SEP 2 ^' +RE D:

^, Copy No. /

COWIDENTIAL

RN No. SIAM1

*4%NACA-.CIA& ..,ATG,H C^"'^I^A

RESEARCH MEMORANDUMfor the

Bureau of Aeronautics, Department of the Navy

LABORATORY INVESTIGATION OF 740 AUTOPILOTS FOR A 4- SCALE

DROP MODS, OF THE GRUMMAN FRF-1 AIRPLANE -

TED NO. NACA 2466

By

Jerome M. Teitelbaum and Ernest C. Seaberg

Langley Aeronautical LaboratoryLangley Field, Va.

CIA& i 4 Ci _ ^ -00N,

{hs lrn l lief ,o ei .̂̂ {be mea

ea n 11 anaml—,, He Act,

a0i31 std 92, tta r ieefon or the

elation of lta contents in any manner to m

^

r eI torized person fa Probf bl ted by 4w.information a Uaaafffedo may be imparted

oNy to persons In the military std [ovalservices M the U pped States, appro

^.dwas^nffiWleedflarr^ d eta^lgaesd ihry^^^r^:x w'pL`- ^iGovernment wtK, have le^ltlmate Itderesl 'therein, ad m United States ettls.ns of Imownl ty and dterrst:- wta o! neCesslty must be

y,^ed Mar f. r

NATIONAL ADVISORY COM fFOR AE RONAUTI CS th8

lo t), 00 r v ^,WASHINGTON Ad^4 0l

r

t 0

r/ to, Al.. *rt/)11e.v,^rAWL ft

CONFIDEInIAL

44

https://ntrs.nasa.gov/search.jsp?R=20050029396 2020-03-25T10:54:37+00:00Z

- NACA RM No. SL8109 / jlllllj t I;lll~~i~~~~~i~i~~~lllii~iii~ t iill 1 N c LAsSl Ff ED 3 11760143'7~763 '

NATIONAL ADVISORY COMMITTEE FOR AERONAUTICS

k-

RESEARCHMEMORANDUM

for the

Bureau of Aeronautics, Department of the Navy

LABORATORY INVESTIGATION 0% TWO AUTOPIUXS FOR A

DROP MODEL OF THE GRUMMAN F8F-1 AIRPLANE -

TD NO. NACA 2466

By Jerome M. Teitelbaum and Ernest C. Seaberg

SUMMARY

Performance investigation and frequency response analyses were con- ducted en two autopilot control systems designed for use in the 1

h-scale model of the Grumman F8F-1 airplane. The first system, based on

the action of a displacement gyroscope only, was investigated to find the cause of a small-amplitude pitch oscillation which had been noted in previous flight tests. The results of the investigation conducted revealed that, although the autopilot-model combination was dynamically stable, a hunting oscillation was possible due to a change in autopilot charac- teristics in a dive. This hunting condition can be elim2nated by increasing the gyroscope pickoff dead spot without greatly reducing dynamic stability of the autopilot-model combination.

The second system based on the combined action of displacement and rate gyroscopes was tested to determine the autopilot characteristics desirable and to predict the control-surface linkage ratios required for dynamically stable flight. The results of the tests conducted sho;J that satisfactory autopilot characteristics can be obtained by selective mixing of the signals from the displacement gyroscope, the rate gyroscope, and the feedback system. By using this adjusted autopilot and by selec- tive gearing between the control surface and the servomotor, conditions for stable flight based on the frequency response analysis can be predicted.

A comparison of the two autopilot systems shows that slight Uprove- ment in the dynamic stability of the model can be expected when the second system mentioned above is used.

u~C~A%%=&J _. .; _ . . . ..-. ---1_p--- .--. ., - --

2

INTRODUCTION

NACA RM No. S&O9

At the request of the Bureau of Aeronautics, Depar-tment of the Navy, the NACA conducted oscillating-table tests on two separate autopilots designed and constructed by the Naval Air Experimental Station, Philadelphia, Pa., for use Fn the L-scale Gr

10 umnmn F8F-1 drop model in

order to stabilize and program the flight of the model. In the.first autopilot (to be referred to as system l), the problem involved was to determine the cause of the oscillatory flight tendencies of the model during the diving portion of flight. The investigation conducted was based on an examination of the following:

