A summary of results obtained during flight simulation of several aircraft prototypes with variable-stability airplanes

ORAND

By Walter E. Mc?JeiLl and Brent Y. Creer

- - - i c NATiONAL ADVISORY . . .

~

F O R AEROMAl

.

WASHINGTON NATIONAL ADVISORY COMMITTEE FOR AERONAUTICS RESEARCH MEMORANDUM A S U W Y Ut’ K E S U M S OBTAINED DURING FLIGHT SIMULATION OF SEVERAL AIRCRAFT PROTOTYPES WITH VARIABU- STABILITY AIRPLANES By Walter E. McNeill and Brent Y. Creer SUMMARY Two airplanes, an ~ 6 ~ 3 and an F-86A, each fitted with servo equipment for varying in flight the lateral and directional stability and handling characteristics, have been flown bytest pilots of the aircraft industry and the NACA to simulate the predicted dynamic behavior of six prototype airplanes. During these simulation programs, flight experience was obtained with lateral oscillatory characteristics represenkative of those predicted for each prototype and with other unusual characteristics asso- In cases where unusual character- ciated with certain specific designs.

istics were predicted, or where use of auxiliary damping devices was con- templated, the company test pilots gained familiarity with the trends in lateral behavior involved and were able to define ranges of acceptable characteristics which could be used as design criteria.

i

The methods of simulation and the types and ranges of variables considered are presented and the results of the individual programs are discussed. In addition, trends in pilot opinions of the lateral oscilla- tory characteristics are discussed in relation to current flying-qualities specifications.

INTRODUCTION Design trends associated with recent increases in operational speeds and altitudes of military and research aircraft have resulted in dynamic stability and control characteristics which sometimes differ widely from Several criteria for satisfactory those previously experienced in flight.

flying qualities have been developed as guides to aid airplane manufactur- ers in judging the suitability of their designs from the stability and control standpoint and as minimum requirements to be met by designers of military aircraft. The most recent of the military specifications for flying qualities of piloted airplanes is presented in reference 1; other published criteria for the more limited case of lateral oscillatory characteristics are given in references 2, 3, 4, and 5.

d While established requirements for stability and control may be useful as design guides, it has been emphasized in reference 6 that it is diffi- cult to include all factors which may be important in the over-all lateral dynamic behavior of an airplane. For example, the airplane's intended mission or peculiarities of a given design may have an important bearing on whether the associated flying qualities will be satisfactory to pilots.

One way of investigating the flying qualities of new designs is to use variable-stability airplanes as flight simulators of the predicted lateral dynamic behavior. In this manner, unusual behavior inherent in a particular design can be investigated in flight long before completion of the prototype; the pilot can become familiar with the lateral dynamic characteristics predicted for the airplane he is later to fly; and problems relating to a given design can be discussed with company engineers who are directly concerned. When this experience is provided in the early stages of design or prototype construction, design modification or installation of artificial-stability equipment can usually be made without causing pro- To obtain such experience, seven company test pilots have duction delays.

*

~ 6 ~ - 3 and F-86A variable-stability airplanes in prototype simu- flown the lation programs at the Ames Aeronautical Laboratory.

It s h o u l d be noted that these simulation programs were not the usual

research-type investigations - techniques varied; no standardized config-

I urations were tested; and results usually differed in type and complete- ness. In these programs, the variable-stability airplanes served as development tools (in much the same way as wind tunnels) for use by the contractors in connection with their individual prototypes. The purpose of this report is to describe the diverse problems and unique procedures involved and to summarize and discuss qualitatively the results of these flight-simulation programs. In addition, the novel features of the F-86A variable-stability equipment (developed by Mr. H. C. Patton, Jr., of the Ames Aeronautical Laboratory) are discussed.

NOTATION lateral acceleration at center of gravity, g units AY C1l2 cycles required for lateral oscillation to damp to half amplitude, T, /2

-

P rn 'I2

c 2 cycles required for lateral oscillation to double amplitude, -

P r o l l i n g moment c , rolling-moment c o e f f i c i e n t , c2 qSb ac1

-, per radian

c z B as

3%

C -, per r a d i a n

lg, a s , acz

-, per radian

C IP 2v *z

-, per radian

‘2r r b a , wing moment yawing-moment c o e f f i c i e n t , Cn

- qSb

acn

-, per r a d i a n

l -

CnP as

acn

Cn -, per radian

6a a s ,

-, per r a d i a n

2v

acn

-, per r a d i a n

cnr side f o r c e side-force c o e f f i c i e n t , CY

ss

&Y

-, per radian

a$

d D

d i f f e r e n t i a l operator, -

d t moment of i n e r t i a about longitudinal p r i n c i p a l axis, s l u g - f t 2 1% moment of i n e r t i a about v e r t i c a l p r i n c i p a l axis, s l u g - f t 2 1% moment of i n e r t i a about longitudinal s t a b i l i t y axis, I X

~ Q c o s ~ q + IZosin2q, slug-ft2

moment of inertia about vertical stability axis,

I cos27 + Iksin2q, slug-ft2

ZO product of inertia with respect to longitudinal and vertical

stability axes, (Izo - Ixo)sin 7 cos 7, slug-ft2

Mach number period of lateral oscillation, sec wing area, sq ft time required for lateral oscillation to damp to half amplitude, sec time required for lateral oscillation to double amplitude, see

true airspeed , ft/sec

indicated airspeed, knots stalling speed in landing configuration (power off, gear down, high-lift devices at landing setting) weight, lb wing span, ft acceleration due to gravity, 32.2 ft/sec2 pressure altitude, ft

J i

mass, slugs rolling angular velocity, radians/sec dynamic pressure, lb/sq ft yawing angular velocity, radians/sec time, see sideslip angle, radians total aileron deflection, positive for right aileron down, radians pilot-applied total aileron deflection, radians

.IIo

rudder deflection, positive for trailing edge left, radians pilot-applied rudder deflection, radians I - servo-applied rudder deflection, radians inclination of the longitudinal principal axis with respect to the flight path, positive when the principal axis is above the flight path at the nose ratio of air density at test altitude to that at sea level bank angle, radians ratio of bank-angle amplitude to sideslip amplitude for the oscillatory mode

-- Io1 57.3 deg

I P I v &' ft/sec

angle of yaw, radians EQUIPMENT AND INSTRUMENTATION Because descriptive material on the variable-stability ~ 6 ~ - 3 airplane and servo equipment already has been published, only brief discussions of special additions to the equipment are included in this report. However, since published information on the variable-stability F-86A rudder-servo system is extremely limited, a relatively complete description of that apparatus is presented.

Variable-Stability ~ 6 ~ - 3 Airplane A photograph of the ~ 6 ~ - 3 variable-stability airplane used in the simulation programs reported herein is shown in figure 1 .

Servo equipment.- The apparatus for varying the dihedral effect of this airplane through servo actuation of the ailerons is described in detail in reference 7 . Brief descriptions of similar methods used to vary and to provide simulated enry Cnpy and C2 the stability derivatives CnRY P rough-air disturbances may berfound in references 2 and 8 .

In addition to the variable parameters mentioned above, two special features were included for use in studying individual stability and con- One provided artificial variation of rolling moment due trol problems.

.. e.. e.. e.

e.. e.

e . . Y..

.. NACA RM ~ 5 6 ~ 0 8

e . . . e e. e..

e. e.. e.

to pilot-applied rudder angle C ; the other allowed servo-applied roll ‘ 6 , damping C to be varied automatically as a function of lateral stick 2 P position. The ways in which these devices were used in the particular simulation programs are discussed later.

Recording instrumentation.- Where data records of specific flight maneuvers were desired, the following quantities were measured: yawing velocity, rolling velocity, sideslip angle, rudder-servo position, aileron-servo position, pilot-applied rudder deflection, and pilot- applied aileron deflection. These quantities were recorded by standard NACA photographic recording instruments synchronized by a 0.1-second instrument timer.

Flight conditions.- A l l simulation flights in the variable-stability F6F-3 were performed in the clean condition at the following airspeed and altitude : vi = 200 knots hp = 7000 feet Variable-Stability F-86A Airplane A photograph of the F-86A variable-stability airplane is shown in figure 2 and a two-view drawing is presented in figure 3.

