Matlab/Simulink Tools for Teaching Flight Control Conceptual Design

Matlab/Simulink Tools for Teaching Flight Control Conceptual Design: An Integrated Approach Hanyo Vera Anders Tomas Melin Arthur Rizzi The Royal Institute of Technology, Sweden.

1

Presentation Outline • • • • • •

Computer Tools for Preliminary Aircraft Design QCARD Conceptual Design Tool Tornado Vortex Lattice Method CIFCAD Flight Simulator Case study: Student Project for Conceptual Design. Questions-Comments

2

Problems on Preliminary Aircraft Design •

The simplified methods used in the early phases of design do not give sufficient fidelity, which may result in mistakes which are costly to correct later in the design cycle.

•

Some examples pertaining to the Flight Control System are: DC-9: unexpected pitch-up and deep stall of T-tail lead to costly redesign DC-9-50 & MD-80: inadequate directional stiffness at high angles of attack in sideslip; adoption of low-set nose strakes SAAB2000: larger than expected wheel forces caused delay in certification; costly redesign of control system Boeing 777: missed horizontal tail effectiveness led to larger than needed horizontal tail 3

Computer Tools for Preliminary Aircraft Design •

There is work going on into the development of Computer Tools to facilitate the preliminary aircraft design process:

• QCARD • Tornado • SIFCAD Flight Simulator

4

QCARD: Quick Conceptual Aircraft Research & Development

5

QCARD in the Conceptual Design Process Design Requirements Business Case & Objectives Airworthiness Aircraft Morphology, Integration & Optimisation

Geometry Geometry

Sizing Sizing && Positioning Positioning of ofEmpennage Empennage Sizing, Sizing, Positioning Positioning&& Deflection Deflection of ofSurfaces Surfaces

Structures Structures

Weights Weights&& Balance Balance

Aerodynamics Aerodynamics

Propulsion Propulsion

Flight FlightControl Control System System

Mech/Elec Mech/Elec Systems, Systems, Avionics Avionics && Interiors Interiors

Loads Loads

Loadability Loadability&& Stability Stability Margins Margins

High-Lift High-Lift Design Design Philosophy Philosophy

Control Control Augmentation Augmentation due due to to Thrust Thrust Vectoring Vectoring

System SystemModes Modes &&Failure Failure Modes Modes

Ice IceProtection Protection Philosophy Philosophy

Low-speed Low-speed&& High-speed High-speed Design Design Philosophy Philosophy

Control Control Laws Laws && Protection Protection Functions Functions

Longitudinal, Longitudinal,Lateral Lateral && Directional Directional Static-Dynamic Static-Dynamic Stability Stability&&Control Control

Operational Operational Performance Performance && Economics Economics

Noise Noise&& Emissions Emissions

Critical assumptions made for sub-categories generates cross-disciplinary interaction at primary and secondary levels

Controllability & Manoeuvrability

6

Core Simulation Modules CAD geometry data

Iteration loop/feedback on design

Mesh generation Each partners’ CFD solver with moving mesh capabilities Flight state: •Motion and altitude •Trajectory Control surface state: •Position and motion

QCARD core: Comp Static/Dyn Derivativ, limited Aeroelast, buffet effects

QCARD analyzer •Stability margin •Empennage/ con sur. sizing •S&C modes, damping modes

ControllabilityManeuverability • Control surface effectiveness • Handling qualities • Pilot work load • etc

QCARD Environment Modules KTH working on two topics today: • Tornado (Dynamic Derivatives) • SIFCAD 7

Conceptual Prediction Methods: Stability & Control •

•

•

This discipline has lacked any form of sophistication & depth at the conceptual level – fundamental issues: controllability & manoeuvrability – tail volume method was adequate in the past; today, critical scenarios need to be identified & addressed early on Introduction of the Mitchell Code during sizing – original ICL FORTRAN code now converted to MATLAB – estimates: aero derivatives, moments of inertia, eigenvalues of motion equations, forced response and limiting speeds Assessing the suitability of design candidates – avoidance of esoteric figures of merit for uninitiated – extensive use of Cooper-Harper scale correlated with merit function plots, i.e. ESDU, MIL-Spec, ICAO, SAE, etc.