1. Effect of signal pressure cut-off

2. Signal pickoff dead spot

3. Stability analysis of the autopilot-model combination

4. Effect,of dive angle on the autopilot response

5* Effect of a roll oscillation on pitch and yaw controls

In the second autopilot (to be referred to as system 2), the problem vas to determine the effect of the rate gyroscope, the displacement gyro- scope, and the servomotor damping on the autopilot, and to predict the control:surface linkage ratios required for dynamic stability of the autopilot-model combination. The tests conducted consisted mainly of frequency-response table tests in which the signal strengths of the component parts were var$ed. As the autopilot was constructed identically in all tiree planes (that is, pitch, yaw, and roll), the tests were conducted in pitch only with and without simulated air load on the servo and the results applied to all three motions. All tests xere made by the Pilotless Aircraft Research Division of the Langley Aeronautical Laboratory.

SyMBOlSAIiDDEFINITIONS ?

K

Km

%

control-amplitude ratio (ratio of control deflection to airplane displacement)

control-amplitude ratio obtained from solution of aerodynamic equations of motion

control-amplitude ratio obtained from oscillating-table tests of autopilot

NACA F&M No. SL'3IOg

E

em

eP

(u

8 -

QI

tf

&e

'a

'r

V

Hunting

phase angle, degrees (positive indicates l&d of 6 ahead of 0)

phase angle obtained.from solution of aerodynamic equations of motion, degrees

phase angle obtained from oscillating-table tests of auto- pilot, degrees

angular frequency of oscillation, radians per second

angle of pitch;degrees

angle of bank, degrees

angle of yaw, degrees

elevator deflection, degrees

aileron deflection, degrees

rudder deflection, degrees *

true airspeed, feet per second

- a self-excited steady-state oscillation

Gyro pickoff - the part of a gyro unit -tihich detects the error between gyro reference and airplane attitude and supplies the corrective signal to the servo

Dead spot - the angle included within the limits of gyro displacement which results in no servo motion

Control-surface linkage ratio - the ratio of the angular deflection of the control surface to the angular rotation of the servo arm

Cross couplin63 - the response of an autopilot mechanism in one plane to a disturbance of the model in another plane

APPARATUS

The two autopilot control systems tested were mounted on trays and were supplied by the Navy Department.

NACA RY No. SL8103

System 1

.

. . .

:

. .

: 0.

. . . .

. .

. : . . .

This system consists of three air-driven Diehl Directional Gyros and three Lear Linear Actuators, model &OAR. The setup used is similar to the one shown in figure 1 and described in reference 1. A schematic representation is shoim in figure 2. Although the autopilot is a three-axis system, only one axis is outlined.

Movement of the servos (actuators) which operate the elevators, ailerons, and rudder is controlled by pneumatic electrical gyro pickoffs. Gyro displacement moves a cut-off plate between a jet and orifice type pressure feed to the pickoff. The pressure variation due to the position of the plate causes movement of a spring-loaded diaphragm. The diaphragm controls the motion of a switch arm between two contacts which, in turn, energize a pair of relays that engage azld control the direction of motion of the cons,tant-speed servos. The gyro cut-off plate is so constructed that either a fast or slow rate of change in pickoff pressure with gyro displacement can be selected, as shown in figure 2. The dead spot is adjusted by changing the spacing bet-,reen the contacts in the pneumatic electrical pickoff.

The follow-up system is superimposed on the gyro pickoff by means of a cable linked between the servo and a pulley geared to the pickoff orifice. Movement of the servomotor due to the gyro displacement causes the orifice in the.pickoff to move in the direction of the gyro displace- ment. This motion continues until the air pressure in the orifice is such that the switch arm breaks contact and the servo disengages. By this method, servo motion becomes proportional to the amount of gyro displacement. The dive mechanism consists of a means of varying the active length of the folio-d-up cable, thus moving the pick-up orifice or base reference location to give the desired dive or climb.

System 2

This system is composed of three air-driven Diehl displacement gyros, three Pioneer, P-l, rate gyros, and three Pioneer two-phase, induction-type rotary servos geared to give the required torque output. The autopilot containing the gyros and amplifier is shown in figure 3. A block diagram showing the operation of one axis of the autopilot is contained in figure 4. Four-hundred-cycle autosyn pickoffs are used to detect both displacement and rate errors- In order to vary the rate and displacement response sensitivities of the autopilot, three controls are added to the system for each plane of %otionh a rate' control varies the effect of rate error on the system II a range control varies the effect of the displacement @r-o, and the follow-up' control varies the amount

.

