Aeroelasticity

13 January 2016

LECTURE L02. 2D AEROELASTICITY

1- Explain the physical concept of divergence (with your own words)
Physical phenomena in which, for a specific conditions, the elastic restoring forces acting
on the wing are lower than the aerodynamic forces, producing a high increase of the twist
angle and a high increase in lift that can yield to the wing break.
Divergence occurs when a lifting surface deflects under aerodynamic load so as to increase
the applied load, or move the load so that the twisting effect on the structure is increased.
The increased load deflects the structure further, which eventually bring the structure to
the divergence point.

2- 2D airfoil divergence. (Equations. From the equilibrium equation

of a 2D airfoil forces deduce the expression for qdiv)

Terms that factorize to the left, others to the right. || There is a mathematical expression
for the equilibrium, showing the twist angle in which the elastic restoring forces equal the
steady aerodynamic forces. The dynamic pressure that makes denominator in twist angle eq.
equal to 0 is qdiv. Then:

(0 +
)
1 /

3- Is the divergence speed dependent on the mass?

Since the mass is independent of in the equilibrium equation, the denominator will
be independent of the mass so Qdiv will be independent on the mass. From this
statement can be deduced that the effect of mass modifies the elastic twist behavior
but not Qdiv. It can be appreciated in equations of next question.

4- From the equilibrium equation of a 2D airfoil forces, deduce the

expression for qdiv when Nz 0.

Same equations as in question 2 but adding NzmgXcg.

5- Compressibility effects on divergence dynamic pressure. Introduce
the concepts of unmatched solution and matched solution.
Unmatched solution does not take into account the standard ISA. For a constant divergence
dynamic pressure, the relation with V2 is constant. In order to add the compressibility
effects, a correction factor qdiv=qidv|inc1 2 is introduced. Unmatched solution is
obtained and can be plotted into a Mach & TAS vs Altitude plot (Page 27) in which it can
be appreciated that for higher Mach, higher is the difference between the incompressible and
compressible Qdiv.

For the case of the matched solution, the standard ISA is introduced as the altitude affects
the value of the density as well as the value of speed of sound.

So once obtained the Mach for different altitudes, a divergence envelope can be obtained for
the Matched Solution (to plot in the Mach & TAS vs Altitude plot). This divergence
envelope should be at least a 15% outside of the aircraft flight envelope (1.15VD/MD).

6- 2D Quasi-steady aerodynamics (heave/pitch). Why is it not

suitable for aeroelasticity?
Having an airfoil undergoing a general motion in heave and pitch and assuming:

In this simplified approach, the changes in the wake due to the airfoil motion are not
accounted, hence, 2D quasi-steady aerodynamics is not suitable for aeroelasticity.

7- Unsteady aerodynamics. Response to a sudden change of angle of

attack. Description of Wagner function. Physical insight.
In a 2D quasi-steady aerodynamic model if it is subjected to an instantaneous change in
angle of attack of = /2 the lift would increase instantaneously by 50%, but this does not
occur in practice.

L = Lsteady () where () is the Wagners function.

Unsteady lift changes instantaneously by a 25% increase and then increases asymptotically
towards the final steady value. This asymptotic behavior occurs due to the change of the
vortex strength, releasing the starting vortex rearwards.
By definition, Wagners function is used to model how the lift acting at the quarter chord
on the airfoil builds up following the step change of incidence.

8- Unsteady aerodynamics. Description of Kssner function.

In a 2D quasi-steady aerodynamic model if suddenly encounters a sharp-edged gust of
velocity wg its lift would increase instantaneously due to the increase in angle of attack, but
this does not occur in practice.

L = Lsteady () where () is the Kssners

function which describes how the aerodynamic forces build up upon entering a step gust (lift
increases proportionally as the gust enters to the airfoil without sudden increase of lift).

9- Why Wagner and Kssner functions are different between?

(Especially for small distance of travelled semi-chords). Elaborate
qualitatively the answer.
Wagners function represents the build of lift due to sudden increase of AoA. The entire
airfoil exhibits immediately the increase of AoA. The vortex released from the TE is
responsible that a =0 the increase of lift is only 0.5 and subsequently, for large travelled
semi-chords, the increase of lift tends asymptotically to 1.0 once the released vortex is far
downstream.

Figure 1: Wagners function (Left) & Kssners function (Right) lift increase.

Kssners function represent the build of lift due to airfoil penetration in a step gust. The
airfoil is entering the gust progressively. Just after =0 only a very small portion of the
airfoil is inside the gust, therefore, there is no significant increase of lift yet. This is why
the Kssner function value starts at 0.0. When the entire airfoil is inside the step gust, the
Kssner function value is 0.55 (A similar value 0.55 is obtained with the Wagner function
with 2 semi-chords travelled (2c/2).)

10- Unsteady aerodynamics. Description of Theodorsen function.

Theodorsens function is used to model the changes in amplitude and phase of the sinusoidal
unsteady aerodynamic forces relative to quasi-steady forces for different reduced frequencies.
It behaves as the Fourier transform of Wagners function.
The amplitude and phase of Theodorsens function shows that for an oscillating airfoil at
different reduced frequencies (k=c/V) the amplitude decreases and tends to 0.5 while the
phase lag increases up to a value around k=0.3 and then reduces again.

11- Unsteady aerodynamics. Description of Sears function.

In a 2D quasi-steady aerodynamic model if a sinusoidal gust field having a vertical velocity
as a function of time is encountered, the lift would increase instantaneously due to the
increase of AoA, but this does not occur in practice. L = Lsteady () where () is the
Sears function which is defined as a (approx.) combination of Theodorsen and Bessel
functions.
Amplitude and phase of Sears function shows that frequency decreases to 0 a k increases
and the phase lag increases only for very low k and then reduces significantly.
THEODORSENS AND SEARSS ARE VERY SIMILAR FUNCTIONS FOR SMALL k.

12- Definition (and physical concept) of reduced frequency k

Reduced frequency is interpreted as the relationship between the time for a particle to fly
over the airfoil and the time that the particle oscillates in the airfoil.
The lines of k are straight lines in the frequency-speed diagram. Amplitude attenuation and
phase lag are functions of k.

LECTURE L03. 2D AND 3D STATIC AEROELASTICITY

1. Evolution of first control surface normal mode (frequency and

mode shape) with control-chain stiffness

2. Explain the physical concept of Control Reversal (with your own

words)
Briefly, control reversal is an effect happening in aircraft that makes controls to
reverse, such that for the pilot to carry out a maneuver, it has to order the
aircraft to do the opposite maneuver, since the controls are reversed.
Physically, when dynamic pressure increases, control surface effectiveness falls up
to the point in which 0 effectiveness is reached, that is, the point at which control
reversal occurs and lift vanishes (the flexibility considering lift, ).
As wing has greater flexibility than ailerons, rudder and elevator, reversal will
occur firstly in these surfaces, and particularly in ailerons, the surface with lowest
reversal speed, and consequently the limiting element.

3. 2D control effectiveness equations

EQUATIONS OF SLIDE 12 TO 20

4. Deduce the expression for the twist induced by an aileron

deflection in a 2D airfoil with torsion stiffness
EQUATIONS OF SLIDE 13 & 14

5. Deduce the expression for the flexible lift due to an aileron

deflection in a 2D airfoil with torsion stiffness
EQUATIONS OF SLIDE 13 & 14

6. Deduce the expression for the reversal dynamic pressure

EQUATIONS OF SLIDES 13 TO 19

7. In which conditions Lflex/Lrigid is reduced linearly with dynamic

pressure?
Since;

Then for R<1 Lf>Lrig (rare case). Only obtained with forward sweep
For R>1 Lf<Lrig (normal case). Only obtained with straight or aft sweep.
When R>>1 the expression becomes:

This condition follows for the case of aft swept wings, with a relatively large
swept angle.
The condition of R=1 is extremely unlikely.

8. Why qreversal is not dependent on e?

Control surface reversal happens because there are two effect that compensate
each other. On the one hand, the increase of lift due to control surface deflection
(that is proportional to e) and on the other hand the decrease of lift produced by
negative twist (that is also proportional to e). When equaling these effects, e is
eliminated from the equations.
Mathematical prove:

Recalling the qdiv expression,

Equating both it yields that:

As expected, it is not dependent on e.

9. 3D wing divergence using the slender straight wing approach

EQUATIONS SLIDES 28 TO 36

10.

Qualitative description of the effect of wing sweep angle on

wing divergence speed

Note that for swept angle positive is for swept back and negative for swept
forward.

As the swept angle is increased, divergence dynamic pressure increases and since
dynamic pressure is proportional to V, it can be concluded that for swept forward
wing velocities have to be much lower than for swept back wings.

11.

Describe coupling between bending and torsion in a swept

back wing
It is clear from figure below that for a swept back, wing bending results in a
positive bending (counterclockwise), which in turn produces a negative twisting in
the wing (see red arrow in rightmost image) that increases the negative twist (see
orange arrow).

