Sunteți pe pagina 1din 37

The FLEXOP Project

Flutter Free FLight Envelope eXpansion for ecOnomical Performance improvement


General Overview
Balint Vanek,
Institute for Computer Science and Control, HAS
(vanek@sztaki.hu)
coordinator

09 12 2016
University of Minnesota

FLEXOP, UMN Seminar, Minneapolis, 09-12-2016


Consortium
• Project Duration: 42 months
• Budget: 6.7 m EUR

Scientific Advisory Group


• UofM – Pete Seiler
• Airbus – Frank Theurich
• Aerospace Control Dynamics -Dagfin Gangsaas
• Imperial College – Rafael Palacios

FLEXOP, UMN Seminar, Minneapolis, 09-12-2016 2


Project Context

• Modern large aircrafts have to (among other important points!)


• Minimize the fuel consumption
• Propose a very high level of comfort

Long wing, Low


Fuel long fuselage Flexible frequency,
consumption structure poorly
Light damped
structure modes
requirement design consequence effect

• The trade-off structure/efficiency must guarantee a good level of comfort

DESIGNING THE AIRCRAFT:


absolutely essential to ensure a sufficiently stiff structure
possibility to minimize the structural vibrations with additional means

EURO GNC presentation of Stéphane Delannoy


Airbus expert in Aircraft control FLEXOP, UMN Seminar, Minneapolis, 09-12-2016
The Key elements to design such a
function:
1. Aeroelastic model of the aircraft:
• Pre-flight mathematical models: theroretical model
• Identification in flight: identified model
2. Sensors:
• Defined on the pre-flight model
• Validated in flight
3. Criteria:
• EASA/FAA international standards
• Internal manufacturer know-how criteria
4. Design of the control law

5. Flight tests

6. Certification

FLEXOP, UMN Seminar, Minneapolis, 09-12-2016


Approach

Move towards methods and tools enabling multidisciplinary design analysis and
optimization in the aeroservoelastic domain
Validate the developed tools with the demonstrator

FLEXOP, UMN Seminar, Minneapolis, 09-12-2016


WP structure

FLEXOP, UMN Seminar, Minneapolis, 09-12-2016


FLOW CHART WP1: -0/-2 wing

FLEXOP, UMN Seminar, Minneapolis, 09-12-2016


WP1 – TUD Aeroelastic Tailoring

Stiffness distribution Thickness distribution Buckling constraint

(-0) (-0) (-0)

(-2) (-2) (-2)

FLEXOP, UMN Seminar, Minneapolis, 09-12-2016


WP1 – DLR Aeroelastic Tailoring

• Objective
– tools for aeroelastic tailoring – based on a NASTRAN shell model
framework
• Approach followed:
– Input from TUD’s optimization process
• Wing structural layout – position of spars & ribs, optimal jig-
twist, optimal stiffness design as starting point
– Optimize stacking sequence of wing-box - skins and spar shear
webs using a Stacking Sequence Table –based Genetic Algorithm
– Output to FACC – optimized stacking sequence for manufacturing
of wings -0 and -2
– Flight airworthiness check – man-hole covers, skin-spar glue-
joints at interface
FLEXOP, UMN Seminar, Minneapolis, 09-12-2016
WP1 – DLR High-Dimensional Modelling

• Objective
– Nastran FE model generation with the parametric
modeling engine ModGen for wing -0/-2
– provision for a well-defined interface (wings -0, -1
& -2) to WP2
• Status
– Fuselage – beam element representation using
cross-sectional modeler with planned system
masses
– V-tail – generated as shell structure using ModGen
– Interface established with WP2 (DLR-OP) for
delivery of full a/c + condensed models for eigen &
aeroelastic analysis using preliminary (-0) wing
design FLEXOP, UMN Seminar, Minneapolis, 09-12-2016
WP1 – TUM Flutter (-1) Wing Design

Preliminary Aeroelastic
AX-8 Wing-1 Objectives: model:

- Flutter Speed below 55 m/s - Parametric


- Structural Sizing to 5g - Design development
- Parameter studies
- Low Flutter frequency - in best case <10Hz
- No control Reversal Detailed Aeroelastic
model:
- Sufficient high divergence speed
- Include of Shell offsets
- No smeared stiffness
approach
AX-8 Wing-1 results: - Adaption necessary due to
manufacturing issues
- Opposing trends in stiffness requirements – Low Wing Stiffness: - Detail adjustment of wing
torsional stiffness by wing
+ Low Flutter Speed shell layer
+ Low Flutter Frequency - Detail adjustment of flutter
- low divergence speed trim masses
- high structural wing deformation (aerodynamic problems worsen by wing sweep)
- control reversal

- Additional flutter tuning masses:


