# Nonlinear Analysis of Aircraft Loss-of- hgk22/papers/AIAAJnlGCD_2013.pdf Nonlinear Analysis of...

date post

13-Jul-2018Category

## Documents

view

214download

0

Embed Size (px)

### Transcript of Nonlinear Analysis of Aircraft Loss-of- hgk22/papers/AIAAJnlGCD_2013.pdf Nonlinear Analysis of...

Nonlinear Analysis of Aircraft Loss-of-Control

Harry G. Kwatny1 and Jean-Etienne T. Dongmo2 and Bor-Chin Chang3

Drexel University, Philadelphia, PA, 19104, USA

Gaurav Bajpai4 and Murat Yasar5

Techno-Sciences, Inc., Beltsville, MD, 20705, USA

Christine Belcastro6

NASA Langley Research Center, Hampton, VA, 23681, USA

Loss-of-Control (LOC) is a major factor in fatal aircraft accidents. Although de-

nitions of LOC remain vague in analytical terms, it is generally associated with ight

outside of the normal ight envelope, with nonlinear inuences, and with a signicantly

diminished capability of the pilot to control the aircraft. Primary sources of nonlin-

earity are the intrinsic nonlinear dynamics of the aircraft and the state and control

constraints within which the aircraft must operate. This paper examines how these

nonlinearities aect the ability to control the aircraft and how they may contribute

to loss-of-control. Specically, the ability to regulate an aircraft around stall points is

considered, as is the question of how damage to control eectors impacts the capability

to remain within an acceptable envelope and to maneuver within it. It is shown that

even when a sucient set of steady motions exist, the ability to regulate around them

or transition between them can be dicult and nonintuitive, particularly for impaired

aircraft. Examples are provided using NASA's Generic Transport Model.

1 S. Herbert Raynes Professor, MEM Department, Drexel University, 3141 Chestnut Street, Philadelphia, PA, 19104,AIAA Member

2 Research Associate, MEM Department, Drexel University, 3141, Chestnut Street, Philadelphia, PA, 19104, AIAAMember

3 Professor, MEM Department, Drexel University, 3141, Chestnut Street, Philadelphia, PA, 19104 ,AIAA Member4 Director, Dynamics and Control, Techno-Sciences, Inc., 11750 Beltsville Road, Beltsville, MD, 20705, AIAA Mem-ber

5 Research Engineer, Dynamics and Control, Techno-Sciences, Inc., 11750 Beltsville Road, Beltsville, MD, 20705,AIAA Member

6 Senior Research Engineer, NASA Langley Research Center, MS 308, Hampton, VA, 23681, AIAA Associate Fellow

1

Nomenclature

x State vector

y Measurements

z Regulated variables

Bifurcation parameter

u Control inputs

Angle of attack, rad

Side slip angle, rad

V Airspeed, ft/s

X x inertial coordinate, ft

Y y inertial coordinate, ft

Z z inertial coordinate, ft

p x body-axis angular velocity component, rad/s

q y body-axis angular velocity component, rad/s

r z body-axis angular velocity component, rad/s

u x body-axis translational velocity component, ft/s

v y body-axis translational velocity component, ft/s

w z body-axis translational velocity component, ft/s

Euler roll angle, rad

Euler pitch angle, rad

Euler yaw angle, rad

Heading, rad

Flight path angle, rad

p Quasi-velocity vector

q Generalized coordinate vector

V Kinematic matrix

M Inertia matrix

C Gyroscopic matrix

2

F Generalized force vector

CX Aerodynamic coecient, force x body axis

CY Aerodynamic coecient, force y body axis

CZ Aerodynamic coecient, force z body axis

CL Aerodynamic coecient, moment about x body-axis

CM Aerodynamic coecient, moment about y body-axis

CN Aerodynamic coecient, moment about z body-axis

CD Aerodynamic drag coecient, force x wind-axis

CL Aerodynamic lift coecient, force z wind-axis

c Mean aerodynamic cord, ft

S Reference wing area, ft2

xcg Center of mass location body x coordinate, ft

xcgrefCenter of mass reference location body x coordinate, ft

Quasi-static approximation for angle of attack

A System matrix of a linear system

B Control matrix of a linear system

C Output matrix of a linear system

J Jacobian matrix of a multivariable vector function

, v Eigenvalue and eigenvector of the Jacobian matrix J

T Thrust, lb

lt Engine location, z-coordinate, ft

M Aerodynamic moment about y body axis, lbf-ft

m Mass, slugs

Ij Inertia parameters about j-axis, j=x,y,z

g Gravitational constant

air density, lb/ft3

e Elevator deection, rad

a Aileron deection, rad

3

r Rudder deection, rad

min Minimum singular value

X State space

C Envelope

U Control constraint set

S Safe set

T Trim manifold

I. Introduction

Recent published data shows that during the ten year period 1997-2006, 59% of fatal aircraft