Servo equipment.- The F-86A variable-stability servomechanism operates in essentially the same manner as the ~ 6 ~ - 3 equipment referenced above. In this airplane, however, only the rudder and rudder tab are driven automati- cally and the primary power used is hydraulic rather than electric. A s in the F6F-3, mechanical differentials are used in the rudder and rudder-tab control systems. The yawing-moment derivatives affected are Cn , Cnr, cnp, B

and Cn . Brief information on the F-8& rudder servo may be found in

6a reference 9.

The rudder servo system installed in the F-86A is of the electro- hydraulic type and incorporates a high-performance single-stage hydraulic valve as the controller. This type of system was selected mainly because of the large servo power requirements at the high airspeeds attainable with this airplane. A simplified block diagram of the electrical-signal portion of the installation is presented in figure 4.

The error-measuring portion of the rudder servomechanism includes a phase-sensitive power amplifier, which Senses the difference between the input and follow-up signals.

A typical input circuit consists of a precision-type a-c pickoff, powered by a 400-cycle carrier voltage and mechanically connected to a sensing device, such as a sideslip vane, rate gyro, or pilot's control stick. The output signal from this pickoff is amplified and fed through the pilot's servo-control console, where manual adjustment of servo gearing is made (for example, rudder angle per unit sideslip &rs/&3) . The individual signals are then summed demodulated ., and fed into the aforementioned phase-sensitive power amplifier. The resultant amplified error signal is then used to vary the field strength of the servo-valve torque motor, which positions the single-stage valve, driving the hydraulic servo actuator in the desired direction. A follow-up signal proportional to the servo-actuator movement reduces the error volt- age to zero when the servo reaches the desired position.

The important components of the hydraulic servo-drive system are shown The system operating pressure is supplied by an engine-driven in figure 5.

variable-displacement pump and is regulated to 2700 pounds per square inch by a pressure relief valve. Hydraulic pressure to various parts of the system is controlled by three solenoid-operated two-position valves.

Valve 1 (fig. 5) controls pressure to the servo valve (that is, on or off), while valves 2 and 3 control pressure to the servo actuator. The valves are shown energized (pressure on) and the system is shown in normal opera- tion responding to a "right rudder" command signal. Dashed lines represent corresponding valve positions for the pressure-off condition.

During normal shutdown of the system, operation of the hydraulic- pressure switch by the pilot immediately grounds all inputs to the power amplifier, except for the follow-up signal. This causes the rudder servo to drive to a neutral position under normal hydraulic pressure. After a time delay of about 0.15 second, the locking-solenoid plunger (fig. 5) engages the servo-actuator unit and valves 1, 2, and 3 rotate simultane- ously to the de-energized position. In the event of failure of airplane primary power, valves 1, 2, and 3 operate immediately and the pilot must engage the locking-solenoid plunger by movement of the pedals in order to return the rudder to neutral.

The mechanical differential used in the combined pilot and servo rudder-control system is shown schematically in figure 5 and a cutaway isometric assembly drawing is presented in figure 6. From these two figures, the desired differential action c m be seen. Normal rudder control remains essentially intact; the only alteration was to thred each rudder-control cable from the first guide pulley around the float- ing center pulley, making a 1 8 0 ' wrap angle, and back through the second guide pulley to the rudder control sector. Thus, if the pedals are held fixed, motion of the floating center pulley results in a proportional displacement of the rudder and, similarly, if the center pulley is fixed, Therefore, any pedal motion results in normal actuation of the rudder.

e e.. a.

NACA RM A56CO8

a

e . a .

a 0 . .

.. e.. e.. a.

movement of the rudder ( 6 , ) is the algebraic sum of the angle called for by the pilot (8, ) and that caused by displacement of the center pulley P 7 which is forced to move with the servo actuator.

The aerodynamic hinge moments due to 6rs, which otherwise would be fed back to the pilot, are balanced by driving the rudder tab in response to motions of the servo. A s shown in figure 5, this was accomplished by means of a hydraulic tab actuator connected in series with the rudder- servo actuator. The necessary tab-to-rudder gearing was obtained by proper selection of tab-actuator piston area and by increasing the tab area about l 5 O percent. Normal tab adjustment by the pilot was retained by mounting the hydraulic tab actuator in series with the production lead- screw-type electric actuator.

Photographs of the variable-stability F-86A cockpit interior, showing the important pilot-operated servo controls, are presented in figure '7.

The recording-instrument control units and hydraulic-pressure control switch (on the stick), as well as indicators for sideslip and rudder-servo error signal, are shown in figure 7 ( a ) . The rudder-servo control panel is located on the right-hand side of the cockpit and is shown in figure 7(b).

Indicators for servo position and hydraulic pressure are included, as well as the servo power switches and knobs for setting the variable-stability parameters. Sine-wave and gust disturbances are provided by deflections of the rudder (through the servo) in response to an electrically driven cam. The F-86A gust generator is similar to that used in the F6F-3, except that signals from two cams driven at different speeds are combined to obtain random inputs. (This method greatly increases the time required for the gust pattern to repeat.) The frequency and amplitude con-crols for this sine-gust generator are shown in figure 7(b). This figure shows also provision for later installation of an aileron-servo systc>m.

Servo-system operation.- When the F-86A rudder-servo system is operated in flight, the electrical circuits are energized by setting the master-power and rudder-servo switches to the on position. Ammctprs which indicate the rudder-servo error signal reduce the possibility of abrupt servo motions which might occur as hydraulic pressure is turned on with large inputs to the servo valve. This error signal may be reduced to zero by the pilot, through use of centering potentiom-ters located on the servo control panel (fig. 7(b)). The servo drive system is t . n c . r g i z t d when the pilot depresses the hydraulic pressure switch on the control stick. Desired changes in the variable-stability parameters can then bc made by setting the selector knobs to appropriate positions. Each knob provides, in addition to the normal ~ - 8 6 ~ value, four increased values and four reduced values of a particular,parameter. Estimatfd ranges of the F-8a variable-stability parameters (based on control cffclctiveness and ground-measured servo gearings), as well a b those for thF FbF-3, are given in table I.

2K

0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 Recording instrumentation.- During flights in the variable-stability F-86A7 the following quantities were recorded by means of an 18ichannel photographic oscillograph: yawing velocity, rolling velocity, sideslTp angle, bank angle, normal acceleration? lateral acceleration, total rudder deflection? rudder-servo position, pilot-applied rudder deflection, total rudder-tab deflection? aileron deflection, rudder-servo error voltage, and NACA recording instruments were rudder-servo follow-up voltage. Standard used to measure pedal force and lateral stick force. The three film records thus obtained were synchronized by means of a 0.1-second instru- ment timer.

Flight conditions.- Standardized speeds and altitudes used l i n F-86A variable-stability flight tests are listed as follows: hP7 M ft

-

10,Ooo 0 . 6 0 10,000 0 . 8 0 35,000 0 . 8 0 Variable static and dynamic stability characteristics.- The effects on pedal force of artificial changes in static directional stability

cnB

and displacement as functions of sideslip are shown for the F-%A in fig- 8 . Time histories of lateral oscillations (returns from steady side- ure slips, pilot's controls restrained) with various

cnP and cnI- knob

settings are presented in figure 9. When attempts are made to change the setting alone, large oscillation period through variations in

cnB

changes in damping also occur. This effect is attributed to changes in Cnr resulting from the small phase differences between the p signal and servo-applied rudder deflection 6rs. In order to show the effect setting on period in figure 9 without variations in damping, of CnP compensating Cnr settings were used as indicated. Similar time histo- ries showing effects of changes in setting alone are presented in Cnr figure 1 0 .

SIMUIATION PROCEDURE Predicted controls-fixed lateral oscillatory characteristics and information on any unusual stability or control behavior which might be expected were furnished by the manufacturer in tabular or time-history form. Where these characteristics were not already available? the 1 - necessary stability derivatives and mass parameters were obtained from the manufacturer for use in calculating the lateral period, damping, d and the ratio of bank angle to equivalent side velocity The I ql/IVe I.

method used to calculate the period and dam ing was similar to that

presented in reference 10. The ratio I c p ] / p B I, from which 1 ( P I / ] Ve I

was obtained, was calculated as shown in the appendix by use of the sta-

bility system of axes (ref. 10). Flight values of I cp ] / I p 1 were derived

.

from measurements of made with respect to the airplane body

I p l / l p I

axes; however, at the test flight conditions, any discrepancies resulting from the use of different axes systems were considered negligible.

To provide the characteristics of interest, appropriate variable- stability knob settings were chosen from documented results of previous or from calibration flights made immediately prior to the simu- flights was lation. In cases where unusual airplane response to control inputs anticipated on the prototype, variable-stability settings were selected to give the best approximation of time histories of the predicted motion.

