Determination of flight characteristics of supersonic transports during the landing approach with a large jet transport in-flight simulator

DETERMINATION OF FLIGHT CHARACTERISTICS

OF SUPERSONIC TRANSPORTS DURING

THE LANDING APPROACH WITH A LARGE

r ,

JET TRANSPORT IN-FLIGHT SIMULATOR

, \ y

\ /c, _I I \ ‘ I - c %A,

by Stuff of the Lungley Reseurcb Center

+ 4 L,) ,’* ’ r

Lungley Reseurcb Center

LuagZey Stution, Humpton, Vu. a

N A T I O N A L AERONAUTICS A N D SPACE A D M I N I S T R A T I O N WASHINGTON, D. C. 0 J U N E 1967 TECH LIBRARY KAFB, NM NASA TN D-3971 DETERMINATION OF FLIGHT CHARACTERISTICS OF SUPERSONIC TRANSPORTS DURING THE LANDING APPROACH WITH A LARGE JET TRANSPORT IN- FLIGHT SIMULATOR By Staff of the Langley Research Center Langley Research Center Langley Station, Hampton, Va.

NATIONAL AERONAUTICS AND SPACE ADMINISTRATION ..

For sale by the Clearinghouse for Federal Scientific and Technical Information

Springfield, Virginia 22151 - CFSTI price $3.00

PREFACE This compilation contains results of in-flight simulator tests made to determine the low-speed flight characteristics of several generalized supersonic transport con- figurations. A large jet transport was used as an in-flight dynamic simulator. This of the NASA Langley Research Center investigation was made by members of the staff and is reported in six parts, each covering one aspect of the study. These parts contain discussions of procedures, equipment, performance characteristic s, longitudinal handling qualities, lateral-directional handling qualities, and an evaluation of the pilot workload.

iii I I I E - I 111111 I CONTENTS P R E F A C E . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . iii S Y M B O L S . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . vii 1. INTRODUCTION

By Robert 0. Schade . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1

2. PROCEDURES AND EQUIPMENT

By Harold L. C r a n e . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5

3. PERFORMANCE CHARACTERISTICS

By Albert W. Hall . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37

4. LONGITUDINAL HANDLING QUALITIES

By William D. Grantham and Lee H. P e r s o n . . . . . . . . . . . . . . . . . 51

5. LATERAL-DIRECTIONAL HANDLING QUALITIES

By Robert E. Shanks, Samuel A. Morello, and Jere B. Cobb . . . . . . . . . 101

6. AN EVALUATION OF PILOT WORKLOAD

By Samuel A. Morello and Albert W. Hall . . . . . . . . . . . . . . . . . . . 167

V SYMBOLS The units f o r the physical quantities used herein are presented in both the U.S.

Customary System of Units and the International System of Units. Factors relating these two systems of units may be found in NASA SP-7012.l The moments of inertia are with respect to the body axes. The stability derivatives However, the simulation was set up so that are given with respect to the stability axes.

all these parameters were transferred to the stability axes.

span, feet (meters) mean aerodynamic chord, feet (meters) cycles to damp to one-half amplitude drag, pounds (newtons) force input to control column, pounds (newtons) natural frequency, cycles/second (hertz) frequency dependent parameter acceleration due to gravity, feet/seconda (meters/second2) geometric altitude, feet (meters) pressure altitude, feet (meters) wing incidence, degrees moments of inertia about X-, Y-, and Z-axes, respectively, slug-feet2 (kilogram -meter s2) product of inertia, slug-feet2 (kilogram-meter$) constant lift, pounds (newtons)

1E. A. Mechtly: The International System of Units - Physical Constants and Con-

version Factors. NASA SP-7012, 1964.

vii l i f t per unit angle of attack per unit of momentum, CLaqS/mV, per second distance from angle-of-attack and sideslip vanes to center of gravity, feet (meters) m mass of airplane, slugs (meters) aileron coupling parameter, positive for adverse yaw "Ba n load factor, g units P period, seconds

dynamic pressure, -pV 1 2 , pounds/foo@ (newtons/metera)

dynamic pressure at trim conditions, pounds/foot2 (newtons/metera) q0 S wing area, feet2 (meters2) S Laplace transform operator T thrust, pounds (newtons) roll time constant, seconds TR time to double amplitude, seconds T2 time to damp to one-half amplitude, seconds T1/2 t time, seconds V true airspeed, knots equivalent airspeed, knots Ve trim airspeed, knots V O equivalent side velocity, feet/second (meters/second) ve viii W weight, pounds (newtons) C Y angle of attack, degrees sideslip angle, degrees B flight-path angle, degrees aileron deflection, positive with right aileron down, degrees deflection command, degrees control column deflection, degrees elevator deflection, positive with trailing edge down, degrees thrust modulator deflection, degrees rudder deflection, positive with trailing edge left, degrees spoiler deflection, degrees thrust- modulator deflection, degrees wheel deflection, positive with wheel right, degrees damping ratio pitch attitude, degrees sweepback angle, degrees air density, slugs/foot3 (kilograms/meter3) bank angle, degrees heading angle, degrees damped natural frequency of short-period longitudinal mode, radians/second WD i x

P

undamped natural frequency of Dutch roll oscillation, radians/second ("d ' undamped natural frequency of short-period longitudinal mode, On radians/second undamped natural frequency from lateral numerator quadratic, radians/second CD drag coefficient CL lift coefficient

c z rolling-moment coefficient

Cm pitching-moment coefficient Cn yawing-moment coefficient damping-in- roll parameter czP CY side-force coefficient Subscripts: basic basic configuration max maximum trim trim conditions wh wheel SST supersonic transport

- 80 367-80 airplane configuration

X Abbreviations: c.g. center of gravity IFR instrument flight rules ILS instrument landing system P I 0 pilot-induced oscillation PR pilot rating r m s root mean square SAS stability augmentation system SST supersonic transport VFR visual flight rules The method of indicating partial derivatives is as follows: A dot over a symbol represents a derivative with respect to time.

x i 1. INTRODUCTION By Robert 0. Schade The presently proposed configurations of the supersonic transport (SST) are different from any existing commercial airplane. These airplanes, primarily designed for supersonic cruise performance, introduce geometric and design features which are expected to-affect the low- speed flight characteristics adversely and to cause problems during instrument-flight-rules (IFR) approaches. For example, the following table, which shows a comparison of the characteristics of two supersonic transport configurations (see fig. 1-1) and a typical large subsonic jet transport, indicates that the pitch inertia is approximately 3.5 times that of current subsonic jet transports. This increased pitch inertia may have detrimental effects on pitch- response times, and, consequently, glide- path control, sink-speed control, and touchdown accuracy. The large increases in the yaw-to-roll moments of inertia (3 to 4 times greater than those of the subsonic jet trans- ports) will possibly introduce new or unusual lateral-directional cross-coupling charac- teristics. The low frequencies of the longitudinal short-period and Dutch roll motion resulting from the high inertias may result in undesirable transient response character- istics. In addition, delta-type configurations will be making landing approaches with speed-thrust instability or on the "back side" of the thrust-required curve (at currently proposed approach speeds) in a region where a decrease in airspeed requires an increase areas, need to be further explored and in thrust. These, and other potential problem design parameters changed as required to provide acceptable low- speed handling qualities and to insure adequate flight safety for future SST configurations.

The current military and civil handling- quantities requirements a r e a useful guide but, in some cases, they have already been proven obsolete by experience with present subsonic jet transports.

The requirements for the Dutch roll and longitudinal stability appear to be too restrictive and others, such 3x4-geometry SST Present jet transport 'resent jet transport htio - Landing weight . . . . . . . . . . . . . 1 . 8 1 . 8 Moment of inertia . . . . . . . . . . . . . . . . . 3.6 3.5 Pitch.

Fixed geometry Variable geometry Roll . . . . . . . . . . . . . . . . . . 0.6 0 . 8 Yaw. . . . . . . . . . . . . . . . . . . 2.4 2.4 Figure 1-1.- SST configurations.

Damped period Longitudinal short-period motion. . . 2.3 1.3 1.5 Dutch roll motion . . . . . . . . . . . 1.2 as lateral-control response, appear to be not restrictive enough. It therefore appears that further flight experience is needed on the SST configurations to shed additional light on the possible updating of the handling-qualities requirements and criteria for this type of airplane.

Ground-simulation techniques provide answers for handling-qualities problems of cruise and instrument flight; however, they are not as satisfactory for evaluating landing characteristics as an in-flight simulator since, during the final landing phase, the pilot relies on a combination of airplane and outside visual references and is subjected to situations which can only be fully experienced i n flight. It appeared, therefore, that the best presently available method for evaluating the SST approach and landing characteris- tics would be a large in-flight dynamic simulator which both simulates the airplane being tested and places the pilot i n the most realistic flight environment possible.

A s a result, a contract was negotiated with The Boeing Company to modify a large as a low-speed in-flight simulator.

four-engine transport airplane The modified in-flight simulator was flown in a simulated IFR low-speed approach and landing investigation at the NASA Langley Research Center from May to October, 1965; variations in generalized configurations of the fixed-geometry and variable-geometry SST concepts were incorpo- rated in the flights.