accidents were associated with Loss-of-Control (LOC) [1]. Yet the notion of loss-of-control is

not well-dened in terms suitable for rigorous control systems analysis. The importance of LOC

is emphasized in [2] where the inadequacy of current denitions is also noted. On the other hand,

ight trajectories have been successfully analyzed in terms of a set of ve two-parameter envelopes to

classify aircraft incidents as LOC [3]. As noted in that work, LOC is ordinarily associated with ight

outside of the normal ight envelope, with nonlinear behaviors, and with an inability of the pilot

to control the aircraft. The results in [3] provide a means for analyzing accident data to establish

whether or not the accident should be classied as LOC. Moreover, they help identify when the

initial upset occurred, and when control was lost. The analysis also suggests which variables were

involved, thereby providing clues as to the underlying mechanism of upset. However, it does not

provide direct links to the ight mechanics of the aircraft, so it cannot be used proactively to identify

weaknesses or limitations in the aircraft or its control systems. Moreover, it does not explain how

departures from controlled ight occur. In particular, we would like to know how environmental

conditions (like icing) or faults (like a jammed surface or structural damage) impact the vulnerability

of the aircraft to LOC.

LOC is essentially connected to the nonlinearity of the ight control problem. Nonlinearity

arises in two ways: 1) the intrinsic nonlinearity of the aircraft dynamics, and 2) through state and

4

control constraints. This paper considers control issues that arise from both sources.

First, the implications of the nonlinear aircraft dynamics are considered. Bifurcation analysis

is used to study aircraft control properties and how they change with the ight condition and

parameters of the aircraft. The paper extends results previously introduced in [4, 5]. There it

was shown that the ability to regulate a system is lost at points associated with bifurcation of

the trim equations; ordinarily indicating stall in an aircraft. Such a bifurcation point is always

associated with a degeneracy of the zero structure of the system linearization at the bifurcation

point. Such degeneracies include loss of (linear) controllability or observability, redundant controls

(rank degeneracy of the B matrix) and/or redundant outputs (rank degeneracy of the C matrix). As

such points are approached, the ability to regulate degrades so that the performance of the regulator

(or pilot) may deteriorate before the bifurcation point is actually reached. The equilibrium surface

or set of trim conditions is a submanifold of the state-control-parameter space that is divided into

open sets by the bifurcation points. Within each region a linear regulator can be designed. However,

a regulator designed in one region will fail if applied in a neighboring region [6]. The key implication

of this result is that at the boundary of these sets, i.e., near stall bifurcation points, the strategy

required for regulating the aircraft is super-sensitive to parameter variations. Accordingly, we say

that the property of regulation is structurally unstable at bifurcation points.

Second, the question of how state and control constraints relate to LOC is considered. The

Commercial Aviation Safety Team (CAST) denes in-ight LOC as a signicant deviation of the

aircraft from the intended ight path or operational envelope [7]. The ight envelope represents a set

of state constraints, so the control issues associated with preventing departure from the constraint

set is considered. The notion of a safe set [8] or viable set [9] is central to this analysis. Suppose an

acceptable operating envelope is specied as a domain C in the state space. The idea of a safe set

derives from a decades old control problem in which the plant controls are restricted to a bounded

set U and it is desired to keep the system state within a convex, not necessarily bounded, subset C

of the state space. Feuer and Heyman [10] studied the question: under what conditions does there

exist for each initial state in C an admissible control producing a trajectory that remains in C for

all t > 0? When C does not have this property we try to identify the safe set, S, that is, the largest

5

subset of C that does. Clearly, if if it is desired that the aircraft remain in C, it must be insured

that it remains in S.

The safe set S is the largest positively controlled-invariant set contained in C. Safe set theory

could be used as a basis for design of envelope protection systems, but this idea has not been fully

developed. It is also important to know the extent to which the aircraft can maneuver within

S. Controlled ight requires the existence of a suitable set of steady motions and the ability to

smoot

Recommended

*View more*