~ 6 ~ - 3 or F-86A, With either the it was not possible to duplicate the moments of inertia nor to cover the full range of performance of the simulated airplanes.

Because of the diversity of problems of interest to the various contractors and the specialized techniques required to investigate those problems, no standardized flight procedure or test maneuvers were employed.

In all cases, one or two company engineers who were well acquainted with development of the prototype accompanied the visiting test pilot to aid in arranging and evaluating the simulation program.

Following preliminary discussion of the program with the contractor representatives and familiarization of the visiting pilot with the variable-stability airplane and associated servo equipment, the planned simulation flights were made. During these flights, the visiting pilot commented on each condition simulated, either in writing or by radio.

Where desired, instrument records of specific flight maneuvers also were obtained.

On completion of the simulation flights, discussions wert held with the contractor representatives for the purpose of reviewing the p i l o t ' s opinions of the particular conditions simulated and offering suggestions for improvement of marginal or unsatisfactory behavior through possible design changes or artificial stability augmentation.

RESULTS Airplane A Airplane A was designed as a high-speed flight-research vehicle powered by two turbojet engines. A two-view drawing and table of prin- cipal dimensions of this airplane are presented in figure 11.

0 0

NACA RM ~ 5 6 ~ 0 8 .. . 11

0 .

0 0 0 0 0 0 0.0 0 .

The predicted lateral oscillatory characteristics ( l/Cl,2 and

I cp 1 / 1 Ve [ ) of airplane A, calculated for several Mach numbers at altitudes

of 3,000 and 35,000 feet from stability derivatives and mass parameters furnished by the contractor, are compared in figure 12 with those meas- ured in the variable-stability ~ 6 ~ - 3 . Although airplane A was not designed a s a.n operational type, the configuration may be representative of future fighter designs. For this reason, the pilot-opinion boundaries of reference 2 are included for comparison. In figure 12, it may be seen that the predicted characteristics of airplane A at all flight conditions considered were well simulated.

As simulated in the ~ 6 ~ - 3 , all lateral oscillatory characteristics of the basic airplane A corresponding to the 3000-foot altitude were considered satisfactory by the contractor pilot. The characteristics M = 0 . 6 0 and 0.90 at 35,000 feet were rated tolerable, predicted for while those for M = 1 . 0 0 were considered intolerable due to the very lightly damped oscillations.

The opinions given by the contractor pilot in the simulation of airplane A tended to be more lenient with regard to high oscillatory roll coupling and low damping than would be indicated by the pilot- opinion boundaries of reference 2 . This might be reasonable due to the .

intended use of airplane A as a research airplane, wherein stringent lateral-oscillation requirements such as those placed on operational aircraft would not be expected to apply. This pilot also had previous flight experience with oscillations having high oscillatory roll coupling and low damping in an earlier research airplane.

Airplane B A two-view drawing and table of principal dimensions of airplane B are presented in figure 13.

Lateral oscillations.- The lateral oscillatory characteristics of airplane B, calculated for the design cruise (M = 0.90, hp = 35,000 feet) and landing-approach conditions by the contractor, are compared in fig-

ure 14 with corresponding values of 1/C,,2 and I (pl/lvel measured in the

variable-stability ~ 6 ~ - 3 and with the pilot-opinion boundaries of refer- Comparison with the boundaries of reference 2 indicated intol- ence 2 .

erable damping and oscillatory roll-coupling characteristics for the basic airplane B in both the cruise and landing-approach conditions; the contractor was interested in assessing the effects of reducing dihedral effect C z p and using a yaw damper to increase Cnr in the manner shown in figure 14.

The predicted oscillatory roll-coupling characteristics of airplane B and three values of Cnr were not simulated as closely with design c 2 B

: : .: : : NACA RM ~ 5 6 ~ 0 8

0 . 0 . . . 0 . .

0 . 0.. .

as desired; however, the damping in these three cases was represented well by the F6F-3. Figure 14 shows that reasonable simulation of the predicted lateral oscillatory characteristics was provided for the reduced C conditions.

For the simulated design cruise condition, opinions of the contractor pilot indicated that airplane B would have intolerable lateral oscillatory characteristics. This appeared safe to assume since the opinion was based on an amount of oscillatory roll coupling less than that actually pre- dieted. With one-half design CzB and 3 and 6 times design C n , , the resulting characteristics were Considered satisfactory. The zero c z B condition for all three values of was also rated satisfactory from Cnr the lateral-oscillation standpoint but would probably be undesirable for other reasons, which will be discussed later.

No formal opinion was given regarding lateral oscillatory character- istics in the landing-approach configuration; however, an intolerable rating such as that assigned to the basic airplane in the cruise condition would be expected.

Roll due to rudder deflection.- An unusually large value of rolling moment due to rudder deflection was predicted for airplane B and ‘ 2 8 , the contractor was concerned about possible adverse effects on the roll response to abrupt rudder deflections, especially in the landing approach.

Analog-computer studies by the contractor indicated initial adverse rolling tendencies in response to rudder step inputs, and a flight investigation of these motions was considered desirable. To investigate this feature, could the ~ 6 ~ 3 variable-stability equipment was modified so that ‘ ‘ 6 , be varied in flight. This was accomplished by installation of a precision- type a-c pickoff on the pilot’s input to the rudder-servo differential; the resulting signal was fed into the aileron servo system through a manual gain control.

The motions indicated by the analog time histories obtained by the contractor were approximated in flight through use of appropriate variable- stability settings in the F6F-3. Figure 15 presents flight time histories of bank a n g l e in response to step-type rudder deflections for the variable- stability F6F-3 set up to simulate airplane B with one-half design

%

and with three values of C . In each case, sufficient rudder angle

‘ 8 , was applied to trim the airplane at about 5 ’ sideslip.

of‘ the F6F-3, AS seen from the curve for the normally small C l 6r figure 15, roll was in the direction expected for positive dihedral effect (left roll for right sideslip) and no initial adverse rolling motion wits present. The middle curve indicates the type of rolling motion obtained in the ~ 6 ~ - 3 with a value of C necessary to simulate the response I - ‘ 5 , I predicted for airplane B . The initial adverse roll shown was noticeable to the pilot but was not considered sufficiently large nor persistent to c- interfere seriously with control of the airplane. With the planned increase in yaw damping, three times design Cnr (not shown), this one- C z p configuration was considered satisfactory by the con- half design tractor pilot and appeared to be a reasonable design goal since a satis- was also given to the lateral oscillatory characteristics factory rating

themselves. With the large adverse C , the F6F-3 rolled in the adverse

‘ 8 , direction throughout the maneuver. This continued adverse roll was definitely undesirable in the pilot’s opinion.

The pilot opinions associated with the motions shown in figure 15 indicated that the limiting case of tolerable adverse would be C 2 6, one in which no sustained adverse roll occurs for a gifren value of C z e .

For designs similar to airplane B, this might serve as a rough criterion for determining maximum allowable C .

‘ 8 , Airplane C A two-view drawing and table of principal dimensions of airplane C are presented in figure 1 6 .

To provide improved pilot visibility and permit the use of short landing gear by avoiding large fuselage angles of attack during the approach and landing, this carrier-based day fighter featured a two- position variable-incidence wing (-lo incidence for cruise, 7 O for landing).

Lateral oscillations.- Calculations made by the contractor for the basic airplane in the landing condition indicated the undesirable lateral oscillatory damping and roll-coupling characteristics shown in figure 1 7 .

These characteristics were traced to the predicted high dihedral effect and the lack of favorable positiire inclination of the longitudinal prin- cipal axis of inertia with respect to the flight-path axis with the wing The contractor considered improving these character- at 7 O incidence.

istics by means of a yaw damper; calculations indicated that, although a substantial increase in damping could be provided in this manner, the objectionable high value of oscillatory roll coupling would still remain Further calculations showed that and even increase slightly (fig. 1 7 ) .

simultaneous improvements in the damping and roll-coupling characteristics of airplane C in the landing condition could be achieved by means of a roll damper which, through servo actuation of the ailerons in response to a roll-rate gyro, provides large stabilizing increments in the damping-in- roll derivative C Th use this type of damper 2 P in the landing approach and therefore was inEerested in comparing its effects with those of a yaw damper on pilot opinions of the associated lateral oscillatory behavior.