The main objectives of this investigation were to: (1) Study the handling qualities of the basic SST configurations and evaluate potential handling-qualities problem areas (2) Obtain preliminary indications of stability-augmentation requirements for satis- factory handling qualities (3) Obtain some indications of the tolerable o r minimum acceptable handling quali- ties by parameter variation of: (a) Aerodynamic characteristics (b) Center-of-gravity location (variable-geometry configuration only) (4) Determine effects of speed-thrust instability or operation on the "backside" of the power- required curve (fixed-geometry configuration only) (5) Obtain approach and landing data applicable to criteria and certification require- ments for SST airplanes The basic SST configurations and variations that were tested in this investigation are as follows: Variable-Geometry Configurations (1) Basic airplane

F -

(2) Pitch-rate augmentation (3) Pitch-rate and angle-of-attack augmentation (4) Aft center of gravity (5) Aft center of gravity with pitch-rate and angle-of-attack augmentation (6) Dutch roll augmentation (7) Dutch roll and adverse-yaw degradation Basic Variable- - Geometry Emergency- Landing (Cruise-Sweep) - Configuration Fixed- Geometry . . . Configurations (1) Basic airplane (2) Pitch-rate and angle-of-attack augmentation (3) Improved speed-thrust stabiliiy (4) Roll-damping augmentation (5) Dutch roll and adverse-yaw degradation For the configurations, the pilot-evaluation tasks were simulated IFR o r hooded landing approaches along prescribed flight paths. The pilots' comments along with various measured flight data were used to evaluate each of the conditions flown; the are included in the following parts of this paper.

results of these evaluations 2. PROCEDURES AND EQUIPMENT By Harold L. Crane SUMMARY An in-flight simulation has been made to determine the handling qualities of several supersonic transport configurations during the landing approach.. This part of the compi- lation describes the test program, the SST test configurations, the simulator, and the SST test configurations. The discussion of the simulator stability augmentation for the covers several topics including the test airplane, the simulation technique, the simula- tion equations, the simulator specifications, the simulation procedures, and examples of the quality of simulation.

INTRODUCTION This part of the report includes a discussion of the procedures and equipment used in the investigation. The research program and test conditions a r e discussed and the test airplane and simulation system are described. The test airplane was the Boeing 367-80, a jet transport prototype. The selection of a five-degree-of -freedom simulation using the response feedback technique is discussed. An example block dia- gram and the complete simulation equations are presented. Details of the control system and response specifications for the five simulation input systems are presented. The selection of stability-augmentation techniques for the SST test configurations is discussed.

Simulation test procedures, quality of simulation obtained, and operational experience with this simulator are also discussed.

PROGRAM AND TEST CONDITIONS The object of this program w a s to investigate the landing-approach and touchdown characteristics of SST configurations by means of in-flight simulation. The configura- tions were designed to represent the fixed-geometry and variable-geometry concepts of the supersonic transport. The variable-geometry configuration w a s tested mainly at the minimum sweep angle of 20° with a brief investigation of the fully swept 72O emergency- landing (or cruise) configuration. The dimensions and design aerodynamic parameters f o r the simulated configurations are given in tables 2-1 and 2-2 on pages 33 and 3 4 .

The test program included: (1) pilot familiarization and VFR (visual-flight-rules) evaluation of the three SST landing-approach configurations at an altitude of 4000 to 8000 feet (1220 to 2440 meters) and (2) the evaluation of instrument-landing-approach characteristics and visual-flare and touchdown characteristics of the three SST con- figurations. The approach speed was 135 knots except for the 72O emergency-landing configuration for which an approach speed of 150 knots was used to simulate 182 knots.

The VFR evaluations consisted of seven basic tests which were as follows: 1. Evaluate static longitudinal and speed-thrust stability and longitudinal control capability by varying speed rtl0 knots with the elevator only.

2 . Evaluate the steady-maneuver characteristics by a wind-up turn to a 4 5 O bank angle.

3. Evaluate the transient-maneuver characteristics by performing a 1O0-pitch- attitude change as rapidly and accurately as possible by using the flight director.

4. Evaluate trim characteristics by cutting power and then reestablishing trim speed.

5. Evaluate Dutch roll characteristics by releasing the airplane from a loo side- slip angle.

6. Evaluate roll-control-response characteristics by a 10°-wheel input with the rudder fixed.

7. Evalute the ease of making a precise heading change by performing heading changes of loo and 3 0 ' .

Following these tests, final pilot evaluation was obtained under conditions of simu- lated (hooded) instrument flight rules (IFR) during approaches to landing. Hooded approaches were used in order to provide a precision pilot task that was representative of actual flight operations.

For this task, an intercept of the localizer w a s made with landing gear down approximately 8 miles ( 1 2 . 8 kilometers) from the runway at an altitude of 1500 feet (460 meters). The flaps and airspeed were then adjusted for the landing approach as required by the simulation. At the intercept of the glide slope, approximately 5 miles (8 kilometers) from the runway, a descent was initiated and the pilot attempted to fly the prescribed flight path as closely as possible down to approximately 200 feet (61 meters) and, if conditions were favorable, continue visually to touchdown. Some tests were made with the localizer offset 200 feet (61 meters) from the runway center line during the approach to evaluate the lateral maneuverability. Following the simu- lated IFR breakout at 200 feet (61 meters) with the lateral offsets, the pilot performed a visual sidestep maneuver in order to line up with the runway. Other tests were also made with square-wave vertical offsets of the glide slope approximately halfway down the glide slope to study the speed-thrust stability and longitudinal maneuverability while the pilot w a s under the hood.

All flight tests w e r e conducted during good ceiling and visibility conditions with light-to-moderate winds of 15 knots or less and gusts below 5 knots.

The following variations and changes were included in the basic SST aircraft con- figurations being simulated during the flight- test program: 1. Variable-geometry variations a. A longitudinal stability- augmentation system was developed which con- sisted of adding pitch-rate feedback and increasing the gearing between the elevator and the column. A final system similar to the preceding one also included angle-of-attack feedback.

b. An aft center-of-gravity configuration was used to simulate the airplane flying with the center-of-gravity location near the maximum allowable aft position. (The final augmentation system described in (a) w a s also used during a portion of these tests.)

c. A lateral-directional stability-augmentation system w a s used to improve the Dutch roll damping.

d. Degraded lateral-directional characteristics were obtained by increasing adverse yaw and reducing Dutch roll damping.

2. Variable-geometry cruise configuration (only the basic configuration was flown) 3. Fixed-geometry variations a. A longitudinal stability-augmentation system identical to the one on the variable-geometry configuration w a s used.

b. A lateral-directional stability-augmentation system was used which improved the roll damping.

c. Degraded lateral-directional characteristics w e r e obtained by increasing adverse yaw and reducing Dutch roll damping.

d. Improved speed-thrust stability w a s obtained by making the thrust versus velocity characteristics of the simulated SST stable.

Most of the evaluation flights w e r e flown by two NASA Langley Research Center pilots. However, brief evaluations of the two basic test configurations were made by two pilots from industry and one from the FAA. An NASA Ames Research Center pilot and a third Langley pilot also briefly evaluated both the basic and the augmented SST configura- tions. Besides the descriptive comments from each pilot, Cooper pilot ratings (ref. 1) were obtained for each configuration. (The Cooper rating system is shown in table 2-3.

The Cooper ratings presented a r e usually the average of two o r more pilot ratings.)

A standard questionnaire w a s used during all postflight debriefings to assure that the all pilots and all test configurations would cover the same topics.

comments obtained for TEST AIRPLANE The design of the simulation system was, of course, strongly influenced by the test airplane. The Boeing 367-80 is a prototype airplane which is similar to a Boeing 707, but has a somewhat shorter fuselage. The pilots are located about 55 feet (17 meters) ahead of the center of gravity, or about half as far ahead of the center of gravity as in proposed SST designs. The airplane configuration is shown in figures 2-1 and 2-2, and the mass and inertia characteristics are given in table 2-1. Aerodynamic parameters for the 367-80 airplane with the spoiler and flap deflections required for SST simulation a r e presented in table 2-4. For this project, the airplane was equipped with quick- acting, precise, irreversible servo-operated control systems. The servo specifications are presented in a subsequent section of this part of the compilation.

Figure 2-1.- Test airplane as equipped for SST landing-approach simulation.

L-65-5431

w

46.0 ft(14.02 m) L -129.58 ft (39.50 m ) - ' 5 C >"tal t b 4 4 . 0 ft- (13.41 m) Figure 2-2.- Three-view drawing of test airplane.

..- SIMULATION SYSTEM Simulation Technique The response feedback technique was selected for this simulation project. This choice was influenced by the fact that the simulator w a s intended for short-term use and by the desire to complete the SST landing-approach tests as soon as possible. With the response feedback technique, an analog computer is programed to modify &e test air- plane stability derivatives to represent derivatives and mass and inertia Characteristics the simulated configuration. The proper response for the configuration being simu- of lated is thereby obtained. Although feedback loops are used to modulate control deflec- tions, the response feedback simulation technique uses an open-loop computation. Some cut-and-try manual adjustment of gains is usually required. The response feedback technique should not be confused with the closed-loop model-analog simulation technique with which the airplane response is continuously and automatically matched to that of an analog-computer model.