Therefore, swept back produces a bending moment that increases the negative
twisting of the airfoil (pitch down)

12.

Effect of wing sweep on the evolution of the frequency of

the wing bending mode with flight speed

It is fairly obvious from picture above that for a wing with no sweep, the
frequency of the bending mode is independent of the speed. On the other hand,
swept-back wings increase their their natural frequency with velocity, while
swept-forward wings behave in the other way round.

LECTURE 04. THE STRUCTURAL MODEL AND THE

NORMAL MODES
1. Describe the building of the structural model (for aeroelastics).
A FEM model is created by using elements related with geometry (coordinate system, grid
points), elements (rods, bars, beams, CQUAD4, shells), properties and material
characteristics, CONM2 for lumped masses, RBE3 to connect masses or apply forces, in order
to represent the stiffness properties of the structure and the load path of the structure that
predicts accurately the displacements at the tip of the lifting surfaces (for aeroelasticians) but
many stress engineers are more concerned about the displacements at the root so this may
have consequences.
The best model is a balanced model in complexity and fidelity to physics in which answer the
key questions. Any FE model suitable for aeroelastic analyses should be able to represent
adequately the normal modes up to 50Hz.

2. The stiffness model (and way of validation).

Stiffness model is created to represent the stiffness of the aircraft. Stiffness model is validated
through static tests performed in the aircraft.

3. Properties of the stiffness matrix [K]

Stiffness matrix is real (all components do not contain imaginary numbers), symmetric with
respect to the diagonal and positive definite (all eigenvalues are non-negative). The stiffness
matrix is used with the mass matrix to get the natural frequencies of the system (one for each
DoF) with an eigenvalue problem.

4. Describe the natural mesh approach for FE modeling of structures.

It is the most suitable mesh for structural dynamics and aeroelasticity. Finer meshes are
undesirable because of its cost. It creates a mesh in which their elements are clearly defined
by the components of the aircraft. As an example: For the wings, in the skin there is only 1
element between ribs and stringers, in the spars only 1 element between ribs and upper &
lower skin; for the fuselage skin, there is 1 element between frames and stringers

5. Which are the structural details important for aeroelastics?

The aeroelasticians would have a model that accurately predicts the displacements at the tip
of the lifting surfaces (surface with more aeroelastic behavior). Need of especial attention to
the details (leading edges, pylons, control surface fittings, fairings). For example, in the

control surface fittings, the aeroelasticians need out-of-plane bending of these fittings to avoid
spurious control surface lateral modes while stress engineers work with in-plane only.

6. The building of the mass model (lumped masses; types of aircraft

masses; etc.)
Masses are treated in a matrix form with the same characteristics of the stiffness matrix but
this is not suitable for a complete aircraft, then lumped masses are used. Lumped masses
represents strips/slices of the aircraft splitting structure fixed part from movable part. Once
the aircraft is modelled as lumped masses, it can give mass data like total mass, center of
gravity, moments of inertia and reference grid point. In MSC NASTRAN, designed with
CONM2.
The aircraft mass is split in:

Each lumped mass should be distributed to several adjacent

hard points, this require a specific FEM element: RBE3 in
NASTRAN.
The RBE3 allows connecting masses to the structure without
adding undesirable extra stiffness. The RBE3 allocates the
mass to structural hard points.

7. Conceptual difference between consistent and lumped mass approach

Consistent masses are treated in a matrix form with the same characteristics of the stiffness
matrix. This is very accurate for a single beam but it is not suitable for a complete aircraft.
Instead of using these for an aircraft, it is used the lumped masses which represent
strips/slices of the aircraft and are simplified in some specific grid points.

8. What has to be considered with payload masses (how should be

connected to the structure, considerations)
Payload masses have to be considered as lumped masses. A
generalized grid in the center of the fuselage is connected
with RBE3 to the structural grids in the lower part of the
fuselage representing the seat, passengers and cargo.
Payload is typically less relevant for stability (flutter)
problems but it is very important in response (dynamic
loads) problems.

9. What has to be considered with fuel masses

Within fuel masses, the fueling sequence and the cold and hot temperature must be
considered.

10. How are the lumped masses connected to the structure? (Use as an
example the NASTRAN element RBE3)
RBE3 allows connecting masses to the structure without adding undesirable extra stiffness. It
connects a reference point with the mass allocated to it with the rest of the structure
considered as hard points through many legs.

11. Properties of the mass matrix [M]

It is real, symmetric, positive definite and banded.

12. The mass model is obtained using densities in the FE model?

(YES/NO)
No, the FEM does not represent the volume of the structure, so there is no possibility to
obtain the mass distribution by multiplying by the densities.

13. Why the dynamic model in the g-set has to be reduced?

CPU of Normal modes computation may be affordable for a few runs, but it may be
problematic for thousands of runs. The CPU cost of running thousands of aeroelastic and
dynamic load cases using complete aircraft FE models is not manageable even with today
computers. Therefore, dynamic model in the G-set has to be reduced.
Dynamic FE models size continues growing with time thanks to the improvement of the
computers that follow the Moores law: Microprocessors double in power every 2 years.

14.

Dynamic model reduction: the Guyan theory.

The idea is to make a partitioning of the dynamic equation between A-set and O-set. All grid
points with mass or applied forces (note that points with mass may be interpreted as points
where inertia forces will be applied) and some more needed will constitute the Analysis set =
A-set. Most of the grid points with no applied forces: Omitted set: O-set.

15. Explain qualitatively the super-element technique

The super-element technique assembles together all the reductions of the aircraft components
to get a complete aircraft dynamic model. Super-element technique is an
efficient/simple/versatile method for dynamic models set-up.

16. Equation to be solved to obtain the structure normal modes

Where K is the stiffness matrix and M the mass matrix. NASTRAN uses LANCZOS.

17. Explain a given mode shape plot

Here, one has to study a mode shape plot by taking into account that: red color and green
color accounts for opposite directions. Blue lines accounts for static lines that do not move.
Blue lines are a key point when analyzing the mode shape plot
This is the first symmetric wing
bending mode. The wing tips are
bending symmetrically.
Outer engines have a pitching motion
due to the bending of the wing.
Maximum displacement seems to be
in the Z direction for the wing tips.

This is the first anti-symmetrical

wing bending mode,
Maximum displacement seems to be
in the Z direction for the wing tips in
an anti-symmetrical mode.
HTP have anti-symmetrical bending
as well as the wing.

This is the anti-symmetric O/B engine

pitch mode.
There is anti-symmetric bending in the
wings and also in the HTP.
The HTP as well is yawing.
Outer engines have a pitch motion due to
the bending of the wings.
Maximum displacement seems to be in the
X direction due to the HTP antisymmetrical yaw.

18. The order of magnitude of aircraft normal mode frequencies

They vary from 7.0 Hz for fighters to 0.7 Hz for very large airliners (effect of mass larger than
the stiffness) for the 1st Symmetric Wing Bending Frequency and from N/A for fighters to
4.2 Hz for very large airliners for the 1st Symmetric HTP Bending Frequency.

19. A dynamic model consisting in a clamped beam with 8 grid points

has 3 lumped masses. What is the expected number of modes of
these models?
8 grids have 6 degrees of freedom each, so 48 modes are possible. Since there are only 3
lumped masses, there are only 3x6 modes. 18 MODES.

20. Mathematical properties of normal mode shapes (eigenvectors) []

There are eigenvectors as many natural frequencies are. They are orthogonal with respect to
the mass and stiffness matrix.

The wing have a torsional moment

and also some bending at the tip.
The wing as well has a yawing moment.
The HTP has anti-symmetrical bending
moment as well as a small yawing
moment.
The fuselage is having anti-symmetrical
torsion.
It can be said that we are dealing with a
chordwise torsional antisymmetric mode.

This is the first anti-symmetrical wing

bending mode,
Maximum displacement seems to be in
the Z direction for the wing tips in an
anti-symmetrical mode.
HTP have anti-symmetrical bending as
well as the wing.

LECTURE 05.GROUND VIBRATION TEST (GVT) &

EXPERIMENTAL MODAL ANALYSIS (EMA). DYNAMIC
MODEL VALIDATION
1. Dynamic model validation flowchart

2. The Ground Vibration Test (GVT). Definition and objectives.

Aspects to consider (specimen, bc, excitation, response
measurements, scaffolding).
The GVT is a non-destructive structural test performed on-ground on a complete aircraft
ready for flight with the aim to obtain experimentally the normal modes of the aircraft.
During the GVT the aircraft is dynamically excited in a controlled way in one or more
points and the response of the aircraft is measured. Its objectives are devoted to the

experimental modal analysis (EMA) of the aircraft in frequency range where potentially
dynamic amplifications or destructive aeroelastic instabilities can take place. The results
obtained are natural frequencies, modal shapes, damping and masses. These results are
used to update the dynamic FEM.
In order to perform a GVT, some aspects must be considered:
o

Specimen and configurations: Aircraft in ready for flight condition must be

tested in different configurations. (i.e. OEW, flaps extended, feed tanks filled)

BC: Free-free conditions through the use of bungees (preferred option), pneumatic
platforms (add some uncertainty to damping), deflated tires (not recommended for
new designs) or jacks (bad results because not represent free-free condition).