It is seen from figure 17 that reasonably good simulation of the predicted characteristics of the basic airplane C (no auxiliary damping) in the landing approach ms provided by the ~ 6 ~ - 3 for both approach speeds of 1.2 and 1.5 Vs Figure 17 also shows that, even though the roll-damper L' was not simulated as closely as desired, the relative condition at 1.2 Vs L effects of the yaw damper and roll damper were well represented by the F6F-3.

In the opinion of the contractor pilot, the damper-off condition would at both approach speeds due to the high oscillatory roll be intolerable coupling and poor damping. The characteristics associated with the yaw- damper condition at 1.2 VS although deep in the intolerable region of L ' reference 2, were considered marginally satisfactory. The characteristics represented by the simulated roll-damper point were felt to be highly satisfactory due to the large reduction in roll coupling which accompanied An additional condition (not shown) represented com- increased damping.

bined use of the yaw and roll dampers and was considered even more desira- c ble than with the roll damper alone.

Nonlinear roll damper.- While planning the use of a roll damper of the type mentioned, the contractor was aware that roll maneuverability would be impaired due to the high effective C with the roll damper ZP operating. To avoid this, it appeared desirable to vary the roll-damper gain as a nonlinear function of lateral stick position, as suggested in reference 11. In this way, maximum roll damping would be provided in steady flight or in mild maneuvers, and would be reduced to the normal- airplane value when the pilot applied large stick deflections in order to roll rapidly. To obtain pilot opinions of such a nonlinear roll damper, the ~ 6 ~ - 3 variable-stability equipment was modified by feeding the rolling- velocity signal to the aileron servo through a tapped potentiometer actu- ated by the stick. Several symmetric variations of servo-applied c2P thus were obtained, as shown in figure 18. Results of early flights using this device showed, as expected, that the desired high roll damping and roll maneuverability could be achieved. At the time of the simulation flights f o r airplane C, the variation indicated by the solid line C Z P (variation 1-3, fig. 18) was considered optimum for rapid roll maneuvers in smooth air by the contractor pilot and two NACA pilots. Subsequent flights by one NACA pilot have indicated that a variation providing full roll damping for moderate stick travel (2-2) might be more suitable for other conditions, such as flying in rough air.

Yaw due to aileron deflection.- Subsequent to the simulation flights made in the ~ 6 ~ - 3 , analog-computer studies of airplane C by the manufac- NACA RM ~ 5 6 ~ 0 8 acceleration of the order of 0.2g immediately following abrupt aileron deflections at high speeds with the rudder fixed. This lateral accel- eration resulted largely from yawing acceleration combined with a cockpit location 23 feet ahead of the airplane center of gravity. The initial yawing acceleration was traced to a large favorable variation of yawing moment with aileron deflection Cn6- (positive yaw in response to aileron -tz deflection initiating positive roll), a characteristic of certain inboard aileron installations. The contractor felt that this abrupt lateral accel- eration would be particularly disturbing to a pilot and hence planned to actuate the rudder in response to aileron deflection (i.e., left rudder for right aileron) over a range of low angles of attack. Analog studies indicated that the lateral-acceleration response would be considerably reduced by this method, and, accordingly, flights were made in the variable-stability F-86A to simulate the predicted behavior of airplane C, both with and without the aileron-rudder interconnection.

on the variable-stability F-86A in flight Provision for varying C nga had already been made by feeding a signal proportional to lateral stick deflection into the rudder-servo summing amplifier through a manual gain control. Settings for Cn were chosen on the F-86A which provided sa close simulation of the lateral-acceleration responses of airplane C The lateral oscillatory characteristics of airplane C were (fig. 19).

approximated reasonably well by a moderate reduction in directional stability CnP Opinions of the contractor pilot indicated that the lateral handling qualities of the F-86A set up to simulate the basic condition (without the aileron-rudder interconnection, fig. l 9 ( a) ) were not objectionable; in fact, this condition was actually preferred over that simulating the improved condition (with the aileron-rudder interconnection, fig. l 9 ( b) ) , due to better roll maneuverability which resulted from the favorable side- slip and positive dihedral effect. This opinion was also attributed partly to his previous experience in another fighter-type airplane which exhibited large lateral-acceleration responses ( sometimes estimated at lg) in abrupt aileron rolls. However, during simulated air-to-air gunnery runs (similar t o those employed in ref. 8) in the variable-stability F-86AY the con- tractor pilot encountered more difficulty tracking in the basic configura- This agreed with gun-camera records and tion than in the improved case.

opinions of NACA pilots obtained in subsequent F-86A flights in which the same variable-stability settings were used.

Airplane D A two-view drawing and table of principal dimensions of airplane D are presented in figure 20.

The predicted lateral oscillatory characteristics of airplane D (measured from analog time histories obtained by the contractor and cal- culated from stability derivatives and mass parameters furnished by the contractor) are compared in figure 21 with those measured in the variable- stability ~ 6 ~ - 3 . Several speeds and altitudes, corresponding to power- approach and combat-cruise conditions, are represented. A similar com- parison is made between predicted airplane D combat-cruise characteristics simulated in the variable-stability F-86A and the measured F-86A charac- teristics in figure 22. In both figures 21 and 22, the pilot-opinion boundaries of reference 2 are included for comparison.

In general, simulation of the predicted characteristics of airplane D was satisfactory with both F6F-3 and F-86A variable-stability airplanes.

Because of the large number of conditions involved, no attempt was made to simulate each point specifically. Instead, the no-damper conditions of major interest were approximated by a series of conditions (points 1, 4, 21) having low damping and covering a large range of 6, 9, and 12, fig.

These were then used as basic points to demonstrate the effects [cp\/lvel.

of various dampers.

As simulated in the F6F-3 (fig. 21), the combat-cruise, damper-off conditions of airplane D were considered marginal (point 1) to objection- able (point 4) by the contractor pilot on the basis of moderate oscillatory roll coupling combined with low damping. With a yaw damper (approximated by points 2 and 5) these combat-cruise conditions were considered to have satisfactory damping; however, the contractor pilot felt that improvement could be made, especially in rough air, by reducing oscillatory roll coupling through use of lower dihedral effect.

The power-approa.ch,damper-off conditions (points 6, 9, and 12) drew very unfavorable opinions from the contractor pilot because of the high and generally poor damping. The addition of a yaw damper in

19 l/lve I

these power-approach conditions (points 7, 10, and 1 3 ) increased the damping to an acceptable level. Oscillatory roll coupling I cp I / I Ve I , however, was still considered marginal to objectionable in rough air.

In addition to the yaw-damper conditions, effects of a roll damper (such as proposed for airplane C) were investigated by the contractor pilot in both simulated combat-cruise and power-approach conditions (points 3, 8 , and ll, fig. 21). Only a slight effect of the roll damper was noticed in the combat-cruise condition indicated by points 1 and 3, probably because the oscillatory roll coupling without the roll damper was small. In the power-approach condition, the roll-damper effect was considcred quite favorable at a speed of 1.4 V only a small improvemcnt while at 1.1 Vs sL7 L over the damper-off condition was noted. In over-all suitability, howevcr, the yaw damper was preferred to the roll damper by the contractor pilot.

A l l conditions simulated in the variable-stability F-86A (fig. 2 2 ) were considered satisfactory from the standpoint of oscillatory roll 3K . 0 0.. e .

, . . e 0 .

NACA RM ~ 5 6 ~ 0 8 : : . e e .

e* e . . 0 coupling. The yaw-damper-off conditions had undesirably low damping, resulting in steady snaking oscillations in rough air, but they were felt to be completely satisfactory with addttion of the yaw damper.

.

Airplane E A two-view drawing and table of principal dimensions of airplane E (as simulated in the variable-stability ~ 6 ~ - 3 ) are presented in figure 23.

Airplane E was a high-speed bomber configuration for which unusual lateral oscillatory characteristics (long period, unstable oscillations A with moderate roll coupling) were predicted in the take-off condition.

second and perhaps more serious problem was a progressive reduction of Cnp, which was expected to occur when the static directional stability design bombing-run Mach number was exceeded by more than 10 percent. Use of a directional stability-augmenting device was planned; however, the contractor was concerned that the resulting long-period unstable lateral oscillation (or even a rapid aperiodic divergence) might be objectionable or dangerous in the event of stability-augmenter failure in this critical flight condition.

In the ~ 6 ~ - 3 , variable-stability knob settings were chosen to simulate the predicted lateral oscillatory characteristics of interest was extended to provide the and the available range of reduced

CnP

desired simulation of low static directional stability.