To apply the response feedback technique of simulation, it is necessary to know all the mass and aerodynamic parameters (stability derivatives) of the test airplane. Flight tests were therefore required to measure the 367-80 airplane characteristics in the simu- lation test configurations (such as, at a speed of 135 knots with 30° flap deflection, landing gear down, for spoiler deflections up to loo).

The simulation was designed to match five degrees of freedom of the SST. The force and moment characteristics which were varied for the simulation included lift, drag, pitching moment, rolling moment, and yawing moment. Lift was varied by modulating the spoilers o r air brakes with respect to a 6O initial deflection. Nonlinear spoiler effectiveness was compensated for by driving the spoilers through nonlinear function generators. Thrust and drag were varied by modulating the clamshell doors of the initial deflection of 3 0 ° . The thrust standard Boeing 707 thrust reversers from an response of the simulator, which is indicated in the specifications, was probably slightly faster than it will be for the SST. Moments were produced by supplementary deflections of the elevator, rudder, and the lateral-control system.

Side force was not modified from the basic 367-80 airplane characteristics. Simu- lation of side force would require expensive modification of the test airplane, such as the addition of an all-movable vertical surface. However, a comparison of transient response, including sideslip and lateral acceleration as well as angular velocities, from five- and six-degree-of -freedom analog-computer tests showed that a five- degree-of -freedom simulation using unmodified 367-80 side-force characteristics would be adequate in this case.

I.;' - Thrust settings w e r e adjusted for the effects of altitude and temperature. How- ever, because the test program was to be made at altitudes from sea level to about 5000 feet (1520 meters), no other corrections for altitude effects w e r e considered to be necessary.

The simulation system was designed to permit flare and touchdown in the simula- tion mode. Nonlinear function generators were used to modify the estimated 367-80 ground effects to simulate the predicted SST ground effects. Parameters adjusted for Altitude w a s the effects of ground proximity were lift, drag, and pitching moment.

obtained for this purpose during flare and touchdown from a radar altimeter located near the center of gravity. The ground-effect functions used for the fixed- and variable- These data are based on unpub- geometry configurations are presented in figure 2-3.

lished wind-tunnel data from several sources. (No ground effects w e r e simulated for the variable-geometry emergency-landing configuration.) It w a s beyond the capability of the spoiler system (from a 6O trim setting) to simulate the full ground effects on the l i f t of the fixed-geometry configuration. Therefore, as shown in figure 2-3, only 35 per- cent of the estimated incremental lift could be simulated. In order to maintain the CL,trim Simulated variable geometry 0.89 Simulated fixed geometry 0.54 0 . 5 4 - - - - - Eetimated fixed geometry Height of wing abwe ground, h/E (a) ACL as a function of h/E.

Figure 2-3.- Incremental lift, drag, and pitching moments due to ground effect for the SST test configurations.

map increase U % I - - --.L- ~L I . .- J 1.0 1 . 2 1.4 0 . 2 .4 . 6 .a Height of wing above ground, h/S (b) ACD as a function of h/E.

Figure 2-3.- Continued.

proper lift-drag ratio, the simulation of incremental drag was also restricted to 35 per- cent of the estimated value.

It was considered important to make actual touchdowns in the simulation mode, even though in that case neither SST attitude nor pilot height above the ground for approach and touchdown could be simulated. Away from the ground, this flying simulator could be con- figured to fly within about 4 O of the attitude of even the more highly swept test configura- tions. However, during flare and touchdown, the match could not be this close, and, in the interests of safe operation, no increase in normal 367-80 touchdown attitude was used in the simulation mode. The approach-body attitude used for this test program w a s 0.5' compared with estimated approach attitudes of 3.6' for the variable-geometry SST con- figuration and go for the fixed-geometry SST configuration.

- simulated variable geometry

S i m h t e d fixed geometry Pitch -.oj - - -.w - -.01 0 . 2 .4 .6 . 8 1.0 1 . 2 1.4 K e i g h t of wing above ground, h/E (c) ACm as a function of h/L Figure 2-3.- Concluded.

Equations for Simulation The moments of inertia of table 2-1 are with respect to body axes. The stability derivatives given in tables 2-2 and 2-4 a r e with respect to stability axes. The angular- velocity sensors measured angular velocities with respect to body axes. However, the simulation system was set up so that all these parameters were transferred to the sta- bility axes. The equations of motion for the airplane were arranged as follows for this simulation project:

Lift -

Drag -

m/2qs m/2qE

Pitching moment -

..

ha+ e =

' IIIIIII I Ill 1lll1l1l11 I l l I IN I I 1 1111 I I1 1111111.111.1 111111 I I.

Rolling moment -

Yawing moment -

Inputs to the control surfaces, spoilers, and thrust modulators in the control-fixed simulation mode were determined by the following equations:

Elevator -

Spoilers (symmetric mode) -

Thrust modulators -

- a 6m

a6T A 9 + 2 AV,,

6m - - ( . aT ha! + - ae av

Wheel (lateral-control system) -

Rudder -

a 6 ,

6 , = - p + - $ + - T $

aB 84 w

The gains for these simulation inputs, based on linearized theory for small pertur- bations, were calculated as follows: Similar expressions can be written for the other elevator, wheel, and rudder gains.

The spoiler gains were: For cases which can be flown at the actual speed of the simulation, the denominator Vo values cancel out and m/S can be substituted for A.

The thrust-modulator gains were proportional to The following expressions were used to compensate for interaction or cross- control effects: - - I 1 1 I 1111111111111111 1 1 1 1 1 1 1 1 1 1 1 1 1 - - 1 ,111 B Control-surface authority was simulated as follows:

a6w,-80 - -

a6w, SST Since there were large differences between the attitude of the test airplane and the estimated approach attitudes of the SST configurations, it became necessary to adjust the inertia values and stability derivatives to account for product of inertia differences. An inertia cross-product transformation w a s used. By using moments of inertia about the stability axes, the rolling and yawing moments of inertia were replaced as follows:

IX by Ix-- Ixz2

I Z and

Ixz2

IZ by I z - -

I X The aerodynamic stability and control coefficients were transformed as follows:

Ixz

CzP + + C r y IXZ

c z 6

cn6 - cn* @ +- I x

1 - 1 1 I 1 1 1 I I 1111111111111111 II 111 I II I 111 I II II I which were measured about the body axes were The rolling and yawing velocities transferred to stability axes to be used in simulation equations.

Simulation Equipment To implement the simulation, the right-hand column and wheel w e r e mechanically disconnected from the normal control system and connected into an electrical system which operated the control surfaces through an interface and analog computer. The interface console, shown in figure 2-4, receives electric signals from the airplane con- trol system and the airplane response transducers and modifies these signals to make them compatible with the analog computer. For example, the interface demodulates ac signals from airplane instruments to dc for use in the computer. Switching circuits to engage or automatically disengage the simulation a r e also contained in the interface.

The simulation computer is shown in figure 2-5. The computer was slightly modi- fied for flight use.

Commands for control-surface deflection went from the computer to the interface The servotab- to autopilot electric servovalves which operated the hydraulic servos.

operated elevator and aileron systems were replaced with the irreversible servosys- tems used on the Boeing 727 airplane. The spoiler servos were replaced with an improved system which provided spoiler positioning accuracy of about * 1 / 4 ' . The thrust-reverser-system actuator w a s located in the fuselage with cable runs to the four engines.

*The simulator was designed to properly represent SST response. However, when there was a large difference between 367-80 attitude and simulated SST attitude the 367-80 cockpit motions approximated those of the SST nose wheel. The result was an unrealistic adverse yawing of the 367-80 cockpit during rolling maneuvers. A partial

remedy for this situation was to reduce the gain of C . This adjustment was applied

n6W to the fixed-geometry configuration to provide realistic cockpit response without appre- ciably degrading the simulation. The static directional stability derivative 86 ap for

./

steady sideslips w a s made about 20 percent too high by this modification. Attempts to apply the same fix very quickly to the cruise-sweep configuration were not successful.

Figure 2-4.- Interface console in the test airplane. L- 65-6116 A simplified block diagram of the pitch-control system is shown in figure 2-6. The other four systems for roll, yaw, lift, and drag were similar to the pitch system. Block diagrams of all five control systems a r e given in reference 2.

Specifications of Simulation System The following specifications obtained from reference 2 are listed in terms of 367-80 airplane control deflections to show the approximate system response to pilot and analog- computer inputs.

Figure 2-5.- Analog computer installed i n test airplane. L-65-6772 Control system

Pilot - Elevator +

simulation airplane L actuator dynamics

-

Response feedback

-

Cockpit instruments Instrumentation Figure 2-6.- Block diagram of response feedback technique as applied to pitch-control system.

