Você está na página 1de 16

Aircraft Control Devices

and Systems

Design for Control

Robert Stengel, Aircraft Flight Dynamics, MAE 331,


2010

Control surfaces
Control mechanisms
Flight control systems

Elevator/stabilator: pitch control


Rudder: yaw control
Ailerons: roll control
Trailing-edge flaps: low-angle lift control

Leading-edge flaps/slats: High-angle


lift control
Spoilers: Roll, lift, and drag control
Thrust: speed/altitude control

Copyright 2010 by Robert Stengel. All rights reserved. For educational use only.
http://www.princeton.edu/~stengel/MAE331.html
http://www.princeton.edu/~stengel/FlightDynamics.html

Critical Issues for Control

Effect of control surface deflections on aircraft motions

Generation of control forces and rigid-body moments on the aircraft


Rigid-body dynamics of the aircraft
!E is an input

!!! = "

Command and control of the control surfaces

Displacements, forces, and hinge moments of the control mechanisms


Dynamics of control linkages
!E is a state

! E!! = "

Control Surface Dynamics


and Aerodynamics

Aerodynamic and
Mechanical Moments
on Control Surfaces

Angle of Attack and


Control Surface Deflection

Increasing size and speed of aircraft leads to


increased hinge moments
This leads to need for mechanical or aerodynamic
reduction of hinge moments
Need for aerodynamically balanced surfaces
Control surface hinge moment

H surface = C H surface

1 2
!V Sc
2

Hinge-moment coefficient, CH
Linear model of dynamic effects

C H surface = C H!! !! + C H! ! + C H" " + C H command + ...

Horizontal tail with


elevator control surface

Horizontal tail at positive


angle of attack

Horizontal tail with positive


elevator deflection

C H!!E : aerodynamic/mechanical damping moment


C H! E : aerodynamic/mechanical spring moment
C H" : floating tendency
C H command : pilot or autopilot input

Dynamic Model of a Control


Surface Mechanism

Floating and Restoring


Moments on a Control Surface

Positive angle of attack typically produces a negative moment on the elevator


With stick free, i.e., no opposing torques, elevator floats up due to negative H#

Approximate control dynamics


by a 2nd-order LTI system

Bring all torques and


inertias to right side

!!! " H !!!! " H ! ! = H # # + H command + ...


mechanism dynamics = external forcing

1 2
H elevator C H elevator 2 "V Sc
!!
!E =
=
I elevator
I elevator
1 2
"V Sc
!
$
&
= %C H!!E ! E + C H! E ! E + C H# # + C H command + ...' 2
I

Positive elevator deflection typically produces a negative (restoring) moment, H!" , on


the elevator due to aerodynamic or mechanical spring

elevator

( H !! E!! E + H ! E! E + H # # + H command + ...

Stability and control


derivatives of the control
mechanism

I elevator = effective inertia of surface, linkages, etc.

" ( H elevator I elevator )


" ( H elevator I elevator )
; H! E =
"!
"!!
" ( H elevator I elevator )
H# =
"#

H !! E =

Coupling of System
Model and Control
Mechanism Dynamics

Augmented Short Period Model


State Vector

Second-order model of control-deflection dynamics

!!x SP /" E = FSP /" E !x SP /" E + G SP /" E !" Ecommand

"E
!!x" E = F" E !x" E + G" E !u" E + FSP
!x SP

# !"! E & # 0
%
()%
%$ !"!!E (' %$ H " E

Augmented dynamic equation

# 0
1 & # !" E & # 0 &
!" Ecommand + %
(% ! ( + %
(
H "! E ( % !" E ( % *H " E (
%$ H q
' $
'
'$

0 & # !q &
(
H + ( %% !+ ((
'
'$

$ Mq
$ !q! ' &
)*&
&
%& !#! )( & 1
%

'
$ M"E
) $ !q ' &
) & !# ) + & + L" E
)( &
VN
) &%
%
(

M#
+

L#

VN

FSP /! E

0 '$
) !" E '
)
&
0 ) &% !"! E )(
)(

" F
SP
= $ !E
$ FSP
#

SP
!E

F! E

"
$
% $
'=$
' $
& $
$
#

Mq

M(

M!E

L(

L! E

0
Hq

! SP /" E ( s ) = sI n # FSP /" E =

#1

VN

0
H! E

1
H !! E

C H ! C H" " + C H# E # E + C H pilot input


#M $
s+

L$

VN

#M " E

L" E

VN

#H q

#H $

#H " E

Stick-free case

Control surface free to float

C H ! C H" " + C H# E # E

=0

#1

(s # H )