Lateral oscillations.- The predicted lateral oscillatory character- istics of airplane E (calculated by the contractor and the NACA) are Simulation of compared in figure 24 with those measured in the F6F-3.

the predicted oscillatory characteristics in the take-off condition (M = 0.35, hp = 0) and after refueling (M = 0.60, hp = 30,000 feet) was reasonably good. The remaining conditions at high speeds and high alti- tudes (still with positive C ) were not simulated as well as desired.

nP

In the opinion of the contractor pilot, the lateral oscillatory characteristics predicted for the take-off condition were intolerable.

This opinion was based on the divergent oscillation, which was felt to be especially objectionable in view of anticipated flight near ground as experienced in the F6F-3, were level. The remaining conditions, considered tolerable from the lateral-oscillation standpoint.

Low directional stability.- Some effects of neutral static directional stability are shown in figure 25. Presented are time histories of pilot- applied control deflections and airplane motions with cockpit controls held fixed and with the pilot attempting to hold a steady course in simulated .

instrument flight. Under controls-fixed conditions, the airplane motion 1 8 NACA RM Ag6C08 involved an unstable la.tera1 oscillation of very long period (P =: 14 sec).

Figure 25 also shows that reasonably steady flight could be maintained under instrument conditions; however, considerable attention to aileron and rudder control was required. In general, the contractor pilot felt that flight of the variable-stability ~ 6 ~ - 3 in the region around neutral directional stability (where T2 z z 12 see) w a s not necessarily dangerous but would be bothersome and fatiguing over extended periods of time.

Subsequent to the simulation program involving the ~ 6 ~ - 3 , changes in the design of airplane E had been made. The low directional-stability problem was still expected to occur in the high-speed cruise condition and, in addition, strong favora.ble C (as in the case of airplane C) and low nga roll damping C were indicated by wind-tunnel tests and preliminary cal- JP culations. The contractor was interested mainly in obtaining some indica- with stability-augmenting devices inopera- tion of minimum acceptable

CnP

tive in the high-speed cruise condition, in the presence of predicted Cn 6a The variable-stability F-86A was chosen as the test vehicle in and C 2 P ' this case because of its greater speed and altitude capabilities. A two- view drawing a.nd table of principal dimensions of airplane E as simulated in the variable-stability F-86A are presented in figure 26.

At the test flight conditions of M = 0.80 and hp = 35,000 feet, F-86A variable-stability knob settings were chosen to give the best approximation of predicted controls-fixed lateral oscillatory behavior of airplane E.

Since the variable-stability F-86A was not equipped with an aileron servo drive system, it was not possible to make significant changes in the roll-

was artificially varied to give the damping derivative C However , Cnp

2P * pilot an impression of low roll damping. This was accomplished by select- servo gearing which provided, for example, a right yawing ing a Cnp moment in response to a right roll initiated by the pilot. This in turn resulted in a left sideslip and an additional right rolling moment due to positive dihedral effect, giving the desired end effect of an increase in roll velocity for a given stick deflection (at least during the middle of the roll transient). Through proper selection of Cn and C portion nP 6a gearings,in addition to Cnr and Cn it was possible to obtain a reason- B' able simulation of the over-all lateral and directional response to control inputs predicted for airplane E.

Once the appropriate variable-stability servo gearings had been established, a series of flights were made in which the contractor pilot explored the interesting range of low directional stability Cnp. From these flights, the pilot first concluded that a value of Cn corre- P sponding to 25 to 30 percent of that of the normal F-86A was about the minimum acceptable.

However, as he gained experience in this region of In cases where CnB, he felt that still lower values might be tolerated.

, wide differences in mass and performance characteristics of the simulator and simulated airplanes are involved, such as in the present example, it would not appear wise to apply values of critical stability derivatives (such as minimum C n , ) estimated in the simulation flights as direct numerical criteria in the prototype design.

Airplane F A two-view drawing and table of principal dimensions of airplane F are presented in figure 27.

Airplane F was designed as a two-place, jet-powered trainer which could accommodate variable-stability servo equipment for the purpose of simulating dynamic behavior of modern fighter aircraft about all three stability axes. The contractor pilot flew the F6F-3 to gain familiarity with the variable-stability concept and for simulation of the predicted lateral oscillatory characteristics of airplane F.

The predicted lateral oscillatory characteristics of airplane F, calculated by the contractor for sea-level climb and landing conditions, are compared in figure 2 8 with those measured in the variable-stability ~ 6 ~ - 3 . Although airpla.ne F was not a fighter type itself, the pilot- opinion boundaries of reference 2 are included for simple comparison.

The characteristics represented by the F6F-3 points in figure 28 were evaluated by the contractor pilot both in smooth air and with the rough- The a.ir simulator (ref. 8) set at moderate aileron and rudder amplitudes.

pilot considered the simulated lateral-oscillatory characteristics of airplane F in the climb condition (point 1, fig. 28) to be satisfactory, since good damping was present and he was able to hold a steady course Points 2 and 3 of figure 28 bracketed the pre- in simulated rough air.

dicted damping characteristics of airplane F in the landing condition;

the pilot rated point 2 as he did point 1 - very good damping and easy to

hold on course in rough air. Point 3 was acceptable, though becoming difficult to control in rough air, having marginally satisfactory damping.

The very low roll coupling of airplane F in the climb condition could not be simulated as closely as desired without the use of objectionable nega- tive dihedral effect on the F6F-3; however, previous pilot-opinion studies indicate that such differences in roll coupling are not critical in the Close simulation of the oscillation period in range considered (ref. 2).

the landing condition was sacrificed in order to preserve much of the high directional stability still present in that condition.

In addition to the lateral oscillatory characteristics presented in figure 28, analog time histories furnished by the contractor showed marked Accordingly, spiral divergence of airplane F in the landing condition.

a . a a o m a a a. * a NACA F M ~ 5 6 ~ 0 8 a * a * a. * a * a.

~ 6 ~ - 3 variable-stability settings providing mild to substantial spiral divergence were included in this flight program, but no exact simulation of the computed divergence of airplane F was attempted.

DISCUSSION The results of these various simulation programs have been discussed individually as they were presented. The present section provides a recapitulation of the more important information and experiences gained during these programs. Pilot opinions of the lateral oscillatory charac- teristics, an item of importance common to nearly all the airplanes studied, are discussed first; then follow the special problems which were of interest usually in individual cases.

Lateral Oscillatory Characteristics Written and verbal opinions expressed by the contractor pilots concerning the lateral oscillatory characteristics of airplanes A, B, C, D, and F (as simulated with the variable-stability ~ 6 ~ - 3 ) have been assembled and are shown qualitatively by the shaded areas in figure 29.

Included are the pilot-opinion boundaries of reference 2 and those presented in the current military specification (ref. 1).

The comments indicated by the shaded areas were obtained from information volunteered by each pilot during and immediately following the simulation flights. No formal procedure was used for obtaining pilot opinions; the pilots were not requested to answer standardized questions or to perform specific maneuvers. In most cases, the flight procedure was dictated by the particular problem being investigated.

In substance, reference 1 states that airplanes in the clean or the landing configuration (while not engaged in gunnery, bombing, or other critical duties) must have, in controls-fixed and controls-free lateral- directional oscillations, a value of the damping parameter 1/C,,2 not less than that represented by curve a of figure 2 9 . Reference 1 states further that if an artificial stabilization device is employed, 1 / C l l 2 with the device inoperative shall be at least 0.24 in all configurations, and shall be at least that represented by curve b in the power-approach configuration. In view of this consideration of artificial-stability devices, direct comparison between the two sets of boundaries presented in figure 29 is difficult because reference 2 considers only normal operation of fighter-type airplanes.

Most of the airplanes considered in figure 29 had predicted damping characteristics below bodndary b of.reference 1 only in the landing- pilots indicated that lateral approach configuration.

- oscillatory behavior characterized by shaded area 3 would require stability * augmentation in the landing approach; however, they did not indicate spe- cifically that such behavior would be unsatisfactory for a condition of I damper failure. It may be that in a detailed pilot-opinion survey (with I - appropriate questionnaire, rating scale, and flight procedure) such "emergency" considerations would result in less damping required for s G t T 6 f Z s t o i - y behavior than is specified by boundary b. Since the pilot- opinion data of figure 29 were among those considered in arriving at the lateral-oscillation specification of reference 1, the good agreement between the pilots' comments and boundaries a and b is not surprising.

