Handling qualities of large flexible control-configured aircraft

Document

General Disclaimer

One or more of the Following Statements may affect this Document

This document has been reproduced from the best copy furnished by the

organizational source. It is being released in the interest of making available as

much information as possible.

This document may contain data, which exceeds the sheet parameters. It was

furnished in this condition by the organizational source and is the best copy

available.

This document may contain tone-on-tone or color graphs, charts and/or pictures, which have been reproduced in black and white.

This document is paginated as submitted by the original source.

Portions of this document are not fully legible due to the historical nature of some

of the material. However, it is the best reproduction available from the original

submission.

Produced by the NASA Center for Aerospace Information (CASI)

FROM: Robert 1 . Swain, Associate Dean , Tech. b Arch.

Div. of E Vorth 111 Engineer' • University Oklahoma OK 74074 Stillwate HANDLING QUALITIES OF LARGE FLEXIBLE CONTROL-CONFIGURED AIRCRAFT

(NASA -( ' P-158694) NANELING QUALITIES OF N79-25033

F'XIBLF, -CONFIGURED AIRCRAFT LARGE FI . CONTROL SPUiannual Status Report, 1 Jan. - 30 .Jun.

1979 ( Oklahona State 1 Iniv., Stillwater.) 'Inclds

10 p HC A02/MF 401 CSCL 01L G3/08 22263

Grant No. NSG 4018 Semi-Annual Status Report for Period January 1, 1979 - .June 30, 1919 J Date of Report: June 22, 1979 Principal Investigator: Dr. Robert L. Swaim Associate Dean Division of Engineering, Te r h- nology and Architecture Oklahoma State University 111 Engineering North Stillwater', OK 74074 NASA Technical Officer: Mr. Glenn 6. lyard Vehicle Dynaj,,i,;s and Control Division NASA Hugh L. Dryden Flight Research Center P. 0. Box 213 Edwa-ds, CA 93523 HANDLING QUALITIES OF LARGE FLEnIBLE CONTROL-CONFIGURED AIRCRAFT Introduction This is the first semi-annual status report on Grant No. NSG 4018.

The work began in January, 1Q79 with the appointment of Mr. Supat Poopaka as a one-half time Graduate Research Assistant. He is a Ph.D. student in the School of Mechanical and Aerospace Engineering. Mr. Poopak.a has extensive background and capability in the analytical methods of Modern and Optimal Control Theory, but had no course work or background in aircraft flight dynamics and aeroelasticity. As a result, much of this first reporting period has been a learning process for film. He completed an Aircraft Stability and Control course taught by Dr. Swaim in the spring semester , nd in addition has rapidly become knowledgeable in handling qualities, pilot modeling and aeroelasticity through intensive self-study.

Discussion of Progress As described on pages 18 and 19 of the proposal for this work (Ref. 1), our approach to an analytical study of flexible airplane longitudinal handling qualities is to parametrically vary the natural frequencies of two symmetric elastic modes to induce mode interactions with the rigid body dynamics. Since the structure of the pilot model is unknown fur such dynamic interactions, the optimal control pilot modeling method is being -2- applied (Ref. 2) and used in conjunction with a pilot rating method (Ref, 3). A pole placement algorithm is also used to ^• .-A ntain rigid body dynamics at acceptable values on short-period and phugoid fre- quencies and damping ratios. This should ensure that the pilot ratings are based on the relative amplitudes of rigid and e'astic pitch responses and not on poor rigid body dynamics.

Figure 1 is a block diagram depiction of how the optimal pilot model is structured and fits into the aircraft and display dynamics blocks. The tracking task is to maintain a reference rigid hitch anyle 9 = 0 in the presence of random turbulence as a disturbance input. Table 1 shows the model parameters; Table 2, some response equations; Table 3, the model parameter values; and Table 4, the matrix equations yielding the quadratic optimal control solution.

Our intent has been to use the optimal pilot model results to establish separation boundaries delineating when the pilot can discern rigid pitch e 'from the total pitch angle as viewed on a flight director is given by display or on the outside horizon, where e i - .029^2(t) 0 i (t) = e(t) - .025F, 1 (t) and includes the pitch contributions from two low frequency elastic modes.

The aircraft dynamics being used are basically the B-1 airplane at a sea level, 949 ft/sec flight condition.

We have completed the computer proyrams required for this effort and have preliminary results for three combinations of first and second elastic

o-

r

L V1 ^ •r V

Z

4J I I C

O ^ ^

7 I ^ L ^ > ^ L GJ ^- X L)

I

J

^ k• • fl r - n W H- C ^n I y

O

_ H

C 1

I L .Q O QJ -^ r 3 1 .J F- tn to I w v 4- U O J L W ^ C I D I 10

"J I

e ^ II ^I y j v cx (X E G O L L d O C w C O O X .-, I ^ C D O

p •r Z rU

uj E Cy N O a ^- cn ^ v C N CT f ,n w v Q ^k DC Z C O S- ^ O d a -3 O - - cn J d O I I

y

c

v

3 ^

E

o

r I^

Z J r I I a1 m ao [L' h- + ..