"! E

Normally

C H! < 0 : reduces short-period stability

! SP /" E ( s ) = s + 2# SP$ nSP s + $

2
nSP

)( s

+ 2#" E$ n" E s + $

2
n" E

C H" E < 0 : required for mechanical stability

Coupling of the modes


depends on values of

M!E ,

L! E

Root locus evaluation of design parameters

Inertial and aerodynamic effects


Control surface in front of hinge line

VN , H q , and H "

Increasing C H! improves pitch stability,


to a point

Too much horn area

Assuming high aero/mechanical


damping of the control mechanism

%
'
'
'
'
'
'
&

G SP /! E

" 0
$
0
=$
$ 0
$ H
# !E

%
'
'
'
'
&

Horn Balance

Characteristic equation

(s # M )

VN

0
H(

Roots of the Augmented


Short Period Model

!q '
)
)
)
)
)(

!"
!# E
!#! E

Augmented stability and control matrices

Short period approximation

!!x SP = FSP !x SP + G SP !u SP = FSP !x SP + F"SPE !x" E

$
&
!x SP ' = &
&
&
&%

Degrades restoring moment


Increases possibility of mechanical
instability
Increases possibility of destabilizing
coupling to short-period mode

NACA TR-927, 1948

Control Flap Carryover Effect on


Lift Produced By Total Surface

Overhang or
Leading-Edge
Balance

from Schlichting & Truckenbrodt

Effect is similar to
that of horn balance
Varying gap and
protrusion into
airstream with
deflection angle

C L! E
C L"

vs.

cf
xf + cf

C H ! C H" " + C H# # + C H pilot input


cf

(x

+ cf )

NACA TR-927, 1948

All-Moving Control Surfaces


Shorts SB.4

Particularly effective at supersonic speed (Boeing


Bomarc wing tips, North American X-15 horizontal
and vertical tails, Grumman F-14 horizontal tail)
SB.4!s aero-isoclinic wing
Sometimes used for trim only (e.g., Lockheed L-1011
horizontal tail)
Hinge moment variations with flight condition

Lockheed L-1011

North American X-15

Grumman F-14

Boeing
Bomarc

Control Surface Types

Elevator

Ailerons

Effectiveness of high
mounting is unaffected by
wing downwash at low to
moderate angle of attack

When one aileron goes up, the other goes down


Average hinge moment affects stick force

Effectiveness of low mounting is


unaffected by wing downwash at
high angle of attack

Spoilers

Compensating Ailerons

Asymmetric contour, with hinge line at or below lower


aerodynamic surface
Reduces hinge moment

Spoiler reduces lift, increases drag

Differential spoilers

Frise aileron

Speed control
Roll control
Avoid twist produced by outboard
ailerons on long, slender wings
free trailing edge for larger high-lift
flaps

Cross-coupling effects can be adverse or favorable, e.g.


yaw rate with roll
Up travel of one > down travel of other to control yaw
effect

Plug-slot spoiler on P-61 Black


Widow: low control force
Hinged flap has high hinge moment
Abzug & Larrabee, 2002

North American P-61

Abzug & Larrabee, 2002

Elevons

Canards

General Dynamics F-106

Pitch control
Ahead of wing downwash
High angle of attack effectiveness
Desirable flying qualities effect
(TBD)

Combined pitch and yaw control using


symmetric and asymmetric surface
deflection
Principally used on

Dassault Rafale

Delta-wing configurations
Swing-wing aircraft
SAAB Gripen

Grumman F-14

Yaw Control of Tailless


Configurations

Typically unstable in pitch and yaw


Dependent on flight control system
for stability
Split ailerons or differential drag flaps
produce yawing moment
Northrop Grumman B-2

Rudder

Rudder provides yaw control

McDonnell Douglas X-36

Martin B-57

Turn coordination
Countering adverse yaw
Crosswind correction
Countering yaw due to engine loss

Strong rolling effect, particularly at high #


Only control surface whose nominal
aerodynamic angle is zero
Possible nonlinear effect at low deflection
angle
Insensitivity at high supersonic speed

Wedge shape, all-moving surface on North


American X-15

Bell X-2

Control Tabs

Balancing or geared tabs


Tab is linked to the main surface in
opposition to control motion, reducing
the hinge moment with little change in
control effect

Flying tabs
Pilot's controls affect only the tab,
whose hinge moment moves the
control surface [BAC 1-11 deep stall
flight test accident]

Control Mechanization
Effects

Linked tabs
divide pilot's input between tab and
main surface

Spring tabs
put a spring in the link to the main
surface
BAC 1-11

Control Mechanization Effects


Fabric-covered control
surfaces (e.g., DC-3, Spitfire)
subject to distortion under air
loads, changing stability and
control characteristics
Control cable stretching
Elasticity of the airframe
changes cable/pushrod
geometry
Nonlinear control effects
friction
breakout forces
backlash