. ~~ __ -. - ___- .__.

NACA RM No. SIi%O~ l I----'- 5

of servomotor damping. The output from each pair of gyros (that is, one rate and one displacement gyro for each plane of motion) is fed through an electronic amplifier, which is electrically connected to the servomotor. A 400-cycle autosyn pickoff is geared to the servomotor and furnishes feedback signal. The strength of the feedback signal as compared with the signals from the gyros is a measure of the effective servo damping. The output of an inverter supplies the electrical power necessary to drive the rate gyro motors, autosyn pickoffs, amplifier, and servomotors.

The system operates as follows: As the model moves off course, the autosyn pickoffs on-one pair of gyros are displaced with respect to their gyro axis. This causes electrical signals to be sent to the rate and range control, which are preset. From the rate and range control,'the combined signal travels through the amplifier and energizes the servomotor, causing control-surface displacement at a speed proportional to the strength of the input signal. Movement of the servomotor also causes the autosyn pickoff in the servomotor housing to be displaced and feedback to the follow-up control, a signal which is opposite in sign to the signal produced by the gyros- Mov+ement continues until theso two signals equalize each other, thus giving proportionality between the displacements of the gyros and servo motion.

Equipment

An oscillating table capable of producing sinusoidal oscillations up to 5 cycles per second and with amplitude adjustment up to i-lT" was used to obtain data for the frequency-response tests. A position recorder was attached to the table and a stiilar one attached to the servo in order to record table motion and servo position as functions of time.

PROCEDURE

System 1

The entire autopilot system was mounted on a tray and attached rigidly to the oscillating table in a position so that the motion of the table would simulate a sinusoidal pitch oscillation of the airplane.. The frequency response of the autopilot system was obtained by oscillating the table at various amplitudes and frequencies and recording the table displacement and servo motion as functions of time. Dead spot was measured by disconnecting the servo and recording table displacement and power to the servo. Dead spot, in degrees, was obtained as the difference in table position required for signal reversal to the servo. The tests to determine the effect of the dive mechanism were conducted by uncaging the gyro in the level position and rotating the table approximately 15O at various

6 NACA RM No. SrFjIO9

displacement rates at the same time as the dive mechanism was energized. Cross-couplFng tests were conducted by mounting the autopilot on the oscillating table to simulate rolling of the model and recording the roll -position and motion of the pitch servo due to the-roll displacement. The tests were conducted with the pitch displacement gyro signal pickoffs in both the normal and wide-open positions.

System 2

Thea autopilot tray contatiing the displacement and rate gyros., amplifier, and rate, range, and follow-up controls was attached to the oscillating table to simulate the pitching oscillation of the model. The servomotor, inverter, and air filter were mounted on an adjacent table and connected to the autopilot as shown in figure 5. The air load on the servomotor was sLmulated with a shock-cord arrangement, whereby 8 to 12 foot-pounds of resisting torque were applied to the servomotor upon movement of the servomotor arm in either direction from centered position. The arrangement bed is shown in figure 5. Oscillating-table tests were conducted at various table amplitudes and frequencies and with various group settings of the rate, range, and follow-up controls. The effect of air load was determined by conducting some of these tests with and without simulated air loads and comparing the results.

RESULTS AND DISCUSSION

System 1

This system, based.on the action of a displacement gyro only, was investigated to find the cause of a small-amplitude pitch oscillation which had been noted in previous flight tests. The tests were conducted without simulated air load as this load is not sufficient to affect the operation of the servomotors. The results of the tests conducted to determine the cause of the oscillation noted are as foILLo-ds:

Signal pressure cutoff.- The tests were conducted with the slow rate of chznye of pickoff pressure with displacement only as attempts to use the fast rate produced hunting servo oscillations under all conditions.

Simal pickoff dead spot.- The dead-spot measurements using the slog rate of change of pickoff pressure cutoff were recorded as folloxs: Pickoff normal operating conditions, l&O and pickoff wide-open condition, 2.2O.