Rudder system:

Electrical-command limit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . *loo

Rudder-deflection limit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . i26'

Maximum no-load rudder rate . . . . . . . . . . . . . . . . . . . . . . . . . . 33O/sec

K Open-loop calculated transfer function

. . . . = KG(s) =

(0.063s + 1)(0.028s + 1)

Typical response to step command for 5 O -

Initial response . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 0.06 sec

63% response . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 0.17 sec

Final response . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 0.33 sec

Elevator:

Electrical-command limit at 135 knots . . . . . . . . . . . . . . . . . . . . . . -9'

Elevator-deflection limits . . . . . . . . . . . . . . . . . . . . . . . . . . . 15O, -25O

Maximum no-load surface rate -

Normal system (safety pilot) . . . . . . . . . . . . . . . . . . . . . . . . 50°/sec

Simulation mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25O/sec

6e K

. . . . . . . . . . . . . - = KG(s)

Open-loop calculated transfer function

0.06s + 1

6C

Frequency response (master) * is down . . . . . . . . . . . . . . . . . 3 dB at 2.8 cps

lag exceeds 9 0 ' at . . . . . . . . . . . . . . . . . . . . . . . . . . . >2.25 cps

Phase

-

Frequency response (slave-) is down. . . . . . . . . . . . . . . . . . >3 dB at 1.8 cps

Right hand Left hand

Typical response to step command for 5 O -

(master) (slave)

Initial response . . . . . . . . . . . . . . . . . . . . . . 0.03 sec

0.09 sec

63% response . . . . . . . . . . . . . . . . . . . . . . . 0.18 sec 0.22 sec

Final response . . . . . . . . . . . . . . . . . . . . . . . 0.3 sec 0.3 sec

Ailerons: Electrical-command limit (wheel deflection) . . . . . . . . . . . . . . . . . . .

63'

Aileron-deflection limit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . *25O

Maximum no-load aileron rate. . . . . . . . . . . . . . . . . . . . . . . . . . . 68'/sec

6a K - = Open-loop calculated transfer function . . . . . . . . . .

(0.014s + 1)(0.05s + 1)

6c

Frequency response is down. . . . . . . . . . . . . . . . . . . . . . 3 dB at 1.25 cps

Phase angle exceeds 90° at . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.25 cps

Typical response to step command for 5 ' -

Initial response. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 0.05 sec

*Either of the two elevator systems, left or right, may be selected as the master, then the other elevator system becomes a slave system which follows and closely approx- imates the response of the master.

63% response . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 . 1 2 sec

Final response . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 0.13 sec

Hysteresis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 . 2 0

Spoilers: Electrical-command limit (when used symmetrically) from

initial setting of 6 O . . . . . . . . . . . . . . . . . . . . . . . . . . . . . loo, -6'

Spoiler-deflection limits . . . . . . . . . . . . . . . . . . . . . . . . . . . . 0 ' to 48'

Maximum no-load rates -

Wheel rate (simulation mode) . . . . . . . . . . . . . . . . . . . . . . . . 180°/sec

Surface rate (simulation mode) . . . . . . . . . . . . . . . . . . . . . . . 50°/sec

Open-loop calculated transfer functions -

6s - K

For lateral control. - -

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

s2 + 0.7(23)s + (23)2

6~ 6s - K

For lift control. . . . . . . . . . . . . . . . . . . . . . . - _

(0.1s + 1)(0.03~ + 1)

6c

Frequency response is down . . . . . . . . . . . . . . . . . . . . . . . 3 dB at 1.6 cps

Phase lag exceeds 90° at. . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.5 cps

Typical response to step command for 2.7O -

Initial response . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 . 0 1 sec

63% response . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 0.09 sec

Final response . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 0.5 sec

Hysteresis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . <o. lo

Gearing (typical) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6s/6wh z 0.26

Thrust modulators (clamshell doors): +120 Electrical-command limits (from initial 30°) . . . . . . . . . . . . . . . . .

Normal-deflection limits . . . . . . . . . . . . . . . . . . . . . . . . . . . 0 ' to 55'

. . . . . . . . . . . . . . . . . . . . . . . . . . . 14 O/se c

Maximum - def 1ec ti on rate

6m - K

. . . . . . . . . . . Open-loop calculated transfer function

6c - (0.04s + 1)(0.19s + 1)

Typical response to loo step command -

Initial response . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 0.3 sec

63% response . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 0.7 sec

Final response . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 0.9 sec

The static gain of this system produced a simulated thrust increment of 3000 pounds (13 kilonewtons) per degree of SST throttle deflection. (AT/W z 0.01 per degree.)

Control-System Details The pitch and roll control systems of the simulator worked in parallel with the standard 367-80 control systems. Therefore the safety pilot's controls moved with the control surfaces and provided him with an indication of the control inputs. However, simulation inputs to the rudder were fed into the existing yaw-damper system and did not move the rudder pedals. To permit the safety pilot to monitor overall rudder inputs, a position indicator w a s installed in the cockpit.

The existing experimental thrust-modulation system on the test airplane was operated by four thrust levers located on the center console to the left of the throttles as shown in figure 2-7. For the simulation mode, the evaluation pilot w a s provided with a single electric throttle also located on the center console as shown in fig- ure 2-7. Deflection of the electric throttle drove the complete thrust- reverser system including the four manual thrust levers. The safety Figure 2-7.- Details of cockpit center console. L-65-6775 pilot could therefore observe all inputs to the thrust-modulation system which were made by either the evaluation pilot o r by the analog computer. The position of each set of clamshell doors was also shown by an indicator on the center instrument panel as shown in figure 2-8.

A set of saturation indicator lights and a disengage indicator light (shown in fig. 2-8) were provided to keep the evaluation pilot aware of the simulation status. The tests were designed to stay within the saturation limits.

The evaluation pilot was provided with control "feel" from preloaded centering springs. The pitch-control "feel system'' used a hydraulic spring which provided an The usual gradient was 4 pounds (18 newtons) per degree of column adjustable gradient.

deflection with a 4-pound (18-newton) breakout force. Pitch trim w a s provided for the Figure 2-8.- Evaluation pilot's instrument panel. L-65-6774 evaluation pilot through the normal trim button which actuated the elevator instead of the stabilizer. Pitch trim rates were 2.3O per second for the fixed-geometry configuration and 1.8O per second for the variable-geometry configuration. The wheel force gradient included a 21-pound (11-newton) breakout force and required 1% pounds (55.6 newtons) for full 7 5 ' wheel deflection. Rudimentary roll trim w a s provided by a potentiometer at the rear of the center console that biased the roll-control signal from the computer.

The rudder pedal force in the simulation mode was 40 pounds (180 newtons) for the max- Normal rudder trim w a s used.

imum pedal travel of 1.5 inches (3.8 centimeters).

Safety Provisions The safety pilot is in command of the airplane and has the primary responsibility for the safety of the flight. In the simulation mode, the safety pilot monitors the total control inputs, which are the sum of evaluation pilot's inputs and the simulation system inputs. The safety pilot is prepared to take over the controls if a hardover input occurs or if a maneuver becomes excessive. The simulation can be disconnected electrically by either pilot. It will also disconnect automatically if the interface detects a malfunc- tion in the computer. If the electrical disconnects should f a i l to operate, the safety pilot can overpower the system with manual inputs.

The control forces required to overpower the system are:

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25 pounds (111 newtons)

Elevator

Lateral control . . . . . . . . . . . . . . . . . . 45-pounds (200 newtons) at 75' wheel

Rudder . . . . . . . . . . . . . . . . . . . . . . . 13 pounds (58 newtons) at 6 , = 10'

Thrust modulators . . . . . . . . . . . . . . . . . 60 pounds (267 newtons) (total)

In this test program the safety pilot occasionally disconnected the simulation and took over the controls near the ground when he felt that a poor landing touchdown was in prospect. A large red warning light (shown in fig. 2-8) notified the evaluation pilot that disconnect had occurred. Flight safety w a s further augmented by the limited authority of the simulation system which was designed to prevent overloading the structure. Simu- lator authority limits for each control system are shown in the specifications.

SIMULATION CHECKOUT PROCEDURES Each SST configuration tested w a s programed on a separate computer patchboard.

Each patchboard w a s wired to modify the 367-80 characteristics to simulate the response of the desired SST. Each patchboard included an analog model of the 367-80 airplane.

With this model, it w a s possible to run ground checks on the simulation prior to flight tests. The ground checkout procedure w a s to pulse the 367-80 controls from the com- puter in the simulation mode and determine the response of the analog model of the 367-80. The transient response was

then compared with six-degree-of -

freedom digital- computer results and

_ _ _ _ Calculated

Flight test modified, if necessary, by potentiom- eter adjustments.

Pitch rate, deg/sec 0 After a good match of transient -1 w a s obtained on the ground, responses -2 the analog model of the 367-80 airplane was disconnected and the same check pulses were repeated in flight. The con- trols which were pulsed included the Elevator deflection, deg elevator, rudder, wheel, and spoilers.

I . L L l J (Transient response to thrust inputs - 2 0 2 4 6 8 10 12 14 16 18 Time, sec was checked with step inputs of the thrust modulators.) Each pulse had a (a) Pitch-rate response to elevator pulse.