Scaffolding: To access the aircraft for installation or instrumentation and place

structural exciters (need of stiffer scaffoldings to minimize perturbation from
platform resonances).

Controlled excitation: Structure has to be excited through Operation Modal

Analysis, using control surfaces (FVT), by hand, instrumented impact hammer,
shakers or hydraulic exciters. Applied energy must be in all axes, high enough to
avoid noise threshold and low enough to avoid damaging the structure.

Transducers (Accelerometers and cabling): Accelerometers must be adequate in

terms of frequency range and sensitivity (location known from FEM). Cables are
deployed through structure to connect accelerometers. Special attention to
minimize added weight and allow load without restraints in the structural
displacement under load. Wiring is time consuming (need of verification in
connections and alignments) [FOR A400M 690 ACCELEROMETERS USED]

Data acquisition and test performance: Check 5 critical points:

o

Rigid Body motion and mode shapes. (Largest RBM frequency should be 3
times lower than first flexible mode)

Correct accelerometers identification and orientation using RBM

Verification od excitations level and direct FRFs

Coherence verification (verify measured output due to excitation input and

not from other source)

Reciprocity (Maxwell reciprocity principle: if excite 1 and obtain response

in 2, then if excite 2, obtain response in 1)

3. Provide to the student a 3D plot of an aircraft and ~2-300

accelerometers available. Ask him to define a suitable
accelerometers distribution for a GVT
Depending on what it is asked, accelerometers have to be located. (I.e. in a wing usually
put accelerometers each 2-3 ribs and in the TE and LE)

4. Digital Signal Processing (DSP). Define sampling and design the

suitable sampling for a given problem. Typical problems in DSP
(aliasing, leakage, signal noise) and how to solve them. (Long
Answer)
It deciphers the accelerometers signals to obtain the frequency contents though the Fourier
transformation. DSP considerations:
o

Sampling:

Aliasing: Is a numerical problem in DSP that arises when there is a high

frequency content in the signal but the signal is sampled at low sampling
ratio. The high frequency signal sampled at too low sampling ratio appears
as a fake low frequency mode. Need of use of anti-aliasing filters.

o Leakage: Dynamic Flight Test DFT takes a cyclic extension of data through
rectangular windows. In order to not to lose information, use Hanning
Windows which decreases side-lobes and widens main lobe.

o Signal Noise: To avoid this, split data span into overlapping segments, apply
window to each segment, apply DFT to each windowed segment and make
and average of the DFT modules. (30-50 averages with 50-75% overlap)

5. Why may you need an anti-aliasing filter?

An anti-aliasing filter is a low pass filter that eliminates high frequencies in the signal and
prevents the appearance of aliasing. A suitable anti-aliasing filter is usually 60%-80% of
Nyquist frequency.

6. Why may you need to window the signal?

Because DFT assumes a cyclic extension of the data, so a specific repetitive data can be
obtained for simplicity.

7. The Experimental Modal Analysis (EMA)

Structure characteristics: time invariant, observable, linear and obeys the Maxwell
reciprocity theory.
Test in GVT can be performed through two methods:
o

Phase Separation Method: many modes are excited simultaneously.

Phase Resonance Method or modal tuning: modes excited one by one.

The Phase Separation Method (PSM) excite randomly many modes, hence, the result that
are going to be obtained are not precise (are correct but not exact) because exciting
various modes at the same time it can be obtained one exact mode but the other ones are
not exact. It will measure the FRF and then using Modal Parameter Estimator the
Experimental Modal Model will be obtained.
The Phase Resonance Method (PRM) excite each mode one by one, hence, the result that
is going to be obtained is exact.
One classical excitation of the GVT or FVT is a frequency sweep from very low
frequencies to large frequencies with a constant spectrum. The Frequency Response
Function FRF is the ratio H (w) = X (w)/Y (w). The Dynamic Amplification Factor is
obtained once known the value of H (w) normalized to 1.0 at w0 D = |H (w)|.

The critical damping is the damping needed to change from an oscillating solution to a
non-oscillating solution.
1-dof Experimental Modal Analysis using FRFs
or Dynamic Amplification Factor is a problem of
curve fitting: Determine Omega_n and damping
factor that better match the test result using:

8. Why may it be necessary to update the dynamic FE model?

Because it is important to match the structural and mass model results with respect to the
GVT results. If this is achieved, the Dynamic Model is validated. If not, the structural
model has to be updated until the results match.

9. How to compare experimental and theoretical normal modes.

Modal Assurance Criterion (MAC).
The MAC is defined as a scalar constant relating the degree of consistency between one
modal vector and another reference modal vector as follows:
Through a frequency comparison in which GVT modes and FEM modes are plotted. If a
45 degrees slope is obtained, both modes are considered good matching.

10. Techniques to update dynamic FE models to match GVT results.

There are basically 4 techniques:
o

[Kaa] matrix update: by changing terms or introducing a global factor on

the stiffness matrix

Adding absent components in

FE models: in order to include
the effect of these components
considered important in the
analysis

Optimization techniques: NASTRAN SOL 200: it computes the sensitivities

of the frequency of all modes with respect to the FEM variables.

Delta stick Approach: Missing stiffness due to the assumptions (i.e.

minimum thickness instead of chemical milling is assumed in the FEM).
Thus, extra stiffness is included but not the mass.

11. Highlight the pros and cons of the different techniques to update
dynamic FE models to match GVT results.
The preferred approach for a dynamic FEM updating to match GVT is the delta-stick
approach. But there are other techniques to update the FE models to match GVT.
o

[Kaa] matrix update: Improve the dynamic FEM model but changing
terms do not have a physical interpretation and only applicable to tested
configurations

Adding absent components in FE models: Relevant components are

included in the analysis.

Optimization techniques: NASTRAN SOL 200: Identifies and select

variables that are relevant for the analysis but the resulting dynamic FEM
has different properties than the check stress model used for static test
(diverge from check stress model). May lead to problems with
Airworthiness Authorities.

Delta stick Approach: It keeps the original check stress model and add
missing stiffness to the mode as a backbone of check stress model. The stick
models runs along the elastic axis of wing, HTP, VTP and fuselage.

Conservative assumptions embodied in the check stress model make the model slightly less
stiff than reality (thats why Delta stick is the preferred). This way allows keeping the
check stress model as it is while the delta stick adds or remove the stiffness in the
components that need to be updated.

LECTURE 06. 3D UNSTEADY AERODYNAMICS.

DLM
1. Conceptual introduction to Finite Element Method in 2D steady
aerodynamics, (using 1 element, using 2 elements, using 3
elements)
Using 1 element:

Using 2 elements:

Using 3 elements:

2. Conceptual description of the Doublet Lattice Method

Vortex Lattice Method (VLM) takes into account only the wing, but in unsteady
aerodynamics, the wake needs also to be modeled.
Departing from the unsteady potential equation:

And deriving w.r.t t and x, the two equations obtained are combined yielding:

This is the equation of the pressure potential which solution, that satisfies all
constraints, is the pressure doublet. It can be solved with some integrals and
algebra,
In the pressure potential, the doublets are the equivalent to velocity potential
vortices. If the VLM is extended to account for oscillatory motion by adding a
line of incremental oscillatory doublets of constant strength to the bound vortex
along the c/4 of each box, the DLM is obtained.
DLM scope is: Linear Aerodynamic Potential Theory, Subsonic, and Harmonic
Variation of w (t).
3D cases should be solved by using FEM technique (with strong assumptions like:
linearity, small motions, harmonic motion, neglect viscosity and thickness
effects, subsonic, etc)
DLM is modelled using flat thin panels subdivided in trapezoidal boxes.
Singularities are located in the c/4 line with collocation points at 3c/4 of each
box. The steady part is solved using the VLM where singularities are horseshoe
vortices. The unsteady part is solved by using the DLM where singularities are
pressure doublets.

3. Provide to the student with a 3D view of an aircraft wing an

ask him to design a DLM wing model.

Additional recommendations:
-

Spanwise divisions must be maintained along the different panels located at the
same plane.

Chordwise divisions should be maintained along the span

Surface intersection must be coincident with panel divisions

Locate more aerodynamic boxes where higher pressure gradients are expected:
wing tips, leading edges

Use a regular distribution of aerodynamic boxes, take advantage of the

symmetry of the A/C and use different aerodynamic panels for control surfaces.
4. Provide to the student with a 3D view of an aircraft wing and an
HTP an ask him to design a DLM model of wing and HTP

5. Given a set of normal modes frequencies of interest (and a value of

c and V) ask for definition of 10 reduced frequencies suitable for
the analyses using those modes.
With known c, V and f, reduced frequency range can be calculated with:

6. How to build a good DLM model (mesh size, spanwise and

chordwise divisions, etc.)
-

Mess size as a balance between CPU cost and the highest normal mode that has
to be represented because higher normal modes use to have more nodal lines
both spanwise and chordwise that requires an increased number of boxes

Spanwise divisions must be maintained along the different panels located at the
same plane.