Stability analysis.- By application of the frequency-response method outlined in reference 2 and using the aerodynamic data given

- _- ._ ^---~- -- _..__... .~ _ ., - --_-_- .~ _I

NACA RM No. SI8109 7

in reference 3, the longitudinal stability of the autopilot-model combination was determined for a control-surface gearing of 3.75O control deflection per inch servo travel and for model velocities of 300 and 850 feet persecond. The results of the tests with the pickoff set in the normal operating position are shobm in figure 6, in which control- gearing ratio K and phase angle E are plotted against angular frequency cD in radians per second. The results of the tests with the pickoff set in the wide-open position are shown as frequency-response curves in figure 7. Although some variation in characteristics was noted between the two tests, an analysis of the results obtained showed that in both tests the autopilot-model combination was stable for oscillations having amplitude of approximately *go, +7’, dk”, and 52' and that, in flight, the model should damp to a value below *2'.

Dive effect.- The results of the tests for dive condition with the dive mechanism in operation and with the dive mechanism inoperative are plotted in figure 8 for normal pickoff position and figure 9 for wide- open pickoff position. The tests were conducted at various autopilot displacement rates in order to cover the possible rate of change in air- plane displacement that may be encountered in actual flight of the model, as this procedure was simpler than trying to calculate the theoretical flight path of the model. The results show that, with the pickoff in normal position, a hunting control oscillation became apparent at the end of the dive, while setting the pickoff in the wide-open position eliminated this oscillation. The hunting oscillation noted after a dive with the pickoff in normal position was due to the fact that the servomotor con- ' tinued to coast after the pickoff signal stopped. This overshoot caused the follow-up system to change the orifice pressure in the pickoff to the extent that the spring-loaded diaphragm moved and closed the contact, giving reverse signal to the servomotor. No explanation as to why this hunting existed only after a dive could be found. With the contacts in the wide-open position, movement of the diaphragm was not sufficient to close the contact which previously gave reverse signal to the servomotor and started servo hunting.

Cross coupling.- With the pickoff in normal position (dead spot, 1.&O), no control oscillation in the pitch sense was noted due to roll until an outside source initiated a disturbance of the follow-up pulley. After this disturbance occurred, the servo control continued to oscillate at the same frequency as the forced roll oscillation. With the pickoff in the open position, it was not possible to obtain a coupled oscillation of pitch control due to roll oscillation. Figure 13 is a plot of the results of the tests showing pitch-control motion due to a roll oscilla- tion of 6$' at frequencies of $, 1, and l$ cycles per second with the

pickoff in normal position and with an outside source initiating a disturbance of the follow-up pulley. Although the tests conducted

-. - .i_ _ ., _ __ _..- -.---_ . . .-_. __-__ ~_--_

8 NACA Rt't' No. SI8109

measured the pitch-control response to a roll oscillation, preliminary tests were made and similar coupling appeared to be present between all planes..

System 2

This system employing the use of a rate gyro in addition to the dis- ' placement gyro used in system 1 and having a-proportional servo in place

of a constant-speed control actuator vas tested to determine :qhether or not improved flight characteristics could be expected with this autopilot- model combination. The results of the tests conducted are as follows:

Effect of load.- The relationship between simulated air load and no load for a typical group of control settings with sinusoidal table oscill:- tion of f3.12O is shoim in figure 11 as plots of control-amplitude ratio K and phase angle E versus angular frequency w in radians per second. The two sets of curves contained in this figure indicate that loading the servomotor lowers the control-amplitude ratio approximately proportionally throughout the frequency range. This is accompanied by a decrease in lag at the higher frequencies.

A possible explanation of this change in phase relation bet-Jeen the autopilot oscillation and the resulting servomotor motion at high frequencies is that the maximum servomotor rotational velocity is attained under no-load conditions at a lower frequency than under loaded conditions. As the rate gyro can no longer increase the speed of the servomotor above this critical frequency, it is feasible that the loaded servomotor should still have an effective rate signal at higher frequencies than the unloaded motor. It follows that the loaded condition will have less lag than the unloaded case at high frequencies. The combination of the lower gearing ratio and decrease of lag for the loaded condition is to increase the dynamic stability of the model according to the criterions for stability outlined in reference 2. This increase in stability under loaded conditions was the general result noted throughout the tests.

The subsequent figures are therefore based on no-load conditions in order to make the results, from jrhich the final servomotor linkage ratios w+re obtained, apply to the most critical conditions.