1-second rise time, 2-second dwell, Figure 2-9.- Examples of response to pulses i n simulation mode.

and 1-second return. The magnitudes Variablegeometry configuration at 9OOO ft (460 m) and 136 knots.

, I 1 1 1 1 1 1 1 i . - - C ' i i l l l l l l l i l / 1 Ellflll II I ' 1 1 1 1 of the pulses were limited to produce moderate.airplane response. The in-flight pulse responses were recorded on a direct-writing 18-channel oscillograph and immediately compared with six-degree-of -freedom digital-computer results which were plotted on transparencies for convenient in-flight comparison. Such comparisons of short-period and phugoid responses are shown in figures 2-9 and 2-10 for elevator and rudder pulses.

Some adjustment of parameters by means of potentiometer adjustment was usually

4 t n

-2t \ I

~ Flight test --- Calculated

-t

-6 c

I I I I I I I I 0 5 1 0 15 20 25 30 35 Time, sec (b) Lateral and directional response to a rudder pulse.

Figure 2-9.- Concluded.

I. 1111 I 1 1 . 1 1 1 I I, I I I I . . 1 . 1 , 1 . -

Flight test

--__

Calculated 156.

148 .

- 1 - L-- . t I I I I 1 0 20 40 60 80 1 0 0 Time, sec Figure 2-10.- Comparison of measured and calculated phugoid oscillatory mode of.

variable-geometry SST configuration.

required to obtain the proper SST transient responses. The fact that such adjustments were somtimes required was an indication that some of the 367-80 characteristics were not known with sufficient accuracy or that approximations used or assumptions of linear- ity were not completely valid.

When the oscillatory modes and transient responses were considered to be satis- factory, the static characteristics of the SST configuration were documented. These characteristics include the variation of lift coefficient with angle of attack, the variation of drag or power required with airspeed, control force and deflection for steady turns, static longitudinal stability, control power in pitch and roll, and steady sideslip parameters.

TEST INSTRUMENTATION A comprehensive system of recording instruments was used for this investigation.

The data were recorded on 1-inch (2.5-cm) magnetic tape which was processed by an automatic data-reduction and machine-plotting system. More than 200 parameters were for troubleshooting. Most parameters recorded but about half of these were intended only were sampled 2.5 times per second. However, this sampling rate was not always ade- quate, and therefore 40 variables were recorded continuously. The nominal instrument accuracy was 2 percent of full scale. Sensitivities and full-scale values were adjusted to be compatible with the test program.

Input quantities which w e r e recorded included the deflection of control column, wheel, rudder pedals, elevator, ailerons, spoilers, rudder, stabilizer, electric throttle, and thrust-modulator levers. Engine data w e r e recorded to permit determination of thrust. The commands to the airplane from the analog computer were also recorded.

The recorded airplane response quantities included airspeed, pressure altitude, geometric altitude (when over the runway), ILS localizer and glide-slope errors, angular velocities, linear accelerations, pitch and roll attitude, incremental heading change, angle of attack, and sideslip.

The angle-of-attack and sideslip sensor was a four-element cruciform wooden vane assembly with a natural frequency greater than 20 cycles per second which w a s mounted on a 17-foot (5.2-meter) conical boom ahead of the airplane nose. (See fig. 2-1.) Since the vanes were less than 1.5 fuselage diameters ahead of the nose and approximately three mean chord lengths ahead of the wing-fuselage juncture, the vane e r r o r s due to upwash and sidewash were large and required correction. The angle-of-attack vane w a s calibrated in flight by the plumb-bob method, and the upwash correction w a s determined to be between 29 and 30 percent. The sideslip vane correction for sidewash w a s esti- mated to be 20 percent based on vector analysis of Dutch roll data. The vane angles w e r e corrected for the e r r o r due to angular velocity. The vane data were also corrected with a lag function for the time required for the airflow measured by the vanes to reach the airplane center of gravity. The expressions for corrected flow angles were of the following form:

a = 1*2g~ndicated + e g - 2

( l + S $ ) 2 la2&ndicated

- 4 , P =

( l + S $ ) DATA REDUCTION Most of the data were reduced automatically from the tape by using routine methods and machine plotted. However, some discussion of the determination of l i f t and drag is desirable.

The incremental values of l i f t coefficient and angle of attack from the trim point are used to show the measured variation of normalized l i f t ACLqoS/mVo with angle of attack for the simulator.

The flight conditions used f o r the lift data were also used for obtaining the drag variation with speed since the thrust was held constant and equal to the value required for level flight at the initial trim speed. Measurement of drag in flight is very difficult and usually contains a rather large amount of scatter. Since, in most cases, the airplane was not completely stabilized at a steady speed at any time, the following expression was used to determine the drag or thrust required for the supersonic transport: dh /dt

DSST = - $(%) - w ( :o I-,, ( % ) m-80

The slopes dVe/dt and dhp/dt were measured from 3- to 4-second time histories of Ve and hp when the rates were nearly constant. For the purpose of data reduction the thrust T for each set of data was assumed to be a value which would make the drag for the trim condition agree with that calculated for the SST. Thus the datum for drag variation with speed was somewhat arbitrary; however, the incremental variation of drag

with speed was not influenced by the assumed thrust value. That is, the value of aT/W

ave is measured correctly.

STABILITY AUGMENTATION Quickened Pitch Response The probable need for quickened longitudinal response of very large airplane con- figurations, in particular for the flare and touchdown, has been widely recognized. There- fore, provisions were made to investigate the effects of augmented pitch response on the landing-approach characteristics of the configurations used in this program. A ground- based simulator investigation was made to evaluate techniques for augmenting the pitch response.

As a result of this study, it was decided to use a pitch damper (6 feedback) in com-

bination with increased gearing between the column and the elevator as the stability- augmentation device. The augmentation system can be represented by the expression

K16, + K2fj where K1 is 2.0 and K2 is 1.46. Such a system has the

6, = ( 2 )

C basic advantage of being easier to implement on the test airplane than a second-order lead-lag system. Figure 2-11 illustrates the effect of a pitch-rate feedback system on the response to a step column input. It is evident that the feedback washes out the elevator deflec- tion as the pitching velocity builds up. Therefore it is possible to increase the gearing between the column and the elevator gearing to increase the initial pitching moment due . to column deflection and thereby to quicken the pitch response without any tendency for the peak pitching velocity to become excessive. However, the static stability apparent to the pilot is reduced.

The proposed pitch-augmentation system was further refined by adding angle-of-attack feedback.

In this case the equation for elevator deflection (not including simulation inputs) is

K16, + K26 + K3 ACY

6, = ( F )

C basic The ratio of elevator to column gearing K1 w a s increased to 4. The gain on the 8 feedback K2 w a s still 1.46. The gain on the A a feedback K3 w a s selected to keep the static longitudinal stability ($$ approximately equal to the unaugmented value.

/--- -- - -

The values selected for K3 were 1.5 for the 'I------ variable-sweep configuration and 1.0 for the fixed- geometry configuration. Figure 2-11 indicates Time that the estimated value of elevator deflection in response to a step column input is initially much Figure 2-11.- Examples of longitudinal response to a step column input u s i n g two types of

larger than the unaugmented value, but then, as 6

stability augmentation with increased control gearings.

and A a build up, it approaches the unaugmented value. As is discussed in part 4 of this compilation, this type of augmentation improves the pitch response by increasing the frequency of the short-period oscillation. Although these augmentation systems were not optimized, the gain settings selected initially were found to be satisfactory for this test program.

Dutch Roll Augmentation Dutch roll augmentation w a s used in some tests of the variable-geometry SST con- figuration to increase the damping ratio from approximately 0.2 to 0 . 3 .

The increased damping w a s provided by a sideslip rate yaw damper. The sideslip rate was computed from the expression The rudder w a s driven to oppose the sideslip rate with a gain of -1, so that A 6 r = - P This type of augmentation was devised at the NASA Ames Research Center and is dis- cussed in reference 3.

Augmented Roll Damping To increase the roll damping of the fixed-geometry SST configuration, the aug- mented version of this configuration included the following equivalent incremental wheel input AtiW = -0.454 As a result of this input the calculated roll time constant was decreased from 0.80 to 0.58 second.

QUALITY OF SIMULATION The simulation w a s believed to be valid for a speed range of *lo knots, an angle- of-attack range of *2O to *3O, and for a range of normal acceleration values of k0.3 to *0.4g. A complete documentation of the simulated configurations is given in reference 4.

The simulated steady-state f lying-qualities data usually matched the design values within *25 percent. Figure 2-9 compares examples of measured transient short-period response to elevator and rudder pulses with six-degree-of-freedom computed data. Fig- u r e 2-10 shows a typical example of the realized phugoid mode for one test configuration compared with the calculated phugoid oscillation. Since this simulator has the capability of varying lift characteristics, which is not common to other in-flight simulators, it does permit simulation of the phugoid mode.

0 Simulator (measured i n flight)

- Supersonic transport (calculated)

1 1 . 7 9

/.f

- . -- I I 1 I I 0 1 2 -2 -1 Angle of attack increment Angle of attack increment Angle of attack increment from trim, deg from trim, deg from trim, deg (a) Fixed geometry.