Z ::D — O f

o

V) c-)

^ I a--

I I

^^ J •r

_ d

TABLE 1. DESCRIPTION OF THE OPTIMAL CONTROL MODEL PARAMETERS

MODEL PARAMETERS I RELATION TO PILOT PERFORMANCE

Aircraft motion variab l es Pilot must observe this well enough to X command aircraft and provide stability u Aircraft control variables Pilot must use this to command aircraft and to provide stability A Aircraft dynamics (stability deriva- Aircraft must be stab l e enough for pilot to control tives with inertial and elastic coupling) 4 Aircraft must respond to the comamands in Aircraft control effects (sensitivity to control and throttle deflections) such a way that the pilot can understand C Displayed motion variables must be Motion variable display transforma- sufficient for command stabilization tion D This infers control observation Control variable display transfor- mation A E Dynamic model "learned" by The better the p illot's knowledge of the aircraft and his own capabilities, the -' the pilot, including neuro- n N muscular lags better he can cope with noisy measurements -T L Control gains which transform pilot's Pilot attempts to tradeoff aircraft motions estimates of aircraft motions to with available control power control actions V Large values decrease accuracy of observa- Covariance matrix of observation Y noise tions V Covariance matrix of neuromotor noise Large values indicate pilot is having m difficulties controlling aircraft TABLE 2. OPTIMAL CONTROL MODEL OF PILOT RESPONSE Controlled Element Dynamics Aircraft Short Period Dynamics Augmentation System

x = Ax + Bu

Mission Phase Vestibular Afferent Dynamics Measurement Vector Displays (Visual and Vestibular) y=Cx+Du Cost Functional Weightings Mission Requirements VMC/IMC Cues

J = E tf(y TQYY + U T Q u u + 6TQR6)dt

First Order Lag Neuromuscular/Manipulator Dynamics

Tnu+u=m+vm

Shaping Filters in State Equation Disturbance Environment (Turbulence, Shear, Guidance

z=Ax+Bu+Ew

Error) Cov ( w) I

I

f

i Work Load Expectations

C i Attention Allocation, MIN J, Ef c 1 = f , f cl ^ 0 ci VMC = Visual Meteorological Condition IMC = Instrument Meteorological Condition TABLE 3. MODEL PARAMETERS SELECTION Qy = Qyl = (50)2 Q. = Q = _ Y Y2 (50/s)2 QO (100)2 (a large displacement manipulator, high force gradient T N = .2 device such as elevator control on a large transport aircraft) Q R is chosen to provide T N = .2 5 Controller Time Delay T = .2 s (Typical value for human opera or) Observation Thresholds (10% Full Scale Value) TH y = 2° , THy = TH y = 20/s Observation Noise Ratios -20db is a typical value Additive Motor Noise -25db is a typical value Attention Allocation f c = 5 , f c = .5 Y2 Y The Attention Allocation can be optimized with respect to J, i.e., MIN J , but for the case being studied J s not sPositive to fc f ^ c therefore, the fixed value of f will be used throughout the study.

i TABLE 4. PILOT MODEL SOLUTION EQUATIONS m(t) = -LxW Pilot Control Control Gain M dLrixT o Ao = o Bo = I N -1 (L,I) = Q R -' B0K Co = {CID) g0K 0 +r0 Q o B o Q R -r BOK O = 0 + o T Q y C 0 - K K0Aa + C Pilot Kalman Esti- A, x(t + B,m(t - -r) x(t mator - u(t - T) _ u(t - 1) + F.Cav y l y W - C o X(t - p u(t - A, F. + EA1 + w - EC o V y -1 C O E = 0 = B = ° WEET A ' TN -' J ' IvmTy- z H o x(t - T ) ^(t Pilot Predictorx(t) = t , (t) + e H '' _ T) U(t - T) u(t) l t(t) + BIM(t) ^( t ) = A Pilot Transfer u(S) = -(T N S + I} - '{F + I} ",L[O)e (Al - SI gSI - A) -1 F, Co TVy-ly(S) Function r - - IS)T IS - A) F = U10) ( e ^ - (SI - A,) - ] + (SI - A,) B, ECoTVy_iCo A = Al - Table 4 Continued

r T

(t) F X = Ioe A l°Wo I`^do

F r.'t) {x T (t)u T

Covariance of Pilot

Predicted State

Lu(t)

+ fueAaeA1TECOTVY-^CO^eAlieATado BoTr4-1{LII1 A = A 0 -

i = 1, ny

(C,X,-OT)ii

Efm^( t )l ; ('-'(LIO)X{LIO)TTN-l)ii

i = 1, n

u

E{vy(t)vy((r)).= V Y d(t -o)

Pilot Observation

Noise Covariance a^ Y

V y , i = 1, ..., ny

( ) ii

-.^ ^i

8 a i /11(a i

)

t E

ai = [

{yi(t)^

= Describing Function Gain of Threshold N (o^) = erfc (ai/aiY2) = Full Attention Noise Ratio e' Yi = Attention Allocation for yi f

Efmm(t)vm(Q)) = VW(t - a)

Pilot Neuromotor

Noise Covariance s nu i = 1, Efm i (t)) (V m ) ii = em i em = Motor Noise Ratio i ^ C) M ^' CL E ^ CNJ p

v ^'^' O

^o ^ ^"^ M t7 a U 0+

C O%

S

Q O O a i v.

A 4 _• N ^(l ^ I ^O

C N N

v a o ^ ^' '`r o c C o W

N a

a a a W ,- c Co N Q) (7) ll O Z Q v 3~ I N W a N r d .O C7 r O U L O d co N r 0 O O 4-^ O

c p

U W J )^ a Cfl _q d r O O

O O O O

C +) ^ t a a c ^ r n -Ct M V) 11 1 ,n ^ LC) ^1 U fa a-) >., N 27 r a V O O O 3n M N r-) M jr c U >> N ^J O > O r I L c a ro II p ^ 3 era a E 3N E E L'

Overview