Chordwise divisions should be maintained along the span

Surface intersection must be coincident with panel divisions

Locate more aerodynamic boxes where higher pressure gradients are expected:
wing tips, leading edges

Use a regular distribution of aerodynamic boxes, take advantage of the

symmetry of the A/C and use different aerodynamic panels for control surfaces.

7. Aero-structure coupling: description of the splining matrix [G] and

its properties. Conditions needed to determine [G] matrix
components.
In all professional FEM solvers with aeroelastic capabilities, any number of
structural and aerodynamic models may exist and may be connected to each
other in such a way that the aerodynamic forces can be mapped to the
structural model and the structural deformation can be mapped to the
aerodynamic model to allow aeroelastic forces to be computed. Aeroelastic
coupling bring these two models together using splining concepts that define the
Spline Methods (Displacement, Force or General). Splines provide interpolation
capability that couples the disjoint structural and aerodynamic models for two
purposes:

Force Interpolator to compute a structurally equivalent force

distribution on the structure given a force distribution on the
aerodynamic mess.

Displacement interpolator to compute a set of aerodynamic

displacements given a set of structural displacements.

Here G is the spline matrix, that relates the forces and displacements between
structure and aerodynamic models and allows the force and displacement
interpolations.
This matrix G is computed internally by the FEM solver using several
assumption/constraints:
o

Structural displacements are assumed to be part of an infinite thin plate

Force transformation must be computed such that the resultant

structural load are statically equivalent to the aerodynamic loads

Similar condition must be satisfied for moments

8. DLM model validation flowchart

9. DLM model checking (4 steps)

-

Check geometry: Check correct modeling of the FEM and that a correct
meshing has been performed.

Check interpolated displacements: Check same normal mode shape plotted in

structural (displacements) and aerodynamic (displacements) grids.

Check steady pressures distribution: Check quality of the unitary cases

pressures, unitary angle of attack and unitary control surface rotation.

Check steady aerodynamic derivatives: Quality of main aerodynamic derivatives

comparing Aerodynamic Data Base with Double Lattice Method Results and
verify evolution with Mach is as expected, etc.

10. Describe a process to update DLM model to match CFD or test

data
-

The DLM model of the wing is divided in Spanwise strips. The total
aerodynamic lift and total aerodynamic moment of each strip is obtained by
both, the DLM method and the CFD or test and compared between them

In the process to update DLM model, each box of the DLM model strip will
have a pondering factor a.

An optimization loop is performed to obtain the set of pondering factors ai such

that the DLM strip lift and moment match ADB strip data. The additional
optimization constraint is to select the pondering factors as close to 1.0 possible.

11. Describe the final DLM model output.

LECTURE L07. THE FLUTTER EQUATION AND ITS

SOLUTION
1- Derive the flutter equation from the Lagrange equations and present
the k-method to solve it.
Equations of slides 7 to 18.

2- Describe what a V-g plot is and how to interpret it.

For each reduced frequency (k) the complex eigenvalue-eigenvector problem is solved. Once
the problem is solved the frequency is obtained from the real part of the eigenvalue and the
damping from the imaginary part. The process is repeated for all the reduced frequencies k.
By plotting the root loci of all the eigenvalues in the f-V and g-V, it is possible to track the
evolution of the frequency and damping of the normal modes with the flight speed.
This way of representing the evolution of the frequency and damping of the normal modes is
called the V-g plot. When the curve of damping crosses from the stable region to the unstable
region, the crossing speed is the flutter speed VF (and the corresponding frequency is the flutter
frequency).
In a V-g plot, a negative damping obtained from the eigenvalue means that the system is
STABLE (because negative damping should be added to a damped motion to obtain a
harmonic solution). On the other hand, positive damping means that the system is
UNSTABLE.
When damping curve crosses the g=0 line, this is the flutter speed and when it crosses the
g=0.03 is an indication on how severe is the flutter mechanism (from Abrupt V<10KEAS to
Mild V>200KEAS.)
When the curve never crosses the g=0.03 line, flutter can be categorized as benign. In addition
if the curve never goes beyond 0.01 it could be categorized as a hump mode.

3- Meaning of damping sign in the flutter equation solution.

In the STABLE region, any perturbation that moves the system away from its equilibrium
condition produces a subsequent damped motion but this motion is not harmonic, it converges
after a while. Negative damping should be added to the damped motion to obtain a
harmonic motion (and then fulfill the assumptions used to derive the flutter equation and the
DLM unsteady aerodynamics).

In a V-g plot, a negative damping obtained from the eigenvalue means that the system is
STABLE (because negative damping should be added to a damped motion to obtain a
harmonic solution).
In the UNSTABLE region, any perturbation that moves the system away from its equilibrium
condition produces a subsequent divergent motion, but this divergent motion is not harmonic.
Positive damping should be added to the divergent motion to obtain an harmonic motion.
In a V-g plot, a positive damping obtained from the eigenvalue, means that the system is
UNSTABLE (because positive damping should be added to a divergent motion to obtain a
harmonic solution)

4- Describe some classical flutter mechanisms.

5- Why the engines are located forward from wing elastic axes in the
current airliners design?
In order to shift forward the Xcg the engines are located forward from wing elastic axes.
Shifting forward the Xcg any up-bending motion will cause a nose down movement, thus
stabilizing the motion from the aeroelastic standpoint.

6- Why are the external stores located forward from wing elastic axes
in current fighters design?
In order to shift forward the Xcg the external stores are located as forward as possible in
military airplanes. Shifting forward the Xcg any up-bending motion will cause a nose down
movement, thus stabilizing the motion from the aeroelastic standpoint.

7- Why a control surface should be mass balanced?

In an aircraft with manual controls, a control surface rotation mode is a mechanism at almost
zero frequency. The frequency of this mode increases with flight speed. On the other hand the

lifting surface bending mode frequency has a smooth evolution with flight speed. Therefore,
there is a range of velocities in which there is a coalescence of frequencies between the control
surface rotation mode and the lifting surface bending mode.
If the control surface is not mass balanced, the control surface center of gravity lies well behind
the hinge line. An up movement of the lifting surface will generate a positive rotation on the
control surface due to inertia coupling by the rear c.g. position. This will create a curvature in
the airfoil that will create additional positive lift that in turn will tend to increase the
movement. Similarly, a down movement of the lifting surface will generate a negative rotation
of the control surface. This creates a curvature in the airfoil that will generate negative lift
again tending to increase the movement. This is an unstable behavior: flutter.
In the plot showing the damping evolution with flight speed, the mode corresponding to the
control surface will be always well damped but the mode corresponding to the lifting surface
will exhibit flutter in the same velocity range of frequency coalescence.

The stability situation of the system can be reserved by mass balancing the control surface. If
weight is added ahead of the hinge line, the coupling between control surface and bending will
be progressively removed. If the control surface c.g. lies ahead of the hinge line, the physical
behavior of the control surface following a movement of the lifting surface will be exactly the
opposite as the one described above, and therefore, the effect of the resulting airfoil induced
curvature will always have a stabilizing effect.

8- Flutter unmatched points vs. flutter matched points

Up to the end of the XX century, the flutter analyses were unmatched, meaning that the
Mach number selected, the altitude and the obtained flutter speed were not a single point of
the standard atmosphere. Therefore, unmatched flutter boundaries were calculated to estimate
conservative flutter margins.

In the 90s, the increase of the computer

power allowed to perform matched
flutter analyses

meaning

that the

analyses are made at constant Mach

number. The flutter solver includes
another

internal

loop:

V-g

plot

computed in KEAS. For each value of

KEAS and at that Mach number, there
is an altitude that matches the ISA
conditions. The loop is continued speed
after speed until the maximum speed
(KEAS) of interest is reached.

9- Explain how it is performed the flutter analysis in industry

Elaborated plots of flutter boundaries versus flight envelope for each mass configuration are
obtained from the successive V-g plots calculated for each Mach number. From them the
flutter mechanisms are identified and classified: Symmetric or Antisymmetric, Flutter speed
and flutter frequency, Whether they are inside, close to or outside the flight envelope, whether
they are mild, moderate, severe, abrupt & Level of reliability of the mechanism.
One way to assess the reliability is by the isolation of the mechanism. In the aeroelasticians
jargon, isolation of flutter mechanism means determine the minimum number of normal modes
that reproduces the instability.

10-

Explain what are the aeroelastic stability margins required by

civil airworthiness regulations in the paragraph 25.629

11-

Flutter is a stability problem: Does it depend on the initial

conditions?

Like all the stability problems, flutter does not depend of the initial conditions.