(b) Variable geometry. (c) Variable geometry (cruise).

Figure 2-12.- Variation of normalized lift with angle of attack using increments from trim.

0 Simulator (measured in flight) - Supersonic transport (calculated)

- 4 1

-9 L I I I I I 1 1. I --

-120 130 140 150 120 130 140 150 130 140 1 5 0 160 Ve, knots Ve, knots Ve, knots Variable geometry Fixed geometry Variable geometry, (Cruise sweep) Figure 2-13.- Variation of thrust required with airspeed for supersonic transport configurations.

(Landing-approach condition.)

Figure 2-12 shows the calculated and measured variations of normalized l i f t with angle of attack. The agreement was good, except for the emergency, cruise-sweep La was about 15 percent high.

landing configuration for which the measured value of Figure 2-13 presents the calculated and measured variation of power required for speed changes from the trim speed.

REMARKS ON OPERATIONAL EXPERIENCE The 367-80 airplane was flown approximately 125 hours in connection with this test program. The simulation equipment proved to be very reliable with small loss of flight time due to equipment malfunction. The approximate efficiency of this simulator in t e r m s of productive use of flight time for its initial test program is indicated by the fol- lowing tabulation: Approximate percent of flight time Flight tests required to determine 367-80 airplane

characteristics with spoilers at 0 ' to 1 0 ' . . . . . . . . . . . . . 5

Functional check of simulation equipment . . . . . . . . . . . . . . 10

Setup and checkout of test configurations . . . . . . . . . . . . . . 25

Simulation check runs (on each flight) . . . . . . . . . . . . . . . . 15

Documentation of test configurations . . . . . . . . . . . . . . . . . 15

Pilot evaluation of SST configurations at altitude

and during landing approach . . . . . . . . . . . . . . . . . . . . 30

The percentage of flight time required for setup and checkout should be lower for additional test programs with this system. However, it should be noted that the setup and checkout times as well as the simulator capabilities are influenced by the character- istics which are to be simulated. For example, the 367-80 airplane with the center of

gravity at 30 percent c had a large static stability margin. It was found to be difficult

and time consuming to set up the simulator f o r the small static margin of the fixed- geometry configuration. This problem might be alleviated in other such projects by actually shifting the center of gravity of the test airplane.

The accuracy of simulation is affected by the amount of time that can be allotted to setting up and checking out a simulated configuration. In the SST landing-approach sim- ulation program, the relatively small amount of flight time which could be budgeted to each test configuration required that the matching of actual to desired response be done as quickly as possible. Therefore, the accuracy of simulation obtained in this program may not represent the full potential of the simulation equipment.

REFERENCES 1. Cooper, George E.: Understanding and Interpreting Pilot Opinion. Aeron. Eng. Rev., vol. 16, no. 3, Mar. 1957, pp. 47-51, 56.

2. Robbins, R. E.; and Person, S. D.: 367-80 Airplane Variable Stability Simulation System (NASA Langley Supersonic Transport Simulation Program). No. D6- 19856 (Contract No. NAS 1-4096), The Boeing Co., 1965. (NASA CR-66126.)

3. Quigley, Hervey C.; Innis, Robert C.; Vomaske, Richard F.; and Ratcliff, Jack W.: A Flight and Simulator Study of Directional Augmentation Criteria of a Four- Propellered STOL Airplane. NASA TN D-3909, 1967.

4. Eldridge, W. M.; Condit, P. M.; Schwanz, R. C.; and Taylor, C. R.: Simulation of Three Supersonic Transport Configurations With the Boeing 367-80 In- Flight Dynamic Simulation Airplane. No. D6-10743 (Contract No. NAS 1-4096). The Boeing CO., 1965. (NASA CR-66125.)

TABLE 2-1.- MASS AND DIMENSIONAL CHARACTERISTICS OF SIMULATED TEST CONFTGURATIONS Variable- Variable- Fixed- geometry SST 367-80 airplane Characteristics geometry geometry (cruise SST SST sweep) ( 4 (b) Weight: 150 000

Ib . . . . . . . . . . . . . 280 000 280 000 270 000

N . . . . . . . . . . . . . 1 245 500 1 245 500 1 201 020 667 000

Center-of-gravity location, 46 35 46 30 percent C . . . . . . . .

1x:

slug-ft2 . . . . . . . . . 2.86 X lo6 2.22 x 106 1.667 X lo6 2.57 X 106

7.11 X 106 5.52 X lo6 4.14 X lo6 6.38 x l a 6 kg-m2 . . . . . . . . . .

Iy: 17.57 x lo6 18.58 X lo6 2.25 X lo6

slug-ft2 . . . . . . . . . 18.11 X lo6

kg-m . . . . . . . . . . 43.65 X lo6 45.00 X lo6 46.16 X l o 6 5.59 x 106

Iz: 20.00 x 106 4.73 x 106

slug-ft2 . . . . . . . . . 20.00 x 106 20.00 x 106

49.69 X l o 6 49.69 X lo6 49.69 X lo6 11.76 X l o 6 kg-m . . . . . . . . . .

Ixz: 0 0 0 0.160 X l o 6 slug-ft . . . . . . . . .

0.22 x 106

kg-m . . . . . . . . . . 0 0 0

20 63 72

A, deg . . . . . . . . . . . 35

S: 5000 2821

ft2 . . . . . . . . . . . . 1 5000 1 8000 1

I 464.50 I 743.20 I m . . . . . . . . . . . . 464.50 2 57

- C: 70 89 20.1 f t . . . . . . . . . . . . .

21.336

m . . . . . . . . . . . . . 21.336 27.127 6.12

b:

f t . . . . . . . . . . . . . 85 111 130.8

25.91 33.83 25.91

m . . . . . . . . . . . . . 39.8

35 for variable-geometry SST 182 135 for fixed-geometry SST

vtrim, knots . . . . . . . . 135 135

c 150 for SST at cruise sweep

SST 12.3 5.45 for fixed-geometry SST

deg . . . . . . . . . 6.6 12

0 0 2.0 iw, deg . . . . . . . . . . .

- bAl1 parameters for variable-geometry configuration a r e based on geometry of cruise-sweep configuration.

II I I I I I I I I I II IIIIIIIII1111 111 I I 1 I 111111 11111111111.11111111111.1111.

TABLE 2-2.- DESIGN AERODYNAMIC PARAMETERS OF SUlIILATED SST CONFIGURATIONS Flxed geometry Variable geometry E m ~ r g e n c Increase Augmenta Augment# 72 Swep1 speed- Augmente (4 + A d Degrads Augment< 4ugmepte A"gFe"l (4 + A 4 Degrade :on€iguratit thrust Basic Basic changed + Ad, C"@ C", (8 +Am), degrade c . c , stabilily 0, B 8, aft c.