Effect of varying: control settings.- The effect of varying the rate, range, and follow-up control settings are shown in figures I2 to 14, which are >lotted for a table amplitude of f3.E'. These are plots of control-amplitude ratio K and phase angle E versus angular frequency w in radians per second and are typi'cal examples of the results used to determine the optimum control settings. Optimum operating condition was selected -t;o be the combination of control settings that would produce a

NACA RM No. SL8109 9

constant ratio of control (servomotor) motion per unit autopilot displace- ment throughout the frequency range with a minimum amount of phase-angle lag. The effect of varying rate (fig. 12) shows that high rate cohtrol settings produce a nonproportional control-amplitude ratio K with increasing frequency, while low rate settings increase the servomotor lag. The effect of varying range (fig. 13) shows that high range settings increase the servomotor lag, while low range settings make the curve of K versus frequency less proportional. The effect of varying follow- up (fig. 14) shows that decreasing the follow-up control setting causes the curve of the control-a+$itude ratio versus the frequency to peak sharply and also increase the lag of the servomotor. From these tests, the control settings which give the best performance were selected to be rate 65, range 70, and follow-up 100.

Stability analysis.- The frequency-response tests of system 2 were measured for sinusoidal table frequencies of O3 to 2 cycles per second and for table amplitudes of ?O.j2O, +1.45O, and 23.16O in order to predict ' stability for various disturbances- To determine the linkage ratios between control surface and servomotor for stable flight, frequency- response plots shown in figures 15 to 17 were made assuming unit gearing between control surface and servomotor. The model frequency-response curves I& and em versus c

0) for airspeeds of 300 and 950 feet per second and based on the aerodynamic data obtained in references 3 and 4 were plotted for the longitudinal and lateral planes of motion so that the stability requirements could be predicted in pitch, yaw, and roll. In pitch 6, a linkage ratio of lo elevator control for 8' of servo motion, in roll f a linkage ratio of lo total aileron control deflection for 12 'of servo motion and in yaw J# a linkage ratio of lo rudder control for 10' servo motio; will satisfy ihe criterion for stable flight.

Nyquist diagrams of the frequency-response curves (based on the method outlined in references j and 6) in which the control gearing values mentioned above were included are shown in figures 19 to 20. Since the resultant curves in figures 18 to 20 do not enclose point (I-,0'), the critical point when NACA nomenclature for control deflection angles are used, the stability criterions are satisfied.

CONCLUSIO~XS

With the use of system 1, based on the action of a displacement gyroscope oLnly, the autopilot-model combination will be longitudinally stable for disturbances in the range tested (that is, and 2'). Under normal operation,

9 = 230, 7", 40, diving the model appears to start a

,.. -... -..-_.. --_

10 NACA R'4 No. SL!UO9

hunting oscillation. This hunting oscillation encountered may be eliminated by increasing the dead spot in the &yroscope pickoff from the normal value of 1.4O to 2.2”. Although doing this would change the frequency-response characteristics, the autopilot-model combination would remain stable-in the amplitude range tested. Cross coupling of the auto- pilot was present with the pickoff in normal position. This also can be eliminated by increasing the dead spot in the gyroscope pickoff to 2.2O.

For system 2, based on the combined action of displacement and rate gyroscope; dynamic stability for the autopilot-model combination in flight can be obtained using the recommended control settings and control- surface gearings.

A comparison of the two autopilot systems shads that slight improve- ment in the longitudinal stability of the model can be expected when system 2 is used since the frequency response for this autopilot-model combination she-ds that the model will operate at a higher frequency with somewhat better damping characteristics.

Langley Aeronautical Laboratory National Advisory Committee for Aeronautics

Langley Field, Va.

Jerome M. Teitelbaum Aeronautical Research Scientist

cLzd&ffl&A Ernest C. Seaberg %

Aeronautical Research Scientist

Approved:

Robert R. Gilruth Chief of Pilotless Aircraft Research Division

. ‘:*. ,. B..

: I..

NACA RM No. X8109

RIiTEtRENCES

, . : l :

11

1. King, L. A,: Laboratory Test of the Perfomce of Autopilotc for 4/10-Scale Model *Jf the G-rumman F8F-1 Airplane - TED No. NACA 2466. NACA MR No. LjHO3, Bur. Aero., 1945.