It should be noted that the characteristics represented in fiere 29 involve only lateral-oscillation periods greater than 1.9 seconds (the minimum normally attainable with the variable-stability ~ 6 ~ - 3 ) .

A s indicated in reference 4, shorter periods associated with high-speed and low altitudes may place more stringent requirements flight at medium on damping and oscillatory roll coupling.

Special Problems

I -

Design information related to particular stability and control problems (other than lateral oscillatory behavior) investigated during these simulation programs is summarized in the following paragraphs.

R o l l due to rudder deflection.- Airplanes having unusually high values of C may exhibit adverse rolling tendencies in response to ' 6 ,

rudder deflections (such as those predicted for airplane B) , particularly

if dihedral effect is low. Pilot opinions associated with such motions simulated in the variable-stability ~ 6 ~ - 3 indicated that the limiting case of tolerable adverse C 2 would be one in which no sustained adverse 6r roll occurs for a given value of This might serve as a rough C z o .

criterion for maximum allowable C for designs similar to airplane B .

' 6 r Yaw due to aileron deflection.- In the variable-stability F-86A, yawing motions similar to those excited by deflection of inboard ailerons Reduction (airplane C ) were found to make air-to-air tracking difficult.

of these motions, simulating the effect of an aileron-rudder interconnec- tion, brought about improvement in tracking performance.

Nonlinear roll damper. - For certain airplanes (e.g., airplane C ) , use

of a roll ( C ) damper to provide improved damping and reduced oscillatory 2 P Pilot opinions roll coupling in the landing approach appears promising.

obtained in flights of the variable-stability F6F-3 indicated that reduc- tion of damper-applied with lateral stick deflection is desirable in czP .I . a a . a a a a . . . a , a a a a a . a a .

NACA RM A5608 a. .a. .a. a.

order to maintain good roll performance. The manner in which C should 2P be varied with stick deflection appears to be similar to that shown by curve 1-3 or 2-2 of figure 18.

Low directional stability.- Tests made in the variable-stability F6F-3 and F-86A indicated that substantial reductions in directional stability could be tolerated, though it was felt that flight under such condi-

CnP

During tions for extended periods of time would be fatiguing to a pilot.

flights in the F-86A, values of as l o w as 25 percent of the normal

CnP

value were tolerated. Caution should be used, however, in applying figures such as this as direct design criteria when wide differences in mass and performance characteristics occur between the simulator a n d prototype airplanes.

CONCLUDING REMARKS Through use of the NACA variable-stability ~ 6 ~ - 3 and F-86A airplanes, flight experience was obtained with lateral dynamic characteristics repre- sentative of those predicted for six prototype airplanes. From these studies, it was found that where unusual stability or control-response characteristics were predicted, or where auxiliary damping devices were to be employed, the test pilots who were to fly these airplanes gained familiarity with the trends in lateral behavior and were able to define ranges of acceptable characteristics. The flight experience obtained was in most cases directly applied to particular flying-qualities problems associated with the individual prototype development programs.

In the investigation of new fighter designs by means of variable- stability airplanes, higher performance airplanes should be used because of the greater speed capabilities and, hence, shorter oscillation periods attainable. This would allow more satisfactory simulation of high-speed lateral oscillatory characteristics than could be provided by the variable- stability ~ 6 ~ - 3 .

Ames Aeronautical Laboratory National Advisory Committee for Aeronautics Moffett Field, Calif., Mar. 8, 1956 APPENDIX METHOD USED TO CALCULATE

I Cp I / \ f3 I

I 1 / I

The lateral-oscillation bank-to-sideslip ratio l q . ) l / I i j I was

calculated, neglecting effects of airframe flexibility, from the following three linearized equations of motion referred to the stability system of axes presented in reference 10 (for level flight): Rolling moment (A,D~ + A,~D)V + ( A ~ D ~ + A , D ) J ~ + A,B = A,, Yawing moment (BID2 + B$)(p + (B4D2 + BgD)l(l + B& = B,, Side force where A, = Ix A, = -qSb(b/2V)Czp A , = I= A5 = -qSb(b/2V)Czr AS = -qSbCz P Bl. = 1x2 B, = -qSb(b/2V)Cnp B, = Iz B, = -qSb(b/2V)Cnr B, = -qSbCn B

c , = -w

C, = C = mV

c , = -qscyB

A,, = rolling-moment disturbance = 0 B,, = yawing-moment disturbance = 0 I C,, = side-force disturbance Solutions for rp and fi, respectively, in response to a unit side- force disturbance, are expressed as

0 A4D2 + A5D A S

0 B4D2 + B5D B S

1 C5D CeD + C, (4) r p =

A,D, + A,D A,D, + A$ o

B,# + B,D B , D ~ + B@ o

c3 C5D 1

p = - ( 5 ) where

A , D ~ + A,D A , D ~ + A,D A S

F(D) =

B , D ~ + B,D B , D ~ + B ~ D B S

C5D C a D + C s

c3

Expanding the determinants, dividing expression (4) by expression ( ? ) , and simplifying results in

For the free oscillation, let D = a + ib, where a and b are the

. real and imaginary parts, respectively, of the complex roots of the char- acteristic equation (from F(D) = 0). Then

I ? = a2 - h2 + Pahf

and 3 3

D = a - 3ab2 + (3a2b - b3)i

Expression (7) then reduces to: where

Q1 = a ( A , B , - AsB4) + (A5B, -

R , = b(A4BS - A94)

Q = (a3 - 3 a b 2 ) ( A , B 4 - A , B , ) +

(a2 - b2)(A1B5 - AsB, + A $ * - A4B2) +

a ( A , B 5 - A92)

R , = (3a b - b3)(A1B4 - A , B , ) +

2ab(A,B5 - A , B , + A s 4 - A , & ) +

b(A2B5 - A921

The ratio of the amplitudes of cp and f3 at a given instant of time during the oscillation is then expressed as .

The parameter was then calculated using the relation

I cp l/lve I

REFERENCES 1. Anon.: Military Specification - Flying Qualities of Piloted Airplanes.

MIL-F-8785(ASG), Sept. 1, 1954. (Amendment 1, Oct. 19, 1954).

2. Liddell, Charles J., Jr., Creer, Brent Y., and Van Dyke, Rudolph D.,Jr.: A Flight Study of Requirements for Satisfactory Lateral Oscillatory Characteristics of Fighter Aircraft. NACA RM ~51~16, 1951.

Bull, Gifford: A Flight Investigation of Acceptable Roll to Yaw Ratio 3.

of the Dutch Roll, and Acceptable Spiral Divergence. Rep. No.

TB-574-F-6, Cornell Aero. Lab., Inc., Buffalo, Feb. 12, 1952.

4. Moore, Norton B . : Artificial Stability Flight Tests of the XF-88A Airplane. TR 52-298, Wright Air Dev. Center, July, 1954.

5. Graham, Dunstan, and James, Clarence W . : A'Flight Investigation of

Minimum Acceptable Lateral Dynamic Stability. Rep. No. TB-574-F-3, Cornell Aero. Lab., Inc., Buffalo, Apr. 30, 1950.

6. Kauffman, William M., and Drinkwater, Fred J., 111: Variable-Stability Airplanes in Lateral-Stability Research. Aeronautical Engineering Review, vol. 14, no. 8, Aug. 1955, pp. 29-35.

7. Kauffman, William M., Liddell, Charles J., Jr., Smith, Allan, and Van Dyke, Rudolph D., Jr.: An Apparatus f o r Varying Effective Dihedral in Flight with Application to a Study of Tolerable Dihedral on a Conventional Fighter Airplane. NACA Rep. 948, 1949.

8. McNeill, Walter E., Drinkwater, Fred J., 111, and Van Dyke, Rudolph D., Jr.: A Flight Study of the Effects on Tracking Performance of Changes in the Lateral-Oscillatory Characteristics of a Fighter Airplane. NACA RM A53H10, 1953.

5. Smith, G. Allan, and Belsley, Steven E.: Artificial Lateral Aircraft Stability by Servo Control and Some Artificial Feel Experiments.

BuAer. Rep. A~-61-5 pt. 1, Report of the Second Piloted Aircraft Flight Control System Symposium, June 2-5, 1952, pp. 47-63.

10. Heinle, Donovan R., and McNeill, Walter E.: Correlation of Predicted and Experimental Lateral Oscillation Characteristics for Several Airplanes. NACA RM ~52506, 1952.