czs cn, B, aft C.6 " 4 ' Cn$ Cn -~ CD, 0.125 0.125 0.125 1.125 1.115 0.115 0.115 0.115 0.115 0.115 0.115 0.145 1.203 C D ~ , per radian 1.203 1.203 .61 ,418 .418 .418 ,418 .418 .418 ,418 .573 .54 .54 .54 .54 a 9 3 .a93 ,893 A93 .893 .a93 .e93 ,4507 CL,trim 3.266 3.266 3.266 8.266 7 4.7 CL,, per radian 4.7 4.7 4.7 4.7 4.7 3.209 .a022 ,8022 C L ~ , per radian ,8022 .8022 .487 ,467 ,487 ,487 .487 ,487 .487 .487 c , , : per radian -.0802 >-.3672 a-.3672 ,0802 .45M -.4584 -1.533 -1.215 -.I41 -1.533 -.4584 -.3438 Cmh, rad/sec 1 0 0 I ,1335 -.1335 -.1335 -.1335 -.1335 -.1335 -.1335 -.a288 -.5947 -.5947 .1751 cmB' rad/sec -.1757 ,2149 -1.261 -1.261 -1.261 -.2149 -1.261 -2149 -.1596 cm6e, per radian - 2 8 7 - 2 8 7 -.287 2 8 7 .'I163 -.7163 -.7163 -.'I163 -.7163 -.7163 -.7163 -.7163 ,045x10-f ,045X10' .045X10-' .045XlO-' 231x10 231x10 .231X10 cmAT, per Pound .231x10 .231x10 .23IXlO- 23lX10- .1275YIO~ per newton .0101x10- .0101X1C .0101x10 .0101x10 .052X10 .052XlO .052X10 .052X10 .052X10 .052X10- .052X10- .0287X10-' czB. per radian ..0825 -.0825 -.0825 ,0825 .1547 -.1547 -.E47 -.1547 -.1547 -.1547 -.E47 -.la91 -SI438 -.0696 -.0438 ,0696 cL +, rad/sec ,2269 -.2269 -.2269 - 2 2 6 9 -2269 -.2269 -.2269 -.0249 cz,p rad/sec .073 SI73 .073 ,073 ,0744 ,0744 .0744 ,0744 .0744 .0744 .0744 .0208 ,0573 .0573 ,0573 ,0573 ,1146 .1146 .1146 .I146 .1146 .1146 .1146 ,0129 C Z v per radian per radian 0 ) 0 0 0 0 1 0 0 I Cz .131 .131 ,131 .I31 ,2006 ,2006 .2w6 2006 ,2006 ,2006 ,2006 .1604 -.0152 -.0352 ,0152 .a049 ,0223 -.0223 -.0223 -.0223 ..0223 -.a76 -SI76 .0067 .lo2 -.lo2 -.lo2 ,102 ,0874 -.0874 -.0874 -.OS74 ..OS74 -.0874 -.0874 ,0554 I 0 -.138 .OS59 .0859 ,0859 ,0859 -.1204 -.1204 I .OB9 .a229 .0229 ,0229 0424 .a424 .a424 .0424 .a424 ,0424 ,0424 ,002 ,0745 -.0745 -.0745 ,0745 086 - 8 8 6 -.OS6 -.OS6 .086 -.OS6 -.OS6 .086 cy@, per radian -.5272 ,5272 -.5272 ,5272 573 -.573 -.573 -.573 ,573 -.573 -.573 ,4928 cy., rad/sec ,0487 ,0487 .0487 ,0487 0253 ,0253 ,0253 .0253 ,0253 ,0253 .0253 ,0346 @ Cy+, rad/sec ,146 ,146 146 ,146 093 ,093 .093 ,093 8 9 3 .093 ,093 ,0692 Cyv Per radian 0 0 0 0 0 0 1 .I146 .1146 .I146 1146 1146 .1146 ,1146 Cya,, per radian .1146 ,1146 ,1146 .1146 .1146 Short-period frequenc) 0.754 1.46 1.46 0.754 0.885 1.303 1.743 0.641 1.743 1.63 0.885 0.981 r a d / W Short-period damping ,793 367 .793 ,846 .672 .938 ,705 .755 .945 .705 .672 .569 ratio Phugoid frequency, ,117 .126 .126 ,117 .170 .114 .156 .149 .132 .156 .168 .129 rad/sec Phugoid damping ratio -.024 ,057 .057 .113 ,019 ,177 .093 ,102 ,047 ,093 ,022 .170 Dutch roll frequency, ,829 .811 .982 ,829 ,628 ,621 ,621 ,621 ,621 .642 .692 1.24 d / S W Dutch roll damping ,381 ,379 .05 .379 2 8 2 ,186 2 8 2 2 8 2 2 8 2 .051 .051 ,172 ratio Spiral-divergence time 74.9 109.2 99.6 149.0 345.0 345.0 ,092 345.0 345.0 397.0 397.0 -17.6 constant, sec Roll-convergence time .so2 .573 ,885 .573 .48 .48 .48 .48 .48 .49 .49 1.7 constant, sec -1.0 -4.0 -4.0 -1.0 -1.3 -2.6 -5.2 -5.2 -1.3 -5.2 -1.3 -1.3 b;/:c 1.46 0 1.46 0 0 1.46 1.46 1.46 0 1.46 0 ae/9 0 0 1.0 1 . 0 0 0 0 1.5 1.5 0 1.5 0 0 27.9 " . / E , Pounds (Paups) 32.8 21.9 32.8 30.8 33.6 28.2 24.8 14.2 30.8 29.7 28.2 143 190 190 143 38 55 165 .65 136 64 138 132 per h o t ,0024 -.0024 -.a024 ,0006 ,0005 ,0005 ,0005 .0005 .WO5 ,0005 .0005 ,0013 Center-of-gravity loca- 35 35 35 35 16 1 6 46 53 53 46 46 l tion, percent i : _ _ 'Values with augmentation on.

TABLE 2-3.- COOPER PILOT-RATING SYSTEM ~~~ ~~ Mode of Adjective Numerical Description operation rating rating iccomplished Yes Excellent, includes optimum Yes Yes Yes Good, pleasant to fly Normal ' Satisfactory Yes Satisfactory, but with some mildly Yes

' I

unpleasant characteristics Yes Acceptable, but with unpleasant Yes characteristics

Emergency 1 Unsatisfactory Unacceptable for normal operation Doubtful Yes

Doubtful Yes Acceptable for emergency condi- tion only: Failure of stability augmenter No Doubtful Unacceptable even for emergency condition: Failure of stability Inoperable Unacceptable augm enter No

Unacceptable - Dangerous No

No No

Unacceptable - Uncontrollable

- Inoperable Catastrophic No No 10 Motions possibly violent enough to prevent pilot escape TABLE 2-4.- AERODYNAMIC CHARACTERTSTICS OF 367-80 AIRPLANE Fixed and Variable geometrj Parameter variable geometrj at maximum wee]

I

, . i

0.1165 0.0892 CD. t r i m . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

0.515 0.327 CD,, per radian . . . . . . . . . . . . . . . . . . . . . . . . . . .

0.856 0.6935 CL, trim . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

4.55

CLa, per radian . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.9

0.244 0.244 CL~,, per radian . . . . . . . . . . . . . . . . . . . . . . . . . .

-1.008 -1.11 Cm,, per radian . . . . . . . . . . . . . . . . . . . . . . . . . . .

-0.261 -0.361 Cmk, rad/sec . . . . . . . . . . . . . . . . . . . . . . . . . . . .

-0.594 -0.425 Cm8. rad/sec . . . . . . . . . . . . . . . . . . . . . . . . . . . .

-0.85 -0.9 Cmg,, per radian . . . . . . . . . . . . . . . . . . . . . . . . . .

2 x 10-6 2 x 10-6 CmAT, per pound . . . . . . . . . . . . . . . . . . . . . . . . . .

45 x 10-8 45 x 10-8 per newton . . . . . . . . . . . . . . . . . . . . . . . . . .

Czp, per radian . . . . . . . . . . . . . . . . . . . . . . . . . . . -0.1572 -0.143

-0.1569 -0.136 CZ 4, rad/sec . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

0.0817 0.0320 Cz4' rad/sec . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

0.0653 0.077 C p e r r a d i a n . . . . . . . . . . . . . . . . . . . . . . . . . . .

2 b' 0.0179 0.0202 C . per radian . . . . . . . . . . . . . . . . . . . . . . . . . . .

Z g r

Cnp per radian . . . . . . . . . . . . . . . . . . . . . . . . . . . 0.0797 0.1167

-0.0225 -0.0166 Cn4, rad/sec . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

-0.0467 -0.0189 Cn* rad/sec . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Cnp, rad/sec . . . . . . . . . . . . . . . . . . . . . . . . . . . . . -0.043 -0.027

0.0082 0.0156 Crib, per radian . . . . . . . . . . . . . . . . . . . . . . . . . . .

-0.0725 -0.068 Cng,. per radian . . . . . . . . . . . . . . . . . . . . . . . . . . .

0.0245

Cn6 . per radian . . . . . . . . . . . . . . . . . . . . . . . . . . . 0.0245

S -0.825

Cyp, per radian . . . . . . . . . . . . . . . . . . . . . . . . . . . -0.831

0.1492 0.0864 Cy+. rad/sec . . . . . . . . . . . . . . . . . . . . . . . . . . . .

0.0865 0.0764 Cy+, rad/sec . . . . . . . . . . . . . . . . . . . . . . . . . . . .

-0.0128 . 0 . 0 128

Cyb, per radian . . . . . . . . . . . . . . . . . . . . . . . . . .

Cyg,, per radian . . . . . . . . . . . . . . . . . . . . . . . . . . 0.1712 0.0177

1.68 1.53 Short-period frequency. rad/sec . . . . . . . . . . . . . . . . . .

0.702 0.698 Short-period damping ratio . . . . . . . . . . . . . . . . . . . . .

0.134 0.138 Phugoid frequency , rad/ s ec . . . . . . . . . . . . . . . . . . . . .

0.282 0.096 Phugoid damping ratio . . . . . . . . . . . . . . . . . . . . . . . .

0.844

Dutch roll frequency, rad/sec . . . . . . . . . . . . . . . . . . . . I 0.799

0.0419 0.091 Dutch roll damping ratio . . . . . . . . . . . . . . . . . . . . . .

. 188.8 127 Spiral-divergence time constant. sec . . . . . . . . . . . . . . . .

0.665 0.657 Roll-convergence time constant. sec . . . . . . . . . . . . . . . .

Flap deflection, deg . . . . . . . . . . . . . . . . . . . . . . . . .

- 6 L- Initial 6,.deg . . . . . . . . . . . . . . . . . . . . . . . . . . . .

-6 I

3. PERFORMANCE CHARACTERISTICS By Albert W. Hall SUMMARY Some performance characteristics are presented which were determined during the in-flight simulation study of supersonic transport landing-approach configurations. The normal load factor and attitude changes resulting from maneuvers during instrument approaches and landing flares a r e presented. The landing flare and effects of speed- thrust stability a r e illustrated and discussed.

INTRODUCTION In this part a r e presented results which are applicable to some of the future per- formance certification requirements of supersonic transports during the landing approach.