2. Greenberg, Harry: Frequency-Response Method for Determination c:f Dynsmic Stability Characteristics of Airplanes with Autcmatic Corlt2- 210 . NACA TN No. 1229, 1947.

j. Co&n, Doris: Analytical Investigation of the Stability of an F8F Dropping Model with Automatic Stabilization. NACA MR No. LSLO3, 1345.

4.. Watson, James M., and Comisarow, Paul: Mind-T-unnel Tests sf a l/j-Scale Po-qered Model of the Grumman XF8F-1 Airplane. III - bterzl Stability and Control - TED No. NXA 2344. NACA MR No. LjFOj, Bur. Aero., 194j.

f. James, Hubert M.., Nichols, Nathaniel B., and Phillips, Ralph S.: Theory of Servomechanisms. McGraw-Hill Book Co., Inc., 1947.

6. Hall, Albert C.: The Analysis and Syn"&esis of Linear Servomechanisms. Servomechanisms Lab., Mass. Inst. Technology, Cambridge, Mass., 1943.

t

CONFIDENTIAL

f'ik ure 1.- System 1 autopilot forlU -scale Grumman F8 P'-1 drop model.CONFIDENTIAL

za

a

z0U1

HOto

. : . .

D o w n + P /ro //o w - u p C a b le --A ----- -----

L

!_ “I”;:

J- L ,e g e n d

- - - - M e c h a n ~ c d L ink L- lec ffm l C o n n e c t/o n

- 2 8 ”Dc l - I Figure 2 .- S c h e m a tic d i a g r a m o f sys tem 1 a u topi lo t fo r h -sca le G r u m m a n F8F-1 d r o p

O n e ax is s h o w n only.

CONFIDENTIAL

ana

z0

r00H0CD

FiI , ure 3.- ',ystern .' autopilot for 10CONFIDENTIAL

F81 , '-1 drop model as installed on oscillating table.CONFIDENTIAL

- .:..

Servomo tar

.

Rute Gyro

--27 Cunbd

I

1 eyed t%/ow-up Control

---- - -/Mechvncal I! mk E/ectr/ca/ Cu/mect/on

Figure 4. - Block diagram of system 2 autopilot for $- scale F8F-1 drop model. One axis shown only.

L-55571

CONFIDENTIAL

;D

yNF-1

110

V^

HUti)

Pij , ure 5. - Pest setup used to obtain frequency responCe records on system 2 autopilot for 10 - scale1 08F-1 drug rriodel.

CONFIDENTIAL

NACA RM No. SL8109

* & 20

e $ d6 P

-% * A? P

k \r, .8

P 8 kJ .4

/4 /6 /6 2 4 6 8 /ii? /2 A?guh?- frquency , uil, rod/&c

Figure 6.- Longitudinal. frequency response for h-scale F8F-1 drop model using system 1 autopilot with normal pickoff opening.

_.. -- I.- ---. - .-. _. .~. __._. -.--.._~-_ ; _ . .-_ ^-. -

. . .

NACA RW l!Jo. SL8109

24 v

/ire nl

850 n’

e I. I I

I /I I I ‘q’ , -1 - ~ -16 - I 0 --D-f

I

-b/5 -- Q Q g--- gy;

P I A ---- t&y

2 I I I I

-44 I 2 -$ .8

? 23 k 4 s;

3

P

401 I I I I I I I I I I I I I I I I I I 1 I w -I // ,

\ I a,\ ‘\ \ I 4\ \ \ /’ --I I 4

. .

A~uhf frequency 3 ti , rd/~ic

Figure 7.- Longitudinal frequency response for &- scale F8F-1 drop model

using system 1 autopilot with pickoff in wide-open position.

-_. ----~-~. _- _ __ ._ .~~---. __~ ~ ~_. -~_ -.--i _

. . . . . l

. .

.:.. .

. .

. . .

: b..

. . .

: . I..

D . :

. . .

NACA RIM No. SL8109

u o-

4 .A=-

8 IO-

/2 /.5- /6

#a-

0 o- 4

5- 8

/u- /K?

a /.5- b 2.0-g ki ;\, $ 0-Q

.A 4 k .5-% % k.8 -@ LO-a

a, /!2 i 4%),J

$20-4 3 2;; is--

EU

CUFYc? AZ’. /

---- \

------------__

curve no. 2

\ \ \ . --- -----

cuwe no. 3 0 ----

o- 4-

.5- 8- \ IO- . \ /2 - \ 15- .