8

Sub-space Coupling & Process Logic until minimum goals achieved

Weights

Geometry

Performance Stability & Control

Propulsion

Aerodyn.

until minimum goals achieved

DOC/P-ROI & Optimal Techniques

9

Aerodynamic Coefficients: TORNADO • • • •

•

•

Developed by Tomas Melin, KTH. Vortex-Lattice Method. Implemented in Matlab Allows the analysis of complex geometry wings (swept, tapper, dihedral, tails,...) Different Flying Condition (Angles of Attack and Sideslip Angles, Roll, Tip and Yaw velocities) For wing-configuration, good results with projection of body along x-z and x-y planes.

10

TORNADO: Basic Assumption-Potential Flow • • • •

Inviscous Incompressible Irrotational Existence of Velocity potential

∇ × ∇φ = 0 ∇ φ =0 2

11

Tornado Implementation •

Sample Output

Tornado Computation Results ilona3 JID: Downwash matrix condition: 870034.5925 Reference area: 74.6078 Reference chord: 2.8041 Reference point pos: 2.8583 0 Reference span: 30.16 Net Wind Forces: (N) Drag: 3859.3759 Side: 35533.5133 Lift: 291384.6908

Net Body Forces: (N) X: -9538.6779 Y: 35533.5133 Z: 291284.6705

Delta cp distribution

0.58619 0

Net Body Moments: (Nm) Roll: 371527.6714 Pitch: -244497.527 Yaw: 516493.7084 -0.5

CL CD CY

0.2834 0.0037536 0.034559

CZ CX CC

0.2833 -0.0092772 0.034559

Cm Cn Cl

-0.084803 0.016656 0.011981

-10

-5

STATE: alpha: beta: Airspeed: Density:

3 0 150 1.225

P: Q: R:

0 0 0

0 0 Rudder setting [deg]: 0 0 0

0 0 0 0 0

5 0 5 0 4 0 30

0 0 0 0 0

-1

0

5 -1.5

2 1

10

0 -10

15 -5 -2

0 5 10

12

SIFCAD Flight Simulator 0.01

• OBJECTIVES:

-K-

Aileron1

0

Aileron_Com

Flap

0.5 States

-K-

Elevator Terminator

0.5

Controls

Rudder1

VelW

-K-

-0.7

R2D

Airspeed

R2D

Mach

0 Sideslip

Euler

-1

• Flight Control System Design. - Analysis of Handling Qualities. - Assessment of Mission Profile.

m

Ang Acc

Rudder_Com

Throttle1

0

Sensors

Elev_Com

Winds

0

AeroCoeff Mass

Throttle_Com

AOA

ECEF

0

MSL

Bank angle

AGL

Axes RST

Buttons joyinput

R2D

REarth

0

m

AConGnd

Pitch angle STOP

Point of View Aerosonde UAV

Joystick Input Axes

0

Stop Simulation when A/C on the ground

Heading m

Bank-angle to ailerons PI control 0

-K-

Auto_Aileron

Bank-angle-to-Aileron Proportional

0 -KAuto_Elev

0

0

Modifications: - Instalation of Turbofan Engine List of Blocksets: - Aerosim Blocket - Virtual Reality Toolbox - Aerospace Blockset

Airspeed

Bank angle Command

FS Interface

RAD DEG Airspeed to elevator PID control

RTCsim

-K-

Aileron

Euler

0

Bank-angle-to-Aileron Bank angle Integral Integrator

rad 2 deg

Elev

Position

1 s

RAD DEG rad 2 deg1

Airspeed-to-Elevator Proportional -K-

Time Error in msec

Real Time Control 25 1 s

Airspeed Command

-KGain

Simulation Time in sec

Airspeed-to-Elevator Airspeed error Integral Integrator -K-

du/dt

Simulation sample time 10 ms Simulation time: 5 min.