The normal load factor and attitude changes resulting from maneuvers during instrument approaches and during the landing flare are given for approaches made during this inves- tigation. The landing-flare characteristics and the effects of speed-thrust stability are illustrated and discussed. These results a r e presented both as flight test data and pilot opinions. The emergency landing configuration (variable geometry with the wings swept in the cruise position) is not discussed in this part because the limited time of this pre- liminary investigation allowed only one instrument approach with this configuration.

RESULTS AND DISCUSSION ILS Approaches Selection of approach speeds.- The design requirements for economical high-speed cruise flight can result in high body attitudes for both the variable-geometry and fixed- For the fixed-geometry geometry supersonic transports in the landing configuration.

configuration, the minimum approach speed which gives adequate tail clearance during a landing probably will be greater than the presently required value of 1.3 times the stall Therefore, approach speeds for the fixed-geometry SST configuration may be speed.

based on attitude rather than on speed margin. It is conceivable, though less likely, that the approach speeds for the variable-geometry SST configuration may also be based on attitude.

The approach and landing attitude requirements should allow a sufficient tail clear- ance margin for operational variations in speed and unexpected maneuvers near the ground. This problem could not be examined during the present investigation because the tail clearance and body attitude at touchdown were not simulated and therefore only the incremental attitude changes could be correctly simulated. (A comparison of touch- down attitude for the simulator and SST configurations is shown in figure 3-1.) There- w a s selected prior to this investigation rather than being an fore, the approach speed objective of the investigation. The selected approach speed was 135 knots since one of the early design objectives of the United States supersonic transport program was to have an approach speed no greater than 135 knots at the maximum landing weight.

Fixed-geometry SST l L 5 L /-Variable-geometry SST ,-Simulator Figure 3-1.- Touchdown attitude for simulator and SST configurations.

Longitudinal maneuvers during approach. - The selected approach speed margin and

corresponding maneuver capability are expected to be more than adequate for the fixed- The maximum variations and variable-geometry SST configurations of this investigation.

of attitude and load factor measured during instrument approaches with these configura- tions are presented in table 3-1. These approaches were made in calm air with no inter- fering traffic and no cockpit distractions. The attitude and load factor variations are probably smaller than those which might occur in turbulent air with minimum weather conditions at a busy terminal with a maximum of aircraft-ground communications after a long flight which has induced pilot fatigue. These data were also affected by the fact that the pilots were trying to keep the airspeed within 4 0 knots of the trimmed approach speed (135 knots) in order to maintain a valid simulation as discussed in part 2 of this is an artificial restriction which would not be present in the actual super- paper. This sonic transport and the effect of this restriction on the approach techniques used herein is not known.

The pitch attitude data for each approach shown in table 3-1 represent the maxi- mum increments above and below the nominal simulator attitude ( 1 / 2 ' nose up for the glide slope). The maximum nose-up attitude increments during each approach were generally less than 4O and the maximum nose-down incre- ments w e r e usually less than the nose-up value. (See table 3-1.)

The distribution of maximum normal load factor is shown in figure 3-2 for the 54 approaches given in table 3-1 for both fixed- and variable -geometry configurations. From figure 3-2 it is seen that 35 percent o f the approaches had a maximum load factor between 1.15 and 1 . 2 , only 9 percent of 1.0 1.1 1.2 1.3 the approaches had load factors between 1.25 and 1.3, and the Load factor load factor did not exceed 1 . 3 for any of the approaches.

Figure 3-2.- Distribution of Time histories of altitude, load factor, body attitude, maximum normal load factor resulting from maneuvers and flight path for the latter part of one instrument approach during 54 instrument approaches for the super- with the variable-geometry configuration are presented in sonic transport configura- figure 3-3 to illustrate the data given in table 3-1. tions investigated.

Time. sec Figure 3-3.- Time histories during the latter part d an instrument approach with the variable-geometry configuration having 6 and 6 augmentation.

P Just prior to the transition from instrument to visual flight the attitude dropped to a negative value. At the transition to visual flight the pilot increased the attitude rather abruptly to about 2.8O, which resulted in a 1 . 3 load factor. This happened for several approaches where the highest load factor during the approach occurred during the transi- tion from instrument to visual flight.

Landing Flare Touchdown attitude.- ~ The distribution of body attitude at touchdown is shown in fig- ure 3-4 for both the fixed- and variable-geometry configurations. These distributions include data from several visual approaches in addition to the instrument approach data 3-1. As mentioned in the discussion of l i f t characteristics in this section, the of table supersonic-transport attitude was not matched by the simulator. The relation between simulator attitude and SST attitude is described in part 2. The simulator and corre- sponding SST attitudes a r e shown in figure 3-4.

lo The measured touchdown zttitudes are 7Lo to 8- less than the maximum ground 2 2 attitude of the simulator but present indications a r e that some SST configurations will be landing at an attitude very close to the maximum ground attitude. The pilots, there- tail-first contact; fore, were landing the simulated configurations with no apprehension of this is not a realistic simulation and it is very probable that the distribution of touchdown

50 r r

1 2 3 4 5 Simulator attitude, deg Simulator attitude, deg L I I I I J I I I I 9 10 11 12 13 4 5 6 7 8 SST attitude, deg SST attitude, deg (a) Fixed-geometry SST (b) Variable-geometry SST (16 landings). (36 landings).

Figure 3-4.- Distribution of touchdown attitude for two supersonic transport configurations.

I' attitudes would be lower (further from the tail-drag limit) than that shown in figure 3-4.

The lower touchdown attitudes would be accompanied by increased approach speeds.

As indicated in part 2, the simulation w a s only valid for a speed range 10 knots above and below the trim speed. Therefore, during the landing flare the pilots had an unrealistic task of keeping the airspeed above 125 knots. The effect of the two simulation deficiencies is not known; however, they tend to be compensating. The improper ground

attitude allows higher than normal touchdown attitudes, while the minimum speed restric -

tion causes lower than normal touchdown attitudes.

In flying the 3 ' approach path, the body attitude for the simulator was about 1/2O nose up for both the fixed- and variable-geometry simulations. From figure 3-4 it can be seen that, generally, the final touchdown attitude for the simulator was from about lo

2' to 3- . This attitude increase represents the increment

required during the landing flare for these configurations with the ground effects that were simulated for this investigation.

Because of simulator limitations only about one-third of the estimated increase in lift coefficient resulting from ground effect could be simulated for the fixed-geometry configuration. If no change in lift-curve slope is assumed, the difference between the ground-effect lift increment estimated for the fixed-geometry SST and that actually sim- a 2' angle-of-attack increment. In other words, the touchdown ulated was equivalent to lift coefficient represented by the data of figure 3-4(a) would have occurred at a body attitude 2' lower than that shown and a large portion of the touchdowns would have occurred with the attitude close to that for the approach. These results would then be in agreement with preliminary data from various sources which have indicated that some fixed-geometry configurations tend to be "automatically flared" through favorable ground is held constant.

effect when the approach attitude The term "estimated ground effect" is used in this discussion and is fairly descrip- tive of most of the available ground effect information applicable to SST configurations.

Considering the relationship between ground effect, touchdown attitude, and approach speed, it is therefore very important that efforts be made to obtain reliable data on ground effects for SST configurations.

Flare-path control.- Precise flare-path control is required in order to have the airplane touch down at a particular point with a low rate of sink (vertical velocity). The ability to touch down near a given runway location is required if each landing is to be completed within a predictable landing-field length. As discussed in part 4, longitudinal augmentation w a s required to improve longitudinal control of the flare path. The longitudinal augmentation was found to be very effective in allowing more precise control of the flare path for both the fixed- and variable-geometry configurations.

The pilots reported that the variable-geometry configuration had an unusually large floating tendency near the ground during the landing flare. The floating tendency was not as pronounced for the fixed-geometry configuration but it should be remembered that the full amount of estimated ground effect w a s not mechanized in this simulation. Although the ground effect could not be changed by longitudinal augmentation, it appeared that the more precise control characteristics of the augmented configuration made it easier to overcome t h e floating tendency.

Measured and calculated flare paths.- .- The major effort in the estimation of landing performance involves the determination of the landing-flare distance.

The calculated and measured flare paths for the fixed-geometry supersonic trans- port configuration are shown in figure 3-5. The measured data are representative of a good landing where the touchdown occurred near the desired location, with a low rate of sink and with no floating or "feeling for the ground." The calculated flare is based on a method explained in reference 1 which a point mass moving in a plane -m involves 6 0 - with two degrees of freedom under the action

- Measured

- 1 5

of known forces. For the calculated flare ---

Calculated E path, thrust was assumed to be constant at -10 the value for a 3O approach and a constant

z

= U < load factor was assumed. For the measured flare path, the thrust was maintained at the

-

approach value but the load factor varied as shown in figure 3-5. For considerably the calculations a constant load factor of 1 . 0 4 was required to decrease the rate of sink from 1 1 . 0 feet/sec to 1 . 2 feet/sec ( 3 . 4 to 0 . 3 7 m/sec) in a vertical distance of 50 feet (15 m). The initial conditions of forward speed and vertical velocity were taken from the flight test data