/6 - _-_------ 2u-

20 - 2.5-

.z4 - m-

0 z/me, ..ec

Figure 8. - Effect of dive for &- scale F8F-1 drop model using system 1 autopilot with normal pickoff opening. Curves Nos. 1 and 2 are with dive mechanism inoperative and curves Nos. 3 and 4 with dive mechanism in operation.

. ~.- -. ___~ -_.. ~-- __ -_~. . ._ _ _.---

. . . . : : . .

. .,.. ..:

NACA RlVI No. SL8109

. . . . . .

B0.b

. . .

: . bea*

B. . t,

. . , .

o- ---\ o-

4- .5-

8-- LO-

/2 - 15-

16 - L----___ ZD-

o-

.5-

LO-

-----------__

. ,

O- .5- LO- &-- 20- L5-- Jo- a--

20 - 28 - 36 _ c‘!h+-b‘? b% 4

I I I I 0 25 30 35 4'0

Figure 9.- Effect of dive for h-scale F8F-1 drop model using system 1 autopilot with pickoff in wide-open position. Curves Nos. 1 and 2 are with dive mechanism inoperative and curves Nos. 3 and 4 with dive mechanism in operation.

._ -_---_--. . . ..~ _ _ . ~~-I_- _ _, _ ;~ _.------ -

I..

NACA RM No. SIB109

8.

: ,.

I. . .

I.

1 : . .

.

f I I 7 E/evaror

I -, , I -9 y4

-------

q * ,Y’/-,, /e /I= k- /id/e ro/hg mof/bn _ i

3 -q /l/f \ , / /

.., i

en mn/ ! \ /

\

\ 4 1 n II f >- s- \ \ / I/I/f 1 1 I,h I I, I I I I ‘.I I I’ I I I I I\I I P .x 2 ,I I I I I P-l-1 1 1 1 1 1 p+-q 1

x-/ 3-4- Q k

$ ~b/erofingh4uetxy= 5cp.s ' I I

-8

s h 4-744-T; I I $- P D 4 -7 /‘iI+ fl\ /(,\ ‘r \\I /I \ I , $0 yo % $ &/ B-4

.$ Qc k -8

/To;bk rohng h-quem-y= /cp.s , t I I f I

,

-4 - h%rihg fieqtx.ency= A 5cps \ I I I I I I I I

-80 / .5 /o h- I I 20 I I 25 30 Tqfii- 35 40

Figure 10. - Cross-coupling effect for 6 -scale F8F-1 drop model us

system 1 autopilot with normal pickoff opening after a disturbance u; initiated by outside source. Amplitude of table oscillation in roll of 1k3.25~.

ing

‘as

2 . : . .

I’,TACA RIX No. 3L8109

. .

. .

: :

-- I I I I I I I I I Cl --m-_-N servo witi smmh+e~ a/r /uud

0 Servo wPYm4t smd77W u/r

? 8 LP \ -8 -i? .8 -K &

-i .4 -& 8 Ll

u

J I l-- V I 1 .

. .

‘0

* -w- 8 f+ -JO- 4 T

-/zoo’ 6 8

h7ph-r frequency , 0, rad/sec

Figure 11.- Effect of loading servo on system 2 autopilot for & - scale F8F-1 drop model with autopilot control settings of rate 50, follow-up 100, and-range 50; with simulated air load on servo of 8 to 10 foot-pounds; and with table oscillation amplitudes of L3.14O.

_ .-.- 1 _ .__- ._ _ - _._. -__ -_____ - _I _. ---

NACA RM No. SL8109

24

2.u

I I I I I I ._ Autqtht cmtrd se ftinys-

472 fo//ow-up rmye-

‘y-j++-+ ;-----!f ‘i” ‘T I_ * /6, 0" Q I I I P

. \I, h

\ \

\ \ \ \

h\ \ N- \ . . .

tl \ \

^_ 1 g -&v I 1 I . , l-l

T --++

. I I

-% 2 4 6710 L? /Y /d I I 1

-v’

’ Any&r frequency 3 U, rvd/sec

Figure 12.- Effect of variation of the rate control setting on system 2 autopilot

for &- scale F8F-1 drop model without simulated air load on servo and with

table oscillation amplitude of L3.12’. .

. . . . .

FJACA RN No. SLi3109

1 : . .

.