12-

Given an aircraft with well-known wing flutter results If a

new under-wing heavy pod is installed in the wing Are the original
wing flutter analysis still valid or have they to be re-computed for
the new system wing-pod?

If the position of the center of mass is changed with this new system, flutter analysis have to
be redo. (Remind that just an additional paint layer can change the CG position).

LECTURE L08. FLUTTER SPEED SENSITIVITIES.

MASSBALANCE. FCS.
1- Explain reasons why flutter sensitivities are needed.
This part of the analysis start when the nominal analysis have been already performed and
the basic flutter mechanisms of the aircraft have been identified. Each flutter mechanism will
have a critical set of specific mass config. and Mach number that produces the lowest flutter
speed from the entire range. For each flutter mechanism, the flutter sensitivities have two
missions:
-

Find solutions. Explore in the design envelope what has to be changed to

remove the flutter instability: mass balance; c.g. position; structural
reinforcement; aerodynamic changes
Cover any possible uncertainty in:
Structural modeling
Mass modeling
Unsteady aerodynamics modeling

2- Control surface aerodynamic hinge moment flutter sensitivity.

The control surface aerodynamic hinge moment has a top relevance in aeroelasticity:
- The frequency of the control surface rotation mode increases with flight speed (V)
- The rate of change of the frequency with V depends on the CH
- The larger the CH the fastest the increase of frequency with V
- Therefore the velocity of the potential couplings with another mode (i.e. a
bending mode) change with CH
The flutter analysis sensitivity is to repeat the flutter analysis by varying the pressure
correction factor in the control surface:
- A typical range of variation is between [0.1 - 2.0] where 1.0 corresponds to the pure
DLM results.
- This wide variation allows to the aeroelasticians to have a good insight on all
-

This sensitivity is performed to all flutter mechanisms and all control surfaces (x3)

3- In an aircraft with manual controls, why control surfaces must be

mass balanced?
In an aircraft with manual controls, a control surface rotation mode is a mechanism at
almost zero frequency. The frequency of this mode increases with flight speed. On the other
hand the lifting surface bending mode frequency has a smooth evolution with flight speed.

Therefore, there is a range of velocities in which there is a coalescence of frequencies

between the control surface rotation mode and the lifting surface bending mode.
If the control surface is not mass balanced, the control surface center of gravity lies well
behind the hinge line. An up movement of the lifting surface will generate a positive
rotation on the control surface due to inertia coupling by the rear c.g. position. This will
create a curvature in the airfoil that will create additional positive lift that in turn will
tend to increase the movement. Similarly, a down movement of the lifting surface will
generate a negative rotation of the control surface. This creates a curvature in the airfoil
that will generate negative lift again tending to increase the movement. This is an unstable
behavior: flutter.
In the plot showing the damping evolution with flight speed, the mode corresponding to
the control surface will be always well damped but the mode corresponding to the lifting
surface will exhibit flutter in the same velocity range of frequency coalescence.

The stability situation of the system can be reserved by mass balancing the control surface.
If weight is added ahead of the hinge line, the coupling between control surface and bending
will be progressively removed. If the control surface c.g. lies ahead of the hinge line, the
physical behavior of the control surface following a movement of the lifting surface will be
exactly the opposite as the one described above, and therefore, the effect of the resulting
airfoil induced curvature will always have a stabilizing effect.

4- Why an unauthorized repaint of a control surface must be

prohibited?
Because a new paint layer will put the control surface out of massbalance tolerances. It is
forbidden in the Aircraft Maintenance Manual to repaint a control surface without removing
previous painting before.

5- Describe a procedure to fine tuning of massbalance during control

surface final assembly.
Mount the control surface fully equipped at the hinge position.
Measure the force, F (N), at a known position, F is considered positive (+) downwards,
meaning that the center of gravity of the control surface is behind the hinge line according to
the flight direction. F is considered negative (-) upwards, meaning that the center of gravity
of the control surface is in front of the hinge line according to the flight direction.
Install or remove the necessary amount of balance weight at the specified location to meet the
balance requirements. Table defining quantities of mass balance will be provided for each
control surface in terms of the measured force F.
Obtain the new value of F by repeating steps 1) and 2). If this force is not within the specified
limits, corresponding to the balance requirements, the amount of balance weight must be
corrected until the final force lies inside the required tolerances.

6- Comment on the amount of mass required for massbalancing. Is it a

significant percentage of the control surface total weight?
The amount of mass required for massbalancing must be sufficiently large to avoid positive
damping (meaning unstable system) for a specific range of velocities.

7- Control surface actuator stiffness flutter sensitivity.

In order to solve the probability of failure of actuators, put two actuators to avoid failure.

8- External stores flutter sensitivity.

The addition of heavy under wing masses modifies the wing normal modes and requires to
perform a complete set of new flutter analyses. Under wing heavy masses include:
-

Engines

Air-to-air Refuelling pods

External Fuel tanks

External stores, etc

The following sensitivities have to be performed:

-

Mass characteristics of the engine, pods or external store

Stiffness characteristics of the pylon

Local stiffness of the attachments

In case of external stores, all subsequent release sequences including failure

cases and antisymmetric configurations.

9- What

of

the

following

design

changes

substantiated?
1. Add winglets at the wing tip (YES/NO)
2. Open the rear door (YES/NO)
3. Deflect the ailerons (YES/NO)
4. Etc

need

to

be

flutter

For the first case, YES. Winglets will modify all aspect related with flutter analysis: winglet
mass affects wing modes, winglet aerodynamics modify wing aerodynamic and stiffness is
affected due to the addition of reinforcements.
For the second case, YES. As the structure pass from a closed section to an open section,
modifying the rear fuselage stiffness and therefore a significant drop in frequency of rear
fuselage anti-symmetric torsion modes. Here, the anti-symmetric tail flutter mechanisms are
significantly affected by this drop of stiffness.
For the third case, NO. As ailerons are already mounted in the wing it will not affect the
mass aspect.
For the rest of cases, every aircraft feature that may have an effect on the normal modes
should be substantiated from the flutter standpoint.

10-

What is the control surface fitting failure usually most critical

for flutter?

The failure cases are modelled in the FEM model by

removing the corresponding elements, one by one. The
most critical fitting failure in a control surface use to
be the one located most outboard.
In this case of a VTP, the Hinge 8.

11-

What is the delamination zone in a control surface usually most

critical for flutter?

In current composite control surfaces, the effect of composite delamination should be assessed
from the flutter standpoint.
The control surface area most critical use to be the area located immediately outboard of the
actuators.
As the delaminated area increases, the flutter mechanism tends to become severe. This
information is used to define the maximum allowable delamination inside the Aircraft
Maintenance Manual.

12-

Describe the analysis needed to cover water ingress in a

sandwich structure of a control surface from flutter standpoint.

Water that remains in the honeycomb cells

during aircraft operation and moves Xcg
rearwards, what is a potential threat of
aeroelastic couplings.
Water ingress in control surfaces is modelled
in two ways: either distributed (yellow dots)
or concentrated in the trailing edge (red
whichever is worst. A typical value accepted
by Airworthiness Authorities is 1 kg/m
concentrated in the trailing edge. The
elaborated plot of (flutter speed) vs (kg/m)
in the trailing edge, shows no problem for 1.0
kg/m, a flutter mechanism starts around 1.4
kg/m and it becomes severe (i.e. crosses also
g=0.03 line) around 1.7-1.8 kg/m.

13-

Definition of aeroservoelasticity and physical description of an

example of longitudinal FCS (symmetric) potentially unstable
coupling.

Aeroservoelasticity is the study of the interaction of all elements relevant for aeroelasticity
plus Flight Control System laws. (Requested in Airworthiness Regulations).

The solution options are limited to filtering the gyro output signal or moving the gyro to a
different location. The gyro can be relocated from the forward position to the anti-node
which eliminates the vibration [Notch filter to eliminate all the undesired responses but adds
a delay]

14-

Example of lateral FCS (antisymmetric) potentially unstable

coupling

LECTURE L09. FLIGHT FLUTTER TEST (FVT).

AEROELASTIC MODEL VALIDATION.
1- Define Flight Vibration Test
Flight Flutter Test is aimed to:
-

Determine experimentally the evolution of normal modes frequency and damping

with flight speed.

Ensure that the aircraft is free from flutter in all its envelope up to Vd/Md,

Validate the aeroelastic model

In order to avoid confusion with the Fast Fourier Transform (FFT) acronym, the acronym for
Flight Flutter Test is FVT.
It is explicitly required by the administrator in CS25.629. The test has to be performed when
a change in the structure is applied, although the change include an improvement in the flutter
behavior. When the change is insignificant such as including an antenna there is no need to
do the test.

2- Describe the objectives of the FVT.

The objectives of the Flight Flutter Test are:
-

Determine experimentally the evolution of frequency and damping of the aircraft

normal modes with flight speed

Ensure that all modes are properly damped and that there is not rapid decrease of
damping of any mode inside the aircraft flight envelope. (Clear the flight envelope
from the flutter standpoint).