The Effects on Dynamic 11. Schade, Robert O., and Hassell, James L., Jr.: Lateral Stability and Control of Large Artificial Variations in the Rotary Stability Derivatives. NACA Rep. 1151, 1953 (Supersedes NACA TN 2781).

m a m m a a a m a . 0 a m a- 0 a a a a . a TABU I . - ESTIMATED RANGES O F VARIABLE STABILITY AND CONTROL PARAMETERS AVAILABLE ON ~ 6 ~ - 3 AND F86-A AIRPIANES F6F-3 F-86A1 Parameter Maximum Normal Minimum Maximum Normal Minimum 0.030 -0.002 0.127

- .080

- 6306

- -197

- .011

-.151 - .012

!

- .080

-9350 - -074

- .450 -1 .G2

- -385

---

.007 . O M lM = 0.80 at hp = 35,000 ft . ... 0.

0. ... 0

0 . * .

... .

- NACA -34 ~56~08 0 0 - 0 . 29 0 0 . . .

r

r

1 - 37.54' *

L- Sideslip vane

Figure 3.- Two-view drawing of t h e variable-stability F-86A airplane.

c W w * v) L m I I I I J

I -

I - C .- c 0 ) cn

t

1 7 A<

.- cn Q ..

L .- a .. 0.. . . . 0.. 0 .

0 . 0 . . ' : i - 0 . . .... 0 . .

NACA RM ~ 5 6 ~ 0 8 0 . 0 . . 0 . .

.. 0.. . . 0. 0 . 0.. 0 .

v) w W W a -I J 3 m m v) a a v) 0 v) > W J J a Lz W U I - - I Right ..

L e f t # U . .

ai (a) Maximum directional-stability setting.

..

E .- I c E 2 Right Right 2 Y- ..

. .

# .c 0 Y- # U L U a 2 Q, U Q U 2 L e f t L e f t Z U U U # K (b) Normal directional stability (inoperative).

I c

- .-

a Right Right io0 IO0 Left L e f t 4 0 4 4 0 4 Left Right L e f t Right

Sideslip angle from trim, p, deg

(c) Minimum directional-stability setting.

Figure 8.- Directional stability and control characteristics during steady, straight sideslips; variable-stability F-86A, M = 0.80, hp = 35,000 feet.

0 0 0 . 0 0 0 0 0 0 . 0 0 NACA RM ~ 5 6 ~ 0 8 0 0 . 0 0 . 0 0 0 Right

0 -"------

L e f t (a) Directional-stability setting: maximum stabilizing.

Directional-damping setting: maximum stabilizing.

I O Right L e f t Right P , deg L e f t normal (inoperative) .

(b) Directional-stability setting: Directional-damping setting: normal (inoperative).

IO Right 6, deg P ' L e f t I O Right P , deg 0 L e f t 0 I 2 3 4 5 6 7 8 9 IO Time, sec ( C ) Directional-stability setting: intermediate destabilizing.

Directional-damping setting: intermediate destabilizing.

Figure 9.- Time histories of typical controls-fixed lateral oscillations with yaw due to roll rate normal; variablc?-stability F-86A, M = 0.80, hp = 35,000 feet.

39 I

I O Right 0 .

Left -

IO Right Left (a) Directional damping: intermediate stabilizing.

I O Right I O Right L e f t (b) Directional damping: normal (inoperative).

I O Right Left IO Right Left 0 2 4 6 8 I O Time. s e r (c) Directional damping: intermediate destabilizing.

Figure 10.- Time histories of typical controls-fixed lateral oscillations with directional stability a n d y a w due to roll rate normal; variable- stability F-86A, M = 0.80, hD = 35,000 feet.

0.. ... 0 .

0 . 0 .

0 . 0 0 0.0 0 . . 0 D O - NACA RM ~ 5 6 ~ 0 8 .... .

4 0

0 . 0 . 0 . .

0 . 0.. 0 0 0' 0 . 0 . 0.. 0 .

# -

I -

Wing

span, f t . . . . . . . . . . . . . . . . . . . . . . . . . . . 22.69

Area, s q f t . . . . . . . . . . . . . . . . . . . . . . . . . 166.3

Aspect r a t i o . . . . . . . . . . . . . . . . . . . . . . . . . 3.09

Taper r a t i o . . . . . . . . . . . . . . . . . . . . . . . . . 0.39

Sweep, 0.25c, deg . . . . . . . . . . . . . . . . . . . . . . 15.9

Dihedral, deg . . . . . . . . . . . . . . . . . . . . . . . . 0

Over-all l e n g t h , f t . . . . . . . . . . . . . . . . . . . . . . 66.75

6K Airplane A ~ 6 ~ - 3 Symbol hP M P P Symbol 0.30 3.4 3 . 7

0 3,000 0.50 3 -0 3 - 2 0

0.70 2- 3 2.2 3 3 o.Go 3.4 3 ' 3 0 35,000 0.90 3.2 3.2 1.00 2.6 2.1 J I I

0 .2

.8 I .o

Figure 12.- Predicted lateral oscillatory characteristics of airplane A and measured lateral oscillatory characteristics of the variable- stability ~ 6 ~ - 3 compared with the pilot-opinion boundaries of reference 2 .

0 . 0 . 0 0 . ......

.. t . ...

e ....... * : ': : : NACA RV ~ 5 6 ~ 0 8

42 .. 0 . .

.. e.. . 0 . .e. .*

Wing

span, f t . . . . . . . . . . . . . . . . . . . . . . . . . . . 22.08

Area, s q f t . . . . . . . . . . . . . . . . . . . . . . . . . 191

Aspect r a t i o . . . . . . . . . . . . . . . . . . . . . . . . . 2.5

Taper r a t i o . . . . . . . . . . . . . . . . . . . . . . . . . 0.38

18 Sweep, 0.25c, deg . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . 0

Dihedral, deg

Over-all length, f t . . . . . . . . . . . . . . . . . . . . . . 48;25

Figure 13.- Two-view drawing and p r i n c i p a l dimensions of airplane B.

n- A i r p l a n e B F6F-3 hP M P Condition Symbol P Symbol

Design Cnr 0 2 . 2 0

3 x Design Cnr 0 2.2 0

0.90 2 . 0 35,000 6 x Design C n ,

- - - o - - - 0 2 . 2

2 x Design C n , A 2 . 3 A

--- - 4-

Sea level 0.24 2 . 5 Landing Config. n

-

I I

0 .2 .4 . 6 -8 I .o 1.2 I .4

l e 1 deg

I ve I ' f t/sec

Figure 14.- Predicted lateral oscillatory characteristics of airplane B and measured lateral oscillatory characteristics of the variable- stability F6F-3 compared with t h e pilot-opinion boundaries of reference 2.

Left

I I 1 1 I I

I I I I 1 I I Right Large adverse G CT Q) TI Left Right CT 0 a , -0 6 4

I I I I I I I

0 I 2 3 4 5

T i m e , sec Figure 15.- Effect of C i on roll response to abrupt rudder deflection Er for for variable-stability ~ 6 ~ - 3 airplane; one-half design

c %

airplane B.

0 0 0 0 0 0 0 - 0 0 0 0.0 0 0 0 . 0 0 0 0 0 0 NACA RM ~ 5 6 ~ 0 8

0 0 . 0 0

0 0 0 - 0 0 0 0 0 0 0 ** 0 0 0 0 0 0 0 0 Wing

span, ft . . . . . . . . . . . . . . . . . . . . . . . . . . . 35.67

Area, sq f t . . . . . . . . . . . . . . . . . . . . . . . . . 375.0

A s p e c t r a t i o . . . . . . . . . . . . . . . . . . . . . . . . . 3.39

Taper r a t i o . . . . . . . . . . . . . . . . . . . . . . . . . 0.25

Sweep, 0 . 2 5 ~ ~ deg . . . . . . . . . . . . . . . . . . . . . . 42.0

Dihedral, deg . . . . . . . . . . . . . . - . . . . . . . . . -5.0

Over-all length, f t . . . . . . . . . . . . . . . . . . . . . . 54.0

Figure 16.- Two-view drawing and principal dimensions of airplane C.

4 6

A i r p l a n e C f F6F - 3

VIVSL Symbol Damper P P Symbol hP 1.5 2.9 3.5

0 None 0

1.2 3 - 5 0 Yaw 3.7

3.5 0

-

1.2 0 Roll 4.2 3.4 I I Figure 17.- Predicted lateral oscillatory characteristics of airplane C and measured lateral oscillatory characteristics of the variable- stability F6F-3 compared with the pi.lot-opinion boundaries of reference 2.

n L .- r m t

E

u) Y C

.-

m

.

a

t9

I ' I I L a . a . * a .