Douglas DC-3

Nonlinear Control Mechanism Effects


Friction
Deadzone

Supermarine
Spitfire

Instabilities Due To
Control Mechanization

Control Mechanization Effects


Breakout force
Force threshold

Rudder Snaking

Control-free dynamics

Nominally symmetric control position


Internal friction
Aerodynamic imbalance

Coupling of mechanical motion with


Dutch roll mode (see Flight Dynamics)
Solutions

Trailing-edge bevel
Flat-sided surfaces
Fully powered controls

Douglas DC-2

Aileron buzz (aero-mechanical instability; P-80)


Rudder snaking (Dutch roll/mechanical coupling; Meteor, He-162, X-1)
Aeroelastic coupling (B-47, Boeing 707 yaw dampers)

Roll/Spiral Limit Cycle


Due to Aileron
Imbalance

Control Surface
Buzz

At transonic speed,
normal shocks may
occur on control surface

Rudder Lock

North American FJ-4

Low directional stability at high


sideslip due to stalling of fin
High (positive) hinge momentdue-to-sideslip at high sideslip
(e.g., B-26)
Negative rudder yawing moment

With deflection, shocks


move differentially
Possibility of selfsustained nonlinear
oscillation (limit cycle)

Solutions

Rudder deflected to stops at high


sideslip; aircraft trims at high $
3 necessary ingredients

Splitter-plate rudder
fixes shock location for
small deflections
Blunt trailing edge
Fully powered controls
with actuators at the
surfaces

Problematical if rudder is
unpowered and requires high
foot-pedal force (rudder float
of large WWII aircraft)
Solutions
Increase high-sideslip directional
stability by adding a dorsal fin
(e.g., B-737-100 (before), B-737400 (after))
Hydraulically powered rudder

Martin B-26

Boeing 737-100

Boeing 737-400

ARC R&M 3364

Downsprings and Bobweights

Control Systems

Adjustment of trim and stickforce sensitivity to airspeed*


Downspring
Long mechanical spring with
low spring constant
Exerts a trailing-edge down
moment on the elevator

Bobweight
Weight on control column that
affects feel or basic stability
Mechanical stability
augmentation

* See pp. 541-545, Section 5.5, Flight Dynamics

Beechcraft B-18

Mechanical and Augmented


Control Systems
Mechanical system

Mechanical, Power-Boosted Systems


McDonnell Douglas F-15
Grumman A-6

Push rods, bellcranks, cables, pulleys

Power boost
Pilot's input augmented by hydraulic servo that
lowers manual force

Fully powered (irreversible) system


No direct mechanical path from pilot to
controls
Mechanical linkages from cockpit controls to
servo actuators

Boeing 767 Elevator Control System

Abzug & Larrabee, 2002

Classical Lateral Control Logic for


a Fighter Aircraft (c.1970)

MIL-DTL-9490E, Flight Control Systems - Design, Installation and Test


of Piloted Aircraft, General Specification for, 22 April 2008
Superseded for new designs on same date by
SAE-AS94900
http://www.sae.org/servlets/works/documentHome.do?comtID=TEAA6A3&docID=AS94900&inputPage=dOcDeTaIlS

Effect of Scalar Feedback Control


on Roots of the System

Closed-Loop Transfer Function


!y(s) = K

kn(s)
kn(s)
!yc (s) " K
!y(s)
d(s)
d(s)

Block diagram algebra


!y(s) = H (s)!u(s) =

=K

kn(s)
kn(s)
!u(s) =
K !" (s)
d(s)
d(s)

!y(s) + K

kn(s)
kn(s)
kn(s)
!yc (s) " !y(s)] = K
!yc (s) " K
!y(s)
[
d(s)
d(s)
d(s)
Aircraft Transfer Function (open-loop): H (s) =

kn(s)
K
!y(s)
d(s)
=
kn(s) %
!yc (s) "
$1 + K d(s) '
#
&

kn(s)
d(s)

Control System Gain : K

Roots of the Closed-Loop System

kn(s)
kn(s)
!y(s) = K
!yc (s)
d(s)
d(s)

Root Locus Analysis of Pitch Rate Feedback to


Elevator (2nd-Order Approximation)

kn(s)
!y(s)
Kkn(s)
Kkn(s)
d(s)
=
=
=
kn(s) % [ d(s) + Kkn(s)] ! closed ( s )
!yc (s) "
loop
$1 + K d(s) '
#
&
K

Closed-loop roots are solutions to


! closed (s) = d(s) + Kkn(s) = 0
loop

or

kn(s)
= !1
d(s)

k q s # zq
kn(s)
!q(s)
=K
=K 2
= #1
s + 2$ SP% nSP s + % n2SP
d(s)
!" E(s)

Number of roots = 2
Number of zeros = 1
Destinations of roots (for k = "):
1 root goes to the zero of n(s)
1 root goes to infinite radius from the origin