Contribute to validate the aeroelastic model.

3- In what situations is the FVT required by the Airworthiness

Regulations in CS25.629?
It is always mandatory for new type designs and for all design modifications unless the effect
is insignificant (even if the change include an improvement of the flutter behavior).

4- Describe the FVT by explaining the different components of this test

The test has to be planned years in advance.
Specimen: It is the first prototype of an aircraft configured in the most critical configuration
from the flutter standpoint.

Controlled excitation: There are several ways to excite the aircraft like natural turbulence,
bonkers (but they change the aerodynamics of the wing), external aerodynamic vanes (located
at the wing tip): Oscillatory vanes & Aerodynamic vanes with slotted Trailing Edge and the
control surfaces of the aircraft. From this last one, in aircraft with manual controls, stick raps
are used (pilot hit the stick with a hammer and it is excited); in aircraft with actuators and
FCS, pulses and sine sweeps are used. When the aircraft can fly at stabilized conditions, it is
better to use sweep and pulse. In flight points that can only be achieved by diving, only pulses
can be performed when aircraft crosses exactly the flight point conditions. At Vd/Md the
aircraft may be excited only with natural turbulence.
Transducers: Accelerometers, normally located at wing and tail tips, engines and rear and
forward fuselage.
Data acquisition: The same considerations as for GVT.
-

Sampling. Should be large enough to capture all phenomena. The typical value are
on the range of 128-512 samples per second. Nyquist frequency is half of the
sampling frequency.

Aliasing: Prevent it using a low-pass filter.

Leakage: Avoid by windowing the signal (Hanning windows)

Noise: To avoid this, split data span into overlapping segments, apply window to
each segment, apply DFT to each windowed segment and make and average of the
DFT modules. (30-50 averages with 50-75% overlap).

Real Time Analysis: FVT are analyzed at the same time as it is being performed. Aircraft
data and voice is received in the telemetry room of the flight test center so decisions can be
made at real time.

5- Means of exciting the aircraft in a FVT

-

Natural turbulence (controlled excitation)

Bonkers (pyrotechnic devices)

External aerodynamic vanes (located at wing tip):

Oscillatory vanes

Aerodynamic vanes with slotted Trailing Edge

The same control surfaces of the aircraft

o

In aircraft with manual controls,

Stick raps

In aircraft with actuators and FCS,

Pulses

Sine sweeps

The preferred option is logically the use of the same control surfaces of the aircraft: in this
way, the configuration of the aircraft being tested in FVT is identical to the series aircraft.
-

In the flight points that the aircraft can fly at stabilized conditions; both, sweep
and pulses are performed

In the flight points that can only be achieved by diving, only pulses can be
performed when the aircraft crosses exactly the flight point conditions

For the final point at Vd/Md the aircraft is excited only with natural turbulence

6- Explain the similitude and the differences between GVT and FVT

7- Describe how are the excitations in a FVT depending on the flight

envelope point that is going to be tested.
The preferred option is logically the use of the same control surfaces of the aircraft: in this
way, the configuration of the aircraft being tested in FVT is identical to the series aircraft.
-

In the flight points that the aircraft can fly at stabilized conditions; both, sweep
and pulses are performed

In the flight points that can only be achieved by diving, only pulses can be
performed when the aircraft crosses exactly the flight point conditions

For the final point at Vd/Md the aircraft is excited only with natural turbulence.

8- Present the aeroelastic model validation flowchart.

9- Explain briefly the contents of each of the boxes of the aeroelastic

model validation flowchart
On one hand, the integrated aeroelastic model is obtained from the unsteady aerodynamic and
dynamic models. With the model, both theoretical normal eigenvalues and numeric response
to excitation are achieved.
On the other hand, with the FVT and the excitation of the aircraft by a pulse or a sweep,
experimental modal and signal analysis can be performed giving the experimental normal
eigenvalues and the aircraft response.
Experimental and theoretical results are compared and if the match is good, the integrated
aeroelastic model is validated, otherwise, dynamic flight loads model has to be updates, thus,
new integrated aeroelastic model is obtained.

LECTURE L10. THE CONCEPT OF LOADS. MS.

TRANSIENT RESPONSE.
1- Which are the structural sizing cases in a modern aircraft?
Symmetric bending of the fuselage During dynamic landing
Wing During pull-up maneuver (Load factor of the aircraft)
Wing up bending of the aircraft During vertical gusts
Wing down bending of the aircraft During taxi loads
HTP area Taking off on a hot day with CG forward (extreme case)
VTP dimensioning Lateral gusts.

2- Define loads
Aircraft structural loads or actions are forces, deformations or accelerations applied to the
structure or its components.
-

Loads causes stresses, deformations and displacements in structures. Assessment of

their effects is carried out by the methods of structural analysis

Excess load or overloading may cause structural failure, and hence such possibility
should be either considered in the design or strictly controlled

Loads are consequences of the aircraft operations (loads due to taxi, take-off, etc.)

3- List several forms of loads classification

Depending on the type of structural analysis they are aimed for, the loads can be classified as:
-

Checkstress loads (limit loads and ultimate loads) for exceptional operation

Fatigue loads for normal operation

Depending on the aircraft operation phase in which they can be met, the loads can be classified
as:
-

Flight Loads

Ground Loads

Depending on whether they are steady or variable, the loads can be classified as:
-

Static Loads

Dynamic Loads

4- What the airworthiness regulations (CS25.301) say about loads?

-

Strength requirements are specified in terms of limit loads (the maximum loads
expected in service) and ultimate loads (limit loads multiplied by safety factor). Unless
otherwise provided, prescribed loads are limit loads

Unless otherwise provided the specified air, ground and water loads must be placed in
equilibrium with inertia forces, considering each item of mass in the airplane. These
loads must be distributed to conservatively approximate or closely represent actual
conditions. Methods used to determine load intensities and distribution must be
validated by flight load measurement unless the methods used for determining those
loading conditions are shown to be reliable

If deflections under load would significantly change the distribution of external or

internal loads, this redistribution must be taken into account.

5- Why static loads are different from dynamic loads?

The difference is mainly due to the inertia
forces. Example: exciting a clamped beam
with a transient input force at the free end.
In case of an applied static load at the beam
tip, inertia forces do not play any role in static deformation nor in the distribution of internal
forces (shear, bending). On the other hand, when the applied load is a transient excitation,
depending on the frequency and shape of the dynamic excitation, inertia forces may be
determinant in the subsequent response of the dynamic system completely changing the
internal loads distribution.
If the frequency of the excitation is
close to the first beam bending
mode, the response will be close to
this

first

bending

mode.

The

evolution of shear force along the

beam will always be positive (line in
the static case) but larger than the
static case because of the dynamic
amplification.

On the other hand, if the frequency of the excitation is close to the second bending mode, the
response will be in the second bending mode, very different from a static solution. The
evolution of shear force along the beam may even change of sign (not intuitive if compared
with a static solution).

6- Use a clamped beam example to illustrate difference between static

and dynamic loads

7- Dynamic loads process flowchart

First Conceptual flowchart:

Final Conceptual flowchart:

8- Describe the different loads loops in an aircraft project development

9- What is a 2d envelope

2D envelopes are the combination

of the Bending Moment (Mx) vs
time and Torsion Moment (My)
vs time plots in a single plot.
The combination of both graphs
forms a 2dimensional envelope.

10-

The concept of nodal loads

Nodal loads are those nodes that has a point/nodal load. In order to select which nodes have
loads, a criteria is followed: Max/Min criteria (looking at moment vs time plot and finding
max and min) & 2D envelopes criteria (selecting the critical points/corners of the 2D
envelope).

LECTURE L11. GROUND DYNAMIC LOADS

1- Define dynamic landing and why this scenario is relevant in
aeroelasticity and structural dynamics
During landing, the vertical velocity or sinking speed of an airplane is quickly reduced to zero
when the wheels strike the ground. This process is accomplished by a transfer of kinetic energy
of the sinking airplane to internal energy in the shock absorption system where it is dissipated.
The vertical velocity of the airplane is brought to zero within a fraction of a second, and hence
the forces applied to the structure through shock strut change from zero to a maximum, also
in a fraction of a second. This rapid change in velocity, or equally rapid application of force,
excites the lower vibration modes of the structure. Therefore the structural dynamics
characteristics of the structure should be taken into account.
The relevant aircraft components affected by dynamic landing are:
-

The landing gear itself

The entire fuselage (symmetric bending and shear force)

The wing down bending

The engine vertical force and vertical acceleration

Dynamic landing loads are used for:

-

Design and Checkstress of the structure Certify the aircraft

Fatigue analysis Define inspections intervals

Analysis of operational firm or hard landings Define maintenance operations.

2- Define taxi and why this scenario is relevant in aeroelasticity and

structural dynamics
Taxiing is the entire phase of straight line motion on the ground prior to final take off and
following landing. (Braking and turning are considered separate phases). Paved runways
have some degree of roughness that may excite the normal modes of the structure when
taxiing. Depending on the taxis speed, a tuning between the roughness excitation and the
normal modes of the aircraft may cause significant response amplification and dynamic
loads on the structure.
The relevant aircraft components affected by taxi loads are:
-

The landing gear itself (strokes)

The nose fuselage (vertical force)

The wing down bending

The phenomenon is exacerbated in military aircraft that must operate in unpaved runways.
Taxi loads are used for:
-

Design and Checkstress of the structure Certify the aircraft

Fatigue analysis Define inspections interval

Operation on unpaved surfaces Defines maintenance operations. Define number of

allowable runs in soft soil, etc.

3- Specific civil requirements in CS25 for dynamic landing.

Complement with examples of more exigent military requirements

For military tactical aircraft Vz = 12 ft/s while for On-board carrier aircraft Vz = 18ft/s.

4- Specific civil requirements in CS25 for taxi loads. Complement with

examples of more exigent military requirements

For military aircraft taxi operations also have to be in unprepared surfaces (DEF-STAN A,B)

5- Explain the effects of landing gear spin-up and spring-back

When the landing gear hits the ground, the tire is not rotating yet. Therefore there is a
difference in speed between the ground
and the tire. This creates a very large
friction force in the contact point that
deforms the flexible landing gear: this is
the spin-up effect that takes place in a
very short period of time ~0.1sec.
The friction force accelerates the wheel that very soon gets the needed rotation speed and the
friction force almost vanishes. Then the flexible landing gear returns to its neutral position:
this is the spring-back effect.
Due to the spin-up and spring-back effects, the forces at the pintle points are not pure vertical
forces: there are significant fore-and-aft forces that must be taken into account in the dynamic
landing simulation.

6- Dynamic Landing model validation flowchart

7- Describe the tests needed for dynamic landing model validation

-

Landing Gear Drop Test: Verification and validation of the landing gear model
including confirmation of the spin-up & spring-back effects. Both effects are

simulated in drop test by making the wheel rotate in the opposite sense at the
corresponding angular speed.
-

Aircraft firm or Hard Landings: Used to validate the coupling between landing gear
and aircraft by comparing numerical simulations with test results.

Free Extension Tests

8- Describe which parts of the aircraft may be sized by dynamic

landing
Relative importance of several dynamic loads in:

Wing shear force (Fz)

Fuselage shear force (Fz)
Wing bending (Mx)
Fuselage bending (Mx)

9- Taxi loads model validation flowchart

10-

Describe the tests needed for taxi loads model validation

Runway EBH curve: Unpaved runway roughness severity may be compared by using
EBH (Equivalent Bump Height) curves. Taxi test preformed in a paved runway in

which an artificial (1-cos) bump of known height and length has been introduced. It
validates the taxi model. [Better explained in next question]

11-

Describe what is a runway EBH curve (Equivalent Bump

Height)

Unpaved runway roughness severity may be compared by using EBH curves (EBH=
Equivalent Bump Height).
These curves show the height (H) of the obstacle that the aircraft is able to pass
without exceeding limit loads. This height H depends on the bump length (L).
Typically the larger the wavelength, the large the height that the aircraft is able to
pass.

The EBH curve can be also computed for each runway depending on its roughness (it
is a pure geometrical curve) obtained by convolution with a (1-cos) mask.

If the EBH curve of the runway is below (i.e. less severe) than the EBH of the aircraft,
then this aircraft can operate (taxi) in that runway without problems.

If the EBH curve of the runway is above (i.e. more severe) than the EBH that the
aircraft can withstand then the aircraft cannot operate in that runway. The taxi tests
are performed in a paved runway in which an artificial (1-cos) bump of known height
and length has been introduced.
The taxi tests are performed in a paved runway in which an artificial (1-cos) bump of known
height and length has been introduced.

12-

Describe which parts of the aircraft may be sized by taxi

loads

Relative importance of several taxi loads in:

Wing shear force (Fz)
Fuselage shear force (Fz)
Wing bending (Mx)
Fuselage bending (Mx)

LECTURE L12. DYNAMIC FLIGHT LOADS. DTG.

1- Define gust and turbulence from aeroelasticity point of view
-

Aircraft regularly encounter atmospheric turbulence (or rough air) of varying degrees
of severity

Turbulence may be considered as movement of the air through which the aircraft
passes

Any component of the velocity of the air (so-called gust velocity) that is normal to
the flight path will change the effective incidence of the aerodynamics surfaces, so
causing sudden changes in the lift forces and hence a dynamic response of the aircraft
involving flexible deformation

The response will involve both the rigid body and the flexible modes, may give rise to
passenger and crew discomfort and will introduce internal loads that need to be
considered for aircraft safety.

Thus it is important for the safe design of the aircraft to calculate the response and
internal loads generated under the conditions defined by the Airworthiness
Regulations, and to evaluate the effect on the fatigue life.

Gust and turbulence loads are significant throughout the aircraft. The most relevant
aircraft components affected by gust and turbulence are:

The wing up bending

The vertical tail plane

The rear fuselage (anti-symmetric bending and shear force)

Gust and Turbulence loads are used for:

o

Design and Checkstress of the structure Certify the aircraft

Fatigue analysis Define inspections intervals

Analysis of operational rough air encounters Define maintenance operations.

2- Classify the severity of gust and turbulence from aircraft point of

view and from pilot point of view
Aircraft point of view
-

Discomfort:
o

Excitation of the structural modes, creating vibrations making some tasks

difficult

o
-

Excitation of rigid body movements, driving passenger to motion sickness

Injuries of unfasten passengers

o
-

Major cause of passenger injuries

Loads: Checkstress and fatigue loads:

o

Gust/turbulence loads is a design condition for nearly all part of the structure
of modern airliners, for both static and fatigue sizing

Current gust loads computation criteria, are adequate: there is nearly no event
of a main structure failure due to a gust in recent aviation history.

Pilot point of view

3- Describe

the

ways

to

measure

the

atmospheric

turbulence,

emphasizing the type of sensors used in the aircraft dedicated to this

measurement
The ways of measuring the atmospheric turbulence:
As example, last classical reference: AGARD report 734 (issued in 1986) describing
dedicated aircraft and test campaigns in US, Canada, Europe
After measuring the characteristics of the atmospheric turbulence in large dedicated flight test
campaigns (plusballoons, satellites information, even kites with suspended sensors, towers)
etc. the results were post-processed and statistics of atmospheric turbulence agreed between
scientists.
o Agreed meaning: consensus in the scientific community, not always unanimous:
o In the civil aircraft realm the evidence is that current gust loads computation criteria,
are adequate: there is nearly no event of a main structure failure due to a gust in recent
aviation history
o For military aircraft the civil requirements are not enough for all weather conditions.
Military requirements are more severe than civil ones. Military requirements depend
on the customer.

 As an example, UK MoD use to require DEF-STAN for aircraft on RAF inventory.

Atmospheric turbulence model in DEF-STAN may results in a much more severe
scenario for aircraft design than other regulations.
But it is not arbitrary (some UK military aircraft lost in 1960s due to in-flight
structural failure attributed to turbulence and gust encounters)

4- Describe a simplified solution of the DTG response (Steady

aerodynamics, Rigid aircraft, No aircraft motion)

5- Describe the 3 factors needed to improve the simplified approach

(Unsteady aerodynamics, Flexible aircraft and aircraft heave and
pitch motion)
-

Aircraft Dynamic Response (Aircraft Flexibility): To account for dynamic effects on

structural loads
o

Structural dynamics can produce and increment or decrement relative to the

rigid A/C case

A dynamic effect depends on the frequency content of the input gust profile.

The response of the aircraft to gust loads modifies the input gust profile experienced
by the aircraft

Unsteady aerodynamics

The steady aerodynamics is not valid in this scenario. A full unsteady

approach is necessary to account for the effects of:

Airfoil movements relative to the flow

Gust penetration

6- The solution of the Discrete Tuned Gust problem

General criteria:

2fg<fmax<3fg

Defining f is equivalent to defining Period T: T=1/f. A typical value of T=20s, so = 0.05

Hz.

7- Selection of the suitable number of modes (of fmax) for DTG

solution

In the sharp edges, the frequency is larger.

8- Selection of the suitable frequency resolution (f) for DTG solution

9- Description of a complete DTG loads loop

A complete DTG loop requires covering:
-

Enough gust lengths between 30ft and 350 ft

Enough flight points in the aircraft envelope:

At Vc: at least sea level and the kink at the highest Mach number. Sometimes
intermediate values and high altitude cases

At Vd: at least sea level and the kink at the highest Mach number

Enough mass configurations in the aircraft weight envelope

o

Operational empty weight (OEW)

Maximum Zero Fuel Weight (MZFW), including several payload configurations

covering all possible centers of gravity

Maximum Take Off Weight (MTOW), both with low fuel and with max fuel,
including several payload configurations covering all possible centers of gravity

Symmetric (vertical gust) and anti-symmetric (lateral gust) cases

In summary: a typical complete DTG loop would require considering several thousands
of cases

10-

Current civil requirements for Discrete Tuned Gust analyses

Current civil regulations for DTG require:

-

The use of a full aeroelastic model made of:

o A dynamic model, properly representing the rigid body modes (RBM) and the
flexible modes of the structure
o

An unsteady 3D Doublet Lattice Method (coupled with the dynamic model)

The gust excitation is represented with (1-cos) shape gust. The analysis should consider
enough gust lengths between 30ft and 350ft

Gust intensity (at Vc, sea level) will be 56 ft/s for long gust lengths. The gust intensity
will be reduced for shorter gust lengths.

Gust intensity is also reduced for altitudes higher than sea level.

11-

Validation of the aeroelastic model for DTG analyses

12-

Indicate what parts of the aircraft may be sized by DTG

loads

Wing shear force (Fz)

Wing bending moment (Mx)
Fuselage shear force (Fz)
Fuselage bending moment (Mx)
VTP shear force (Fy)

LECTURE L13. DYNAMIC FLIGHT LOADS. CT.

1- Define gust and turbulence from aeroelasticity point of view
Aircraft regularly encounter atmospheric turbulence (or rough air) of varying degrees of
severity
Turbulence may be considered as movement of the air through which the aircraft passes
Any component of the velocity of the air (so-called gust velocity) that is normal to the
flight path will change the effective incidence of the aerodynamics surfaces, so causing
sudden changes in the lift forces and hence a dynamic response of the aircraft involving
flexible deformation
The response will involve both the rigid body and the flexible modes, may give rise to
passenger and crew discomfort and will introduce internal loads that need to be considered
for aircraft safety.
Thus it is important for the safe design of the aircraft to calculate the response and internal
loads generated under the conditions defined by the Airworthiness Regulations, and to
evaluate the effect on the fatigue life.
Gust and turbulence loads are significant throughout the aircraft. The most relevant
aircraft components affected by gust and turbulence are:
a. The wing up bending
b. The vertical tail plane
c. The rear fuselage (anti-symmetric bending and shear force)
Gust and Turbulence loads are used for:
d. Design and Checkstress of the structure Certify the aircraft
e. Fatigue analysis Define inspections intervals
f.

Analysis of operational rough air encounters Define maintenance operations.

2- What is the rms of a given signal y(t) and what is the ?

The magnitude of the fluctuations of y around is mean value (zero) is measured by a special
kind of average, the root mean square (rms) value is given by :

If y(t) is the transient response to an excitation f(t) (with its own rms value),
then the term in aeroelastic response is usually used to reflect the ratio
between the rms of the response with respect to the rms of the excitation:

3- Definition of PSD

Why PSD is useful in solving dynamic problems? Because there is a direct relationship
between PSD output and PSD input through the transfer function:

4- Definition and conceptual examples of autocorrelation function

5- Mathematical expression for PSD

6- Units of PSD. Give as an example the units of PSD if y(t) is an

acceleration measured in gs

7- Historical evolution of the continuous turbulence analyses in the

Airworthiness Regulations
Slides 13 to 17

8- Current civil requirements for Continuous Turbulence analyses

The PSD excitation requested by current airworthiness regulations for continuous turbulence
analysis is the Von-Karman Spectrum.
The Von-Karman spectrum concentrates most of the excitation at very low frequencies.

9- The solution of the Continuous Turbulence analyses

Slides 20 to 24

10-

Conceptual description of correlated load cases

11-

What is the shape of a 2D continuous turbulence load

envelope?

An ellipse

12-

Description of a complete CT loads loop

13-

Validation flowchart of the aeroelastic model for CT analyses

14-

Indicate what parts of the aircraft may be sized by

Continuous Turbulence loads

Wing shear force (Fz)

Wing bending moment (Mx)
Fuselage vertical shear force (Fz)
Fuselage symmetric bending moment (My)
Fuselage anti-symmetric lateral bending moment (Mz)
VTP shear force (Fy)
VTP bending moment (Mx)

LECTURE L14. DYNAMIC FLIGHT LOADS. BUFFET.

1- Define buffet and buffet loads
-

Buffet is a response problem result of pressure fluctuations exciting the structural

modes
Buffet flow consists of random, fluctuating pressures caused by separated flow
Buffet Onset = the first appearance of a significant area of separated flow
thereby causing fluctuating pressures
Buffet loads = Loads due to flow separation

2- Classify different types of buffet according to the aircraft

component affected
-

Wing buffet (usually self-generated, i.e., flow separation of the same lifting
surface)
o

Magnitude of response depends on magnitude of unsteady flow

disturbances

May be due to

Alpha (AoA)

Transonic effects (Shock waves)

Empennage buffet (Impingement of the flow detached from a forward lifting

surface or deflected device)

HTP affected by flow separated from the wing

VTP affected by flow separated from the wing leading edge apex

VTP affected by flow separated from airbrakes

Component buffet (local flow separations)

o

Flaps (deployed) buffet

Pods/Pod fin

Landing gear doors

3- Explain the differences between flutter and buffet

Flutter:
- Flutter is a STABILITY problem
- Classical Flutter is produced by the coupling of two or more normal modes.
- There is a coalescence of frequencies in the evolution of frequency of these modes with
flight speed.
- The phase between the modes is such that the coupling allow extracting energy from
the free flow and makes the subsequent structural motion divergent.

Flutter may be a potentially catastrophic event

Flutter use to be a high speed phenomena (Flutter onset usually triggered by dynamic
pressure)

Buffet:
- Buffet is a RESPONSE problem
- Buffet is the structural response to flow separation. Buffet loads are loads produced by
flow separation. Most often is the response of a single mode.
- Buffet use to be a fatigue issue
- Buffet onset usually triggered by angle of attack

4- Describe the Von-Karman vortex street as a classical example of

flow separation
A von Krmn vortex sheet is a repeating pattern of swirling vortices caused by the unsteady
separation of flow of a fluid around blunt bodies. A vortex street will only form at a certain
range of flow velocities, specified by a range of Reynolds numbers (Re).
The range of Re values will vary with the size and shape of the body from which the eddies
are being shed, as well as with the kinematic viscosity of the fluid. Over a large Re range
(47<Re<105 for circular cylinders) eddies are shed continuously from each side of the body,
forming rows of vortices in its wake. The alternation leads to the core of a vortex in one row
being opposite the point midway between two vortex cores in the other row, giving rise to the
distinctive pattern shown in the picture. Ultimately, the energy of the vortices is consumed by
viscosity as they move further downstream, and the regular pattern disappears.

5- Define Strouhal number

6- Use the Strouhal number and speeds probabilities to derive

frequency design requirements in a civil structure

7- Buffet in the airworthiness regulations

CS25.251 Vibration and buffeting
- The aeroplane must be demonstrated in flight to be free from any vibration and
buffeting that would prevent continued safe flight in any likely operating condition.
- Each part of the aeroplane must be demonstrated in flight to be free from excessive
vibration, under any appropriate speed and power conditions up to VDF/MDF. The
maximum speeds shown must be used in establishing the operating limitations of the
aeroplane in accordance with CS 25.1505.
- Except as provided in sub-paragraph (d) of this paragraph, there may be no buffeting
condition, in normal flight, including configuration changes during cruise, severe
enough to interfere with the control of the aeroplane, to cause excessive fatigue to the
crew, or to cause structural damage. Stall warning buffeting within these limits is
allowable.
- There may be no perceptible buffeting condition in the cruise configuration in straight
flight at any speed up to VMO/MMO, except that the stall warning buffeting is
allowable.
- For an aeroplane with MD greater than 06 or with a maximum operating altitude
greater than 7620 m (25,000 ft), the positive maneuvering load factors at which the
onset of perceptible buffeting occurs must be determined with the aeroplane in the
cruise configuration for the ranges of airspeed or Mach number, weight, and altitude
for which the aeroplane is to be certificated. The envelopes of load factor, speed,
altitude, and weight must provide a sufficient range of speeds and load factors for
normal operations. Probable inadvertent excursions beyond the boundaries of the
buffet onset envelopes may not result in unsafe conditions. (See AMC 25.251(e).)

8- Describe qualitatively the relevant characteristics in a wing buffet

case

9- Describe qualitatively the relevant characteristic of empennage

buffet using the F-18 case as example
Slides 15 to 19

10-

Describe qualitatively the relevant characteristics of fin buffet

using the Eurofighter fin buffet due to aero brake deployment as
example

11-

Application of wind tunnel tests in understanding the buffet

problem

12-

The solution of the Buffet Problem (both in the aerodynamic

side and in the structural side)

AERODYNAMIC SIDE:

STRUCTURAL SIDE:

13-

Buffet analysis flowchart highlighting the application of the

PSD technique to solve the buffet problem (in the structural side)

14-

Validation flowchart of the aeroelastic model for Buffet

analysis

15-

Indicate what parts of the aircraft may be sized by Buffet

loads

Wing, VTP, HTP etc.