeroelastic characteristics is
investigated, with constant thickness, surface area and unit layer thickness. From
the results, it is inferred that the flutter characteristics are highly dependent on the
cross-ply angle and the aspect ratio. When aspect ratio is greater than one, the
corresponding flutter velocity also increased with ply angle.
Flutter Analysis of Supersonic Low Aspect Ratio … 371

5 Conclusion

A two-dimensional plate model of low aspect ratio composite wing is developed for
aeroelastic flutter analysis from the extended Hamilton principle. And a solution of
flutter analysis is given based on the fluid–structure interaction method. From the
study, the effects of aeroelastic characteristics such as cross-ply, aspect ratio and
stacking sequence on the supersonic aerodynamic load condition were investigated.
The following conclusions can be made from the results obtained:
(a) Fluid–structure interaction method provides a better solution for the flutter and
aeroelastic characteristics of low aspect ratio composite wings, in comparison
with DQM. Also, the computational time for this analysis is comparatively less
in case of two-dimensional models.
(b) Supersonic aeroelastic characteristics of low aspect ratio composite wings can
occur with very high circumferential frequencies. This proves that the plate
model is better than the beam model in the case of low aspect ratio wings,
because in case of beam model the aerodynamic flutter occurs by the coales-
cence between the bending mode and the torsional mode.
(c) Comparative analysis of flutter characteristics at different aspect ratio, cross-ply
and the stacking sequence shows the influence of such parameters over the
flutter speed. It is to be noted that the flutter characteristics of the composite
laminates vary complicated with the ply angle and aspect ratios.
Since the problem is simplified by CPT into a 2D model, the computational time
taken by the commercial software is comparatively less. However, the computa-
tional power required to couple the two solvers is more to perform simulations.

References

1. Karpouzian, G., Liberscu, L.: Exact flutter solution of advanced anisotropic composite
cantilevered structure. AIAA-93-1535 (1993)
2. Karpouzian, G., Liberscu, L.: Comprehensive model of anisotropic composite aircraft wings
suitable for aeroelastic analysis. J. Aircraft. 31, 703–712 (1994)
3. Giles, G.L.: Equivalent plate analysis of aircraft wingbox structures with general planform
geometry. J. Aircraft. 23, 859–864 (1986)
4. Giles, G.L.: Further generalization of an equivalent plate representation for aircraft structural
analysis. J. Aircraft. 26, 67–74 (1989)
5. Meirovitch, L., Seitz, T.J.: Structural modeling of low-aspect ratio composite wings.
AIAA-93-1371 (1995)
6. Shu, C., Richards, B.E.: Application of generalized differential quadrature to solve two
dimensional incompressible Navier-Stokes equations. Int. J. Numer. Methods Fluids 15, 791–
798 (1992)
7. Shu, C.: Differential Quadrature and Its Application in Engineering, pp. 206–224. Springer,
London (2000)
8. Liu, G.R., Wu, T.Y.: Multipoint boundary value problems by differential quadrature method.
J. Math. Comput. Model. 35, 215–227 (2002)
372 T. Babin and N. Sangeetha

9. Hua, X.G., Chen, Z.Q., Ni, Y.Q., Ko, J.M.: Flutter analysis of long-span bridges using
ANSYS. J. Wind Struct. 10, 61–82 (2007)
10. Tateishi, A., Watanabe, T., Himeno, T., Inoue, C.: Multimode flutter analysis of transonic fan
using FSI simulation. In: Proceedings of ASME Turbo Expo-26702 (2014)
11. Parameshwaran, R., Jathin, S., Rao, D.: Fluid-structure interactions and flow induced
vibrations: a review. Proc. Eng. 144, 1286–1293 (2016)
12. Kamakoti, R., Shyy, W.: Fluid structure interaction studies for aeroelastic application.
J. Progress Aerosp. Sci. 40, 535–558 (2004)
13. Kwon, Y.W., Owens, A.C., Kwon, A.S., Didoszak, J.M.: Experimental study of impact on
composite plates with fluid structure interaction. Int. J. Multiphys. 4, 259–271 (2010)