Angles of asymptotes, %, for the


roots going to "
K -> +": 180 deg
K -> ": 0 deg

Root Locus Analysis of Pitch


Rate Feedback to Elevator
(2nd-Order Approximation)
K

Root Locus Analysis of Angular


Feedback to Elevator (4th-Order Model)*

k q s # zq
kn(s)
!q(s)
=K
=K 2
= #1
s + 2$ SP% nSP s + % n2SP
d(s)
!" E(s)
Flight Path Angle

Center of gravity : doesn!t matter

Locus on real axis

Pitch Rate

K > 0: Segment to the left of the zero


K < 0: Segment to the right of the zero
Pitch Angle

The feedback effect is analogous


to changing Mq

Angle of Attack

* p. 524, Flight Dynamics

Direct-Lift Control-Approach
Power Compensation
Vought F-8

F-8 Crusader
Variable-incidence wing to
provide better pilot visibility
Flight path control at low
approach speeds requires throttle
use, could not be made with pitch
control alone
Engine response time is slow
Flight test of direct lift control
(DLC), using the ailerons as flaps

Direct Lift and


Propulsion Control

Approach power compensation


for A-7 Corsair II and direct lift
control were studied at Princeton
using our Variable-Response
Research Aircraft

Vought A-7

Princeton VRA

Direct-Lift/Drag Control

Direct-lift control on S-3A Viking


Implemented with spoilers
Rigged up during landing to
allow lift.

Speed brakes on T-45A Goshawk


make up for slow spool-up time
of jet engine
BAE Hawk's speed brake moved
to sides for carrier landing
Idle speed increased from 55% to
78% to allow more effective
modulation via speed brakes

Lockheed S-3A

Next Time:
Flight Testing

Boeing T-45

Trailing-Edge
Bevel Balance

Supplementary
Material

See discussion in
Abzug and Larrabee

C H ! C H" " + C H# # + C H pilot input

Aft Flap vs. All-Moving


Control Surface

The Unpowered F4D Rudder

Rudder not a problem under normal flight conditions

Not a factor for upright spin

However, in an inverted spin

Carryover effect

Aft-flap deflection can be almost as effective as


full surface deflection at subsonic speeds
Negligible at supersonic speed

Single-engine, delta-wing aircraft requiring small rudder inputs


Rudder was ineffectual, shielded from flow by the large delta wing
rudder effectiveness was high
floating tendency deflected rudder in a pro-spin direction
300 lb of pedal force to neutralize the rudder

Fortunately, the test aircraft had a spin chute

Aft flap
Mass and inertia lower, reducing likelihood of
mechanical instability
Aerodynamic hinge moment is lower
Can be mounted on structurally rigid main
surface

Powered Flight Control Systems

Early powered systems had a single


powered channel, with mechanical
backup

A4D

Reversion typically could not be


undone
Gearing change between control stick
and control to produce acceptable pilot
load
Flying qualities changed during a highstress event

Hydraulic system failure was common

Alternative to eject in military aircraft

Redundancy was needed

Artificial-feel system
Restores control forces to those of an
"honest" airplane
"q-feel" modifies force gradient
Variation with trim stabilizer angle
Bobweight responds to gravity and to
normal acceleration

Pilot-initiated reversion to
"conventional" manual controls
Flying qualities with manual control
often unacceptable

Advanced Control Systems

A3D

B-47

Fly-by-wire/light system
Minimal mechanical runs
Command input and feedback signals
drive servo actuators
Fully powered systems
Move from hydraulic to electric power

Control-Configured Vehicles

Boeing 777 Fly-By-Wire Control System

Command/stability augmentation
Lateral-directional response

USAF F-15 IFCS

Bank without turn


Turn without bank
Yaw without lateral translation
Lateral translation without yaw
Velocity-axis roll (i.e., bank)

Longitudinal response

Pitch without heave


Heave without pitch
Normal load factor
Pitch-command/attitude-hold
Flight path angle
Princeton Variable-Response Research Aircraft

USAF AFTI/F-16

United Flight 232, DC-10


Sioux City, IA, 1989

Root Locus Analysis of Angular


Feedback to Thrust (4th-Order Model)

Flight Path Angle

Pitch Angle

Uncontained engine failure damaged all three flight control


hydraulic systems (http://en.wikipedia.org/wiki/United_Airlines_Flight_232)

Pitch Rate

Angle of Attack

296 people onboard


185 survived due to pilot!s
differential control of engines
http://www.youtube.com/watch?v=vTXr1QR3rbQ

Propulsion Controlled Aircraft

Proposed backup attitude control in event of flight control system failure


Differential throttling of engines to produce control moments
Requires feedback control for satisfactory flying qualities

Proposed retrofit to McDonnellDouglas (Boeing) C-17

NASA MD-11 PCA Flight Test

NASA F-15 PCA Flight Test