Airplane Dynamics, Modeling, and Control

Dr. Eugene A. Morelli

NASA Langley Research Center May 14, 1997

Overview
q q q q

General Airplane Dynamics Modeling for Control Design Control Design for Airplanes Demonstrations

Airplane Dynamics
q q

The Airplane is a Nonlinear Dynamical System Newtons 2nd Law for a Rigid Body Translational Motion : Rotational Motion :
r r p = F r r h = M

Assumptions
q q q q q

Earth is an inertial reference, no curvature Airplane is a rigid body with lateral symmetry Thrust acts along fuselage through the c.g. Still atmosphere (no winds, no gusts) Constant mass, no internal mass movements

Axis Systems
q

Equations written in body axes Fixed to the airplane, constant inertia Rotating axes nonlinear inertial terms
xb

yb Earth ye ze xe

zb

Nonlinear Equations of Motion

Translational motion of the c.g.
r r r r r r mV + mV = Faero + F prop + Fgravity

Rotational motion about the c.g.

r r r r I + I = Maero

Rotational kinematics
r r =L

States and Controls

r a e
v q u p r w

States and Controls

u V v w r V p p r r x = = q or q r r r th r e u= a r

Steady Flight
q q

Nonlinear Equations of Motion : Define steady state (trim) :

L

r r r r x = f ( x ,u )

r r r r 0 = f ( xo , uo )
un b sta le

T Mo W

D Lt
TRIM
sta ble

Modeling for Control Design

q

Define steady state (trim) : Linearize about trim :

r r r = [ f x ] r x x
o ,uo

r r r r 0 = f ( xo , uo )

r r r x + [ f u ] x r

o ,uo

r u

Linear model for analysis and control design :

r r r x = A x + Bu

Linear Models
Linearized model variables are perturbations Linearization naturally decouples dynamics
Longitudinal V r x= q Lateral / Directional p r x =r

r th u= e

r a u= r

Longitudinal Linear Equations

= ( XV + TV ) V + X g V = ZV V + Z + q q = MV V + M + Mq q =q + T th th + X e e + + Z e e M e e

Laplace Transform
= ( XV + TV ) V + X g sV +X + T th th e e

+ Z = ZV V + q s
+ M + Mq q sq = MV V =q s

+
+

Z e e
M e e

Computing Transfer Functions

s ( XV + TV ) ZV MV 0 X s Z M 0 0 1 s Mq 1 T th g V 0 0 = 0 0 q 0 s X e th Z e M e e 0

Computing Transfer Functions

s ( XV + TV ) ZV MV 0 = e s ( XV + TV ) ZV MV 0 X e Z e M e 0 X s Z M 0 0 1 s Mq 1 0 1 s Mq 1 g 0 0 s g 0 0 s

Modeling Example
Airplane : F-16 c.g. position : 0.2 c (fwd) 10,000 ft 350 kts Flight Condition : 5 AOA
Full Linear Model

2 2 0 . 19 [ s + 0 . 008 ] + 0 . 08 = e [ s + 0. 008] 2 + 0. 07 2 [ s + 1. 3] 2 + 2. 9 2

)(

Short Period Approx. 0.19 = e [ s + 1. 3] 2 + 2. 9 2

Why Feedback Control?

q q q

Modify plant dynamics Accurate regulation or tracking Overcome plant uncertainty

n

r + b

Controller

y Plant K SAS

Sensors

Airplane Control Tasks

q q

Stability Augmentation System (SAS) Control Augmentation System (CAS)

pitch rate command system g-load command system

Autopilots (pilot relief)

airspeed hold altitude hold heading hold turn coordination

Choosing Feedback Quantity

Stability Augmentation
q = M + Mq q + M e eSAS + e PILOT

eSAS = K q = M + K M e + Mq q 14 4 244 3 effective M

M e ePILOT

Stability Augmentation System (SAS)

ePILOT

+ eSAS

e
K

SAS Design Demonstration

Airplane : F-16 c.g. position : 0.2 c (fwd) 10,000 ft 350 kts Flight Condition : 5 AOA
Short Period Approx. 0.19 = e [ s + 1. 3] 2 + 2. 9 2

) )

Full Linear Model

c.g. position : 0.35 c (nom)

= e [ s + 0. 08] 2 + 0. 13 2 ( s + 1. 8)( s 0.1)

0.18 [s + 0 . 007] 2 + 0. 08 2

Choosing Feedback Quantity

Regulation or Tracking
Airplane V th

r = V DESIRED +

e Controller

th

Sensor = 0 to hold trim airspeed

Airspeed Hold Demonstration

Airplane : F-16 c.g. position : 0.2 c (fwd) 10,000 ft 350 kts Flight Condition : 5 AOA

Full Linear Model

2 0 . 17 [ s + 1 . 3 ] + 6.1 2 ( s + 0.8) V = th [ s + 0. 008] 2 + 0. 07 2 [ s + 1. 3] 2 + 2. 9 2

)(

Control System Design

q

Close feedback control loops

one at a time (classical control) many at once (modern control)

q q

Use several linear models

design points

Link individual designs (gain scheduling)

Practical Considerations
q

Control Effectiveness
Deflection limits High AOA Nonlinearity Actuator Dynamics

Time delay
Control surface rate limits Transport delay

Unmodeled effects q Pilot variability

q

Control Design
Nonlinear Airplane Dynamic Model Linear Design Models

Flight Test

Control Design

Piloted Nonlinear Simulation

Nonlinear Batch Simulation

Summary
q q q q q

General Airplane Dynamics Modeling for Control Design Control Design for Airplanes Demonstrations References for Further Study