NACA RM ~ 5 6 ~ 0 8 * 48 a . . a * a. a .

a . a . . * a * a .

a. *.a a a a * . m e a m a

Airplane G, y m p u t e d

---

F - 8 6 8 , fligh test Left -1.2 -.8 -.4

0 .4 .e I .2 I .6 2 .o

T i m e , sec (a) Basic condition (no aileron-rudder interconnection).

Figure 19,- Rolling-velocity and lateral-acceleration responses to abrupt for airplane C, compared wlth pedals-fixed aileron deflections computed measured responses of the variable-stabilj.ty F-86A.

Airplane C, computed F-86A, flight test Left -1.2 -.0

- .4

- . 3 -.2 -.I .I 0 .4 .8 1.2 I .6 2 . 0 Time, sec (b) Improved condition (with aileron-rudder interconnection).

Figure 19.- Concluded.

* .

...... ......

0 . 0 .

0 . 0 . .

: 0 : : : NACA RM A56C08

0 . 0 . . . . .

......

Wing

~ p a n , f t . . . . . . . . . . . . . . . . . . . . . . . . . . . 31-63

Area, sq f t . . . . . . . . . . . . . . . . . . . . . . . . . 250.0

Aspe c t r a t i o . . . . . . . . . . . . . . . . . . . . . . . . . 4.01

Taper r a t i o . . . . . . . . . . . . . . . . . . . . . . . . . 0.50

Sweep, 0.25c, deg . . . . . . . . . . . . . . . . . . . . . . 35 -0

Dihedral, deg . . . . . . . . . . . . . . . . . . . . . . . . -2.5

Over-all length, ft . . . . . . . . . . . . . . . . . . . . . . 40.83

Figure 20.- Two-view drawing and p r i n c i p a l dimensions of ai.rplane 1).

Airplane D I F6F-3 I -

- L

M v/vsL Symbol Average P Average P Symbol

Condition hP

10,000 0.64 --- U 1 . 8

2- 3

1.41 --- 13 1.5

Combat

40,000 ------

..-, 1 : n *

L A U I Y C -

0.70 --- 0 3.1

2.8 ii

55,000 0.98 --- 0 2.7

2 . 8 A

Power --- 1.1 A 3 . 8

approach 0 --- 1.4 0 3.4

3 . 0 0

P/A, reduced Cn

--- 1.1 0 4.0 3.0 n

B

-

~ lntolera ble

- t i - -

0 .2 .4 .6 .8 I .o 1.2

- 101 deg

1 ~ ~ 1 ’ ft/sec

Figure 21.- Predicted lateral oscillatory characteristics of airplane D and measured lateral oscillatory characteristics of the variable- stability F6F-3 compared with pilot-opinion boundaries of reference 2.

0 . e.. .

NACA RM ~ 5 6 ~ 0 8 0 . 0 . .

0 . ... .

Symbol M Symbol Q P P hP 0.91 0 1.1 1 .0 10 f 000 0 0.9 1.5 0 1.15 40 000 1.41 0 1 . 5 1.6 Flagged symbol indicates yaw damper ( Gn,/Gn = 11.74) ‘design

F - 8 6 A yaw - damper variation

Satisfactory I

Intolerable I

0 .2 .4 . 6 .8 I .o

Figure 22.- Predicted lateral oscillatory characteristics of airplane D and measured lateral oscillatory characteristics of the variable- stability F-86A compared with pilot-opinion boundaries of reference 2; combat cruise condition.

dm a m m m m m m m m m m a a m

NACA RM ~ 5 6 ~ 0 8 : : m m m m a a m m m m m a 6 0 0 m a m m m a - a m m a a m m

i

I - Wing

span, ft . . . . . . . . . . . . . . . . . . . . . . . . . . . 56.86

Area, sq f t . . . . . . . . . . . . . . . . . . . . . . . . . 1542.0

Aspect r a t i o . . . . . . . . . . . . . . . . . . . . . . . . . 2.096

T a p e r r a t i o . . . . . . . . . . . . . . . . . . . . . . . . . 0

Sweep, O.25c, deg . . . . . . . . . . . . . . . . . . . . . . 51.5

Dihedral, deg . . . . . . . . . . . . . . . . . . . . . . . . 0

Over-all length, ft . . . . . . . . . . . . . . . . . . . . . . 95.00

Figure 23.- Two-view drawing and p r i n c i p a l dimensions of a i r p l a n e E as simulated by v a r i a b l e - s t a b i l i t y F6F-3.

a - I

LO

.2 I .o

Figure 24.- Predicted lateral oscillatory characteristics of airplane E and measured lateral oscillatory characteristics of the variable- stability F6F-3.

NACA RM ~ 5 6 ~ 0 8 81 I 1 I 1 I I i Right Left .2 Right I, radians/sec Left .2 .6

Right I Controls held fixed I 1

.4 .2 P, radians/sec .2 Left .4 -0 4 8 12 1 6 20 24 28 32 36 Time, sec Figure 25.- Time histories of lateral a n d directional motions of the variable-stability ~ 6 ~ - 3 with low directional stability; Cn z 0.

B .........

0 . 0 . .

. . 0 . . 0 .

0 . * . .

56 NACA RM ~ 5 6 ~ 0 8 .* . e * e .

0 0 Wing

s p a n , f t . . . . . . . . . . . . . . . . . . . . . . . . . . . 56.86

Area, sq f t . . . . . . . . . . . . . . . . . . . . . . . . . 1342.0

A s D e c t r a t i o . . . . . . . . . . . . . . . . . . . . . . . . . 2.096

Taper r a t i o . . . . . . . . . . . . . . . . . . . . . . . . .

Sweep, 0.25c, deg . . . . . . . . . . . . . . . . . . . . . . 51.5

Dihedral, deg . . . . . . . . . . . . . . . . . . . . . . . . 2.23

Over-all l e n g t h , f t . . . . . . . . . . . . . . . . . . . . . . 96.78

Figure 26.- Two-view drawing and p r i n c i p a l dimensions of a i r p l a n e E as simulated by v a r i a b l e - s t a b i l i t y F-86A.

mm oms m m m m m m o m m m 0 0 0 NACA RM ~ 5 6 ~ 0 8 m m m m m m o 0 0 m m m m o m m m m m Wing S p m , f t . . . . . . . . . . . . . . . . . . . . . . . . . . .

33.0

Area, sq f t . . . . . . . . . . . . . . . . . . . . . . . . . 181.8

Aspect r a t i o . . . . . . . . . . . . . . . . . . . . . . . . . 6 .o

Taper r a t i o . . . . . . . . . . . . . . . . . . . . . . . . . 0.64

Sweep, 0.25~2, deg . . . . . . . . . . . . . . . . . . . . . . 0

Dihedral, deg . . . . . . . . . . . . . . . . . . . . . . . . 2

f t . . . . . . . . . . . . . . . . . . . . . . 27.1

Over-all length, Figure 27.- Two-view drawing and p r i n c i p a l dimensions of a i r p l a n e F.

e a e a . e s e e - e e e e e e e a e*: e o e e e e e e o e e NACA RM ~ 5 6 ~ 0 8 58 . e e e e e e e e e e.. e e e e e e e e e e a e e e e e

Airplane F F ~ F - 3

v, mph P P Symbol

Symbol Condition hP 0 Climb 0 220 2.3

2.1 0

0 Landing 0 109 4.8 3 -0 0 Satisfactory

0 .2 .4 . 6 .8 I .o

Figure 28.- Predicted lateral oscillatory characteristics of airplane F and measured lateral oscillatory characteristics of the variable- stability ~ 6 ~ - 3 compared with the pilot-opinion boundaries of reference 2.

a : Normal flight Boundaries of reference I b : Dampers inoperative

--

Boundaries of reference 2 A rea Pi tots' comments I Generally satisfactory 2 Some objectionable characteristics 3 Artificial stabilization required in landing approach I

0 .2 .4 .6 .8 I .o I .2 I . 4

1 $ 1 deg

Fl' ft/sec

Figure 29.- Comparison of pilots' comments on lateral oscillatory characteristics with specification of reference 1; airplanes A, B, C, D, and F.

NACA - Langley Field, Va.