Airspeed-to-Elevator Airspeed error Derivative Derivative

13

SIFCAD: Characteristics • Flight Simulator in Simulink Environment • Based on commercially available Simulink Toolboxes • Graphics provided by Microsoft Flight Simulator • Highly Flexible and easily customable (Simulink format) • Options: Fast-time or Real-time. 14

Simulink Toolboxes: •

•

•

Aerospace Blockset - Mathworks - Aerodynamic - Engine, - Earth and Atmosphere models. Virtual Reality Toolbox Mathworks - Man-Machine interface i.e. Joysticks) AeroSim Blockset – Unmanned Dynamics - Aerodynamic, - Engine, - Earth and Atmosphere models

15

Simulink Toolboxes: •

•

Flight Dynamic and Control Blocket - M.O. Rauw, Netherlands. - Aerodynamic, - Engine, - Earth and Atmosphere models - Avionics. Port and Memory IO for Matlab and Simulink – Werner Zimmermann, FHT Esslingen - Real time execution in Matlab Environment.

16

Interface with Microsoft Flight Simulator • •

•

•

Use of interface provided by AeroSim Blockset Possibility to send information to a Second Computer Running Microsoft Flight Simulator Information sent involves position, attitude and gauges information. The result is high quality graphic interface without the need of extensive programming.

17

Use of Simulator •

Simulator Running at FastTime: -

•

Airplane Model development FCS development and testing Autpilot testing Mission profile Assesment

Real Time Simulation: -

Handling Qualities Assesment Pilot-in-the-loop analysis Research in Aircraft Operational Factors Research in Human Factors.

18

Case Study: Horizon Project •

Conceptual Design Student Project in The Royal Institute of Technology in collaboration with Ecole Polytechnique of Montréal

•

Objective: -

Analysis of 70 PAX regional airliner Unducted Fan Able to achieve speeds close to Turbofan

19

Case Study: Horizon Project •

Procedure: Use of QCARD in the conceptual design process: -

•

Estimation of Low Speed Aerodynamic Properties Estimation of High Speed Aerodynamic Properties Stability and Control Analysis

Conclusion: The initial design has poor stability qualities. Need to improve the design to reach reasonable stability characteristics.

20

Geometric Modifications •

Wing: -

•

Horizontal Tail: -

•

Moved Forward Increased Area Reduced Aspect Ratio

Lowered Increased Area Increaed Aspect Ratio

Vertical Tail: -

Reduced Area.

21

Results of the Modifications Type of motion

Name

Period (s)

Time-to-half (s)

Cycles-to-half

Initial

Impro ved

Initial

Impr oved

Initial

Impro ved

Phugoid

1.76e5

1.81e6

94.98

95.9

5.4e-4

5.3e-4

Short-period

4.35

2.92

0.82

0.58

0.19

0.20

Dutch roll

5.11

5.40

12.94

6.84

2.52

1.26

Spiral

128.4

117.5

Rolling convergence

0.86

0.89

Longitudinal

Lateral

22

Stability and Control Analysis

23

Stability and Control Analysis

24

Stability and Control Analysis

25

Stability and Control Analysis

26

Possibilities of using SIFCAD in Horizon Project • • • • • •

•

Higher understanding of criteria for stability. Analysis of airplane handling with Pilot-in-the-loop. Possibilities of considering relaxed stability in design. Design of Flight Control Systems. Mission Profile Analysis. Response to medium and heavy weather phenomen (i.e. Gusts, windshear, etc.) Explore operational profile (take-off, approach, landing)

27

SIFCAD Demo: Horizon Model

28

SIFCAD: Future Goals •

SIFCAD in Aeronautic Education -

-

-

Teach to student effects of changes in Aerodynamic Coefficients in Airplane Handling Effects of Center of Gravity in Aerodynamic Stability. Suitable for teaching Concepts on Flight Control System Design Suitable for practical examples in Avionics use and Limitations.

29

SIFCAD: Future Goals •

Full Integration with QCARD software: -

-

•

Automatic Load of Aerodynamic, engine and mass properties onto the Simulator Model. Requisite for integration on the Conceptual Design package.

Use in research: -

Airplane Design Operations Human Factor Etc.

30

SIFCAD: Ongoing and Future Work. •

Improve the Aerodynamic Model -

Possibility to manage non-linear aerodynamic phenomena.

•

Develop a stable simulation platform with ”best of commercial packages” plus native development.

•

Improve Human-Machine interface: -

Projectors Glass-Cockpit Improved Joysticks Pedals

31

Questions – Comments?

32