24 I I I I I I I . 5ll +W//Of contra/ se f fmgs

; .4 e

I I I I I I I I I I I I I I I I

#

-8 9@ / o-- 0.

/’

c&J 4 b-4v Q

T 3 jpffffp -I- -lY.?A*I I -8U

I _

I I I I

-/q?

-/2lQ ’ 2 2 4 4 6-T--& I I I v

6 8 10 A? A? /4 /4 /6 16

Ahy.u~r fiquency , w, m&&c

Figure 13.- Effect of variation of the range control setting on system 2 autopilot

for 4 -- scale F8F-1 drop model without simxlated air load on servo and with 10

table oscillation amplitude of L3.12’.

/

I. - ,._. ..- -.I_- -. .~._ ,_. __.___. i .--- _- _I---._- _ - -I -- ~- ~”

t.. : ..: . . . . . : . .

: . . . . . . a*

. . . . . .

NACA RiVl No. SL8109

ran-k 50 /ii 50 a------

---

e- ; zu

t

2 /I6 T CL k L2

% k 3 .8

4 6 6 /cl ~2 14 /6 Anguhr frqueenc+i, 0, rd/sec

Figure 14.- Effect of variation of the follow-up control setting of system 2

autopilot for 4 - - scale F8F -1 drop model without simulated air load on 10

servo and with table oscillation amplitude of 23.12’.

I - ..~ .- _.-__-_ __~_ . . _ .._. ._ __ _~

. . .

: . I. .

. . . . .

NACA RM No. SL8109 4.. .: . . .

: . . . . . .

t.. : . . :

l :

k-3 b-iii I I I I I I 1

Figure 15.- Longitudinal pitch frequency response for h-scale F8F-1 drop

model using system 2 autopilot with unit control-surface gearing ratio, without simulated air load on servo, and using autopilot control settings of rate 65, follow-up 100, and range 70.

. _ _ -_ __--_ -----_ _ .

.: . NACA. RM No. SL8109 , ,..

- . f. : .

. . . . .

.: . G o--

&6-

3

-8 .* L?

p p .8

-3 k 4 .4

4 6- 8 /U A? /4 /6 x Anguhr frEsz/ency, ~2~ f-z&&

Figure 16. - Lateral roll frequency response for A-scale F8F-1 drop model P using system 2 autopilot with unit control-surface gearing ratio, without

simulated air load on servo, and using autopilot control settings of rate 65, hollow-up 100, and range 70.

. ^--. .~ _--- ~-.~ _..- “._. ---_-------_-._ ._.~_ -. --

,.. . . . .

.* NACA R&I No. SL8109

. . . .

. . . .

t

Anguh- frequeficq , u , md/ec

Figure 17.- Lateral yaw frequency response for h-scale F8F-1 model

using system 2 autopilot with unit control-surface gearing ratio, without simulated air load on servo, and using autopilot control settings of rate 65, follow,up 100, and range 70.

t Ia . . I. s.: . . .

8. : .

. : . : . . .

NACA RM No. SL8109

f

Figure 18. - Nyquist diagram for a longitudinal pitch oscillation of k1.45’ for

the h-scale F8F-1 drop model assuming V = 850 feet per second and a

control-surface gearing ratio of 1:8. No simulated air load and autopilot control settings of rate 65, follow-up 100, and range 70.

-,- - - -_ .---- -_ . - .._ . _ _ -.-- -. -__

. . . •~a . .

,

,..: NACA Rhl No. SL8109 .

I..

: . .

. . . . . . l :

. . .

Figure 19.- Nyquist diagram for a lateral roll oscillation of +1.45’ for the $- scale F8F-1 drop model assuming V = 850 feet per second and a control-surface gearing ratio of 1: 12. No simulated air load and autopilot control settings of rate 65, follow-up 100, range 70.

. -. -__---, _ _ - . .

. . . . . . . .

. . : . . t l

.:.. .

. .

. . . : >*.

a* : : 1.0 , . : .* .

NACA RIM No. SL8109

Figure 20. - Nyquist diagram for a lateral yaw oscillation of -1.45’ for tie

&-scale F8F-1 drop model assuming V = 850 feet per second and %

control-surFace gearing ratio of 1:lO. No simulated air load and autopilot control settings of rate 65, follow-up 100, and range 70.

1 3 1176Q1437,9763 j oilelunli~~~i~~i~i~~~liiiiii