+ Low Flutter Frequency
- Additional structural dynamic loads

FLEXOP, UMN Seminar, Minneapolis, 09-12-2016


WP1 – TUM Flutter Wing Evaluation

• Concept N: Assymetric second-bending/torsion coupled flutter with 5Hz @ 45 m/s


• 400 g flutter tuning mass in each wingtip
• Sensitivity for small design changes
• Small margin for divergence speed

FLEXOP, UMN Seminar, Minneapolis, 09-12-2016


WP1 – AGI-G Flap Design

Overall Flap Layout


- 4 equal length flaps per wing
- one flap fits all design:
 all wings use the same flaps
- hinge position based on max. local
wing bending
- flap deformation decoupled from wing
during normal operation
Design Requirements

- min. eigenmode at 50Hz

- precise integration with wing (minimal bearing clearance)


1st mode at 69Hz
- very light weight to reduce rotational inertia (especially

flap #4)

- low structural complexity with a high operational

reliability

FLEXOP, UMN Seminar, Minneapolis, 09-12-2016


WP1 – AGI-UK Aerodynamic Investigation

Objectives/ Activities in WP1

• Development of baseline requirements for 3 Wings based on overall A/C requirements by TUM
• Aerodynamic Design
• High Fidelity Simulation (TAU CFD)
• Static Aeroelasticity – Jig shape computation using CFD-CSM

Span loading – achieved vs. target


• Joint development of requirements
• Wing section design
• 3D wing shape design
• Design to meet lift & span-loading req.
• Evaluation of aerodynamic data
• Investigations into Pitch-up stability
• Trades performed on mass, c.g., twist (rigid trimmed)
• 20m/s 1g is the most challenging flight point
• Trailing edge flap reserved as risk mitigation option
• Assessment of Flexibility effects using CFD-CSM
• Ongoing work to investigate control surface aerodynamics

FLEXOP, UMN Seminar, Minneapolis, 09-12-2016


WP1 – FACC Manufacturing Aspects
– Starting point & approach: Common interface (4 attach. points) for all 3 wing types.
Decoupling of the wing root shear forces and the root bending moment at the interface
→ shear forces transferred between wings and fuselage, moment across the two wing
halves.
– Preliminary sizing of the interface has been completed - sizing of the laminate done by
means of Finite Element Models and sizing of the joints and metal fittings carried out by
hand calculations.
– Material design parameters provided for CFRP, GFRP and foam core material. This
included an own test campaign to establish the baseline material parameters (moduli,
strength) for the thin UD glass tape material.

Deformation W-1 (5g Load Case) Laminate RF W-1 (5g Load Case)
FLEXOP, UMN Seminar, Minneapolis, 09-12-2016
WP2 - DLR Math Modelling

Dynamic Model Aircraft Integration Process

•Nastran data set provided by TU Munich and DLR-AS


Data Sources

•Nastran model required some modifications


•Varloads data base for DLR-SR tools
Preprocessing •Rational function approximation of DLM results

•LTI state space system model integration


Integration

•Matlab state space system objects (delivered to SZTAKI)


Implementation

•Flutter analysis and comparison with TUM results


Application
•Estimation of achievable performance

FLEXOP, UMN Seminar, Minneapolis, 09-12-2016


WP2 - DLR Low Order Flutter Analysis

Flutter Mode Animation 1st model

FLEXOP, UMN Seminar, Minneapolis, 09-12-2016


WP2 – DLR/AGI-G Controllability Study

Flutter speed vs. required actuator bandwidth -> WP1

Investigate
controllability of
flutter mode taking
into account the
actuators selected
and tested in WP3

FLEXOP, UMN Seminar, Minneapolis, 09-12-2016


WP2 - UOB Bottom-Up Flutter Analysis

Emphasis on transfer to industry

Increasing modeling sophistication


and increasing number key parameters
for adequate transfer to industry

FLEXOP, UMN Seminar, Minneapolis, 09-12-2016


WP2 – UOB Flutter Mode Study

Emphasis on transfer to industry Two modern aspects of aircraft design


are investigated
HIGHLY FLEXIBLE AND ADVERSE GEOMETRY LAYOUT
TAIL VOLUME
STIFFNESS VARYING WINGS BENDING (FLIGHT MECHANICS DESIGN)
(AEROELASTIC DESIGN)
STIFFNESS
5m<D (tail distance)<10m
0.05<σ𝑠 <1

Lowest flutter speed calculated as the two parameters are varied

Body Freedom Flutter region Restrained flutter region

FLEXOP, UMN Seminar, Minneapolis, 09-12-2016


WP2 - UOB Uncertainty Analysis

Emphasis on transfer to industry Reconciliation of advanced analysis


with physical knowledge
and as a flutter sensitivity tool
NOMINAL PLANT UNCERTAIN PLANT

• σ𝑠 =0.15 (nominal bending stiffness defined) • LFT1 at V = 150 m/s


• D=6m (nominal tail distance defined) • Two uncertainty levels for (σs , σ𝐷)
𝑚
• BFF speed 175 (15 )
𝑟𝑎𝑑
• Structural (D, EI) and aerodynamic (A##)
𝑠 𝑠
𝑚 𝑟𝑎𝑑
• Restrained flutter speed 162 (70 )
𝑠 𝑠

High-frequency/
Restrained flutter
Low-frequency/
Body Freedom Flutter

FLEXOP, UMN Seminar, Minneapolis, 09-12-2016


WP2 – SZTAKI Control Design Toolchain

Flutter suppression control


Robust control synthesis
for the BAH wing

Aeroservoelastic modeling:

𝑊𝑓 = 0.2𝐼,
𝑊𝑑 = 1, 𝑊𝑢 (𝑠)
𝑊𝑛 = 0.001𝐼

Sigma plots of
the closed and
FEM based structural model + open loop
DLM aerodynamics + linear systems around
actuator = 115 state LPV model flutter freq. the flutter speed
scheduled with the airspeed
Sigma plots of
Flutter speed 343,11 m/s, 11 rad/s the controller
around the
Model order reduction: flutter speed

→ 23-state LPV model


FLEXOP, UMN Seminar, Minneapolis, 09-12-2016
WP2 – TUD NL Flight Dynamics

Aeroelastic analysis

Flight conditions Fuel and mass Static/dynamic


Skin and spar
Manoeuvre/gust cases aeroelastic stability
strains
and response

PROTEUS
Cross-sectional modeller

Aeroelastic analysis
- NL Timoshenko beam
- High-subsonic aero (continuous time)

FLEXOP, UMN Seminar, Minneapolis, 09-12-2016


WP2 – TUD NL Flight Dynamics for MC Analysis

Fast exploration method: example – convergence & comparison


5 5
10 10
Lift coefficient Moment coefficient
10 0 10 0

RMSE [%]
RMSE [%]

-5 -5
10 10

10 -10 10 -10
MT MT
POD POD
BPOD BPOD
BT BT
-15 -15
10 10
0 50 100 150 200 0 50 100 150 200
Modes Modes

Method (states) ROM [s] Simulation [s] Break even number

FOM (2030) - 6.2 -


POD (89) 66.1 0.012 11
BPOD (22) 61.8 0.005 10
BT (78) 631.3 0.012 102
FLEXOP, UMN Seminar, Minneapolis, 09-12-2016
WP3 - TUM Demonstrator Design

Efforts & Outcome:


• Fuselage Structural Design, Internal Layout & CAD Modelling

FLEXOP, UMN Seminar, Minneapolis, 09-12-2016


WP3 – TUM System Components

Efforts & Outcome:


• Propulsion System De-
tailed Design & Testing:

• Detailed Numerical Full Mission Simulation


• Mock-Ups of Airbrakes & Tank Compartment
• Engine Testing & Precise Numerical Modelling

FLEXOP, UMN Seminar, Minneapolis, 09-12-2016


WP3 – TUM Handling Qualities

Main Efforts & Outcome:


• Aircraft Handling Characteristics and Stability Investigations
2
1.8
1.6
1.4
CL(y)

1.2
1
0.8
0.6
0.4
0.2
Trimmed CL(y)_Inboard_Flaps 10° 2D CL_Max
0
0 500 1000 1500 2000 2500 3000 3500 4000

b/y[mm]
• Parachute Release

FLEXOP, UMN Seminar, Minneapolis, 09-12-2016


WP3 – SZTAKI Avionics HW/SW

Flight Control Architecture


• Develop a custom Flight control architecture including HW and
SW
• Minimal delay between the controller and physical components
• Distributed redundant flight control architecture
• High variety of I/O ports for custom sensors and actuators
Starting point
• Working flight control software stack from a previous project –
Sindy
• Working core of the RX-MUX (safety switch) from a previous
project

FLEXOP, UMN Seminar, Minneapolis, 09-12-2016


WP3 – SZTAKI Baseline Control

Baseline control concept


Flight modes
& Control actions
FM1 - IAS HOLD
- WP TRACK
- ALT-HOLD
FM2 - CHI-TRACK
- ALT-HOLD Test track
- ENG-SPEEDUP
FM3 - CHI-TRACK
- IAS-HOLD
- ALT-TRACK
- ACC, DEC
FM4 - ALT-HOLD UTC time
- CHI-TRACK
𝜙, 𝜃, 𝜓
- IAS HOLD
𝑝, 𝑞, 𝑟
𝑥𝐴 , 𝑦𝐴 , −ℎ
𝑣𝑁 , 𝑣𝐸 , 𝑣𝐷
𝛿𝑎1 , 𝛿𝑎2 , 𝛿𝑓 𝑎𝑥 , 𝑎𝑦 , 𝑎𝑧
ℎ, 𝛼, 𝛽
𝛿𝑟𝑅𝑈 , 𝛿𝑟𝐿𝑈 𝑝𝑠𝑡 , 𝑝𝑑𝑦𝑛
𝛿𝑡ℎ , 𝛿𝑠𝑏 𝐼𝐴𝑆
xSense,
Control surfaces reserved for flutter
FLEXOP,
suppression control UMN Seminar,
are marked Minneapolis,
by dark red. 09-12-2016 ADS
WP3 – AGI-G Servo Testing

Results
• Comparison of 2 actuators
HBL880 HBL599
Bandwidth < 20 Hz < 15 Hz
Static moment < 0,6 Nm < 4 Nm

Resonant No resonance
above 1e-4
kgm^2
Significant No dependency
dependency of on control
dynamic surface
behavior on rotational
control surface inertia
rotational
inertia

FLEXOP, UMN Seminar, Minneapolis, 09-12-2016


WP3 – INASCO Fibre Bragg Sensing

Objectives
• Providing a FBG (Fibre Bragg Grating) system is to monitor the strains and
forces on the wings during each flight tests

Schematic representation of the operational principle of FBGs

FLEXOP, UMN Seminar, Minneapolis, 09-12-2016


WP4 Objectives and Summary

• Validation of developed tools for flutter management and aero-elastic


tailoring using the flight test demonstrator
• Application of the developed methods and tools for design of a derivative
aircraft with extended wing-span (scale-up task) providing 7% fuel efficiency
improvement or 20% payload increase at a 50% reduction of development
and certification costs
• flutter testing:
– testing of natural flutter onset speed:
• oscillation measurement after excitation
• evaluation of damping progression for increasing airspeeds
• airspeeds below calculated natural flutter onset airspeed
– tuning of flutter control:
• see above with activated flutter suppression system, goal to exceed natural
flutter onset speed by 20%
• passive load alleviation:
– fly same manoeuvers flight with wing -0 and -2, record wing deformations and wing
loads
FLEXOP, UMN Seminar, Minneapolis, 09-12-2016
WP4 - RWTH Optimization Process

Scale-Up Optimization Process for Configuration Selection

FLEXOP, UMN Seminar, Minneapolis, 09-12-2016


WP4 – RWTH/AGI-G Baseline Aircraft

Scale-Up Configuration Optimization Preliminary Results


Span variable
LE-Sweep 32 °
TE-Sweep variable
Taper Ratio 0.30725
Twist 0°
Dihedral 0°

FLEXOP, UMN Seminar, Minneapolis, 09-12-2016


Status at 18 Months (+Deviations)

• Aircraft Concept is ready, but took slightly longer


– Different partners have different objectives for testing
– Few iterations for size, weight (+ budget )
• Design of -0 wings is merged with design of -2 (90% complete)
– To have similar mold for manufacturing the designs are coupled
– Investigation of -2 only makes sense if -0 is also optimized with state-of-
art techniques
– Definition of performance objectives and measurements is not trivial
• Design of -1 wings is much more challenging than expected (75% complete)
– Highly interdisciplinary task
– Modeling and interfacing problems
– a/c has to be relevant for scale-up
– Actuator performance is questionable
– Flutter tuning masses and retrofit of custom actuators
– Not well established process where the iterations are required
FLEXOP, UMN Seminar, Minneapolis, 09-12-2016
Status at 18 Months

• Flight Control is waiting for the models


– Model reduction and control techniques are developed
– Flutter analysis is done on several benchmark problems
– Baseline controller has to be tuned for the -0 a/c model, with flex effect
compensation
• Aircraft and wing manufacturing is on track, even though the wing design is delayed
– Aircraft systems are selected and custom components are designed (under testing)
– Aircraft overall operating and testing is established
– Wing and fuselage are ready for manufacturing (even with thin ply tech.)
• Scale-up study is better understood
– The goals of the ground/flight test are clarified
– The scale-up approach is harmonized with the proposal objectives (utilizing tools
and methodology)
• Project coordination is waiting for the new surprises @ 18 months Form C
– Successful GA, CA and Amendment
– Many physical and electronic meetings
– Workflows need constant adjustment due to interdisciplinary aspects

FLEXOP, UMN Seminar, Minneapolis, 09-12-2016


End & Thanks

FLEXOP, UMN Seminar, Minneapolis, 09-12-2016

S-ar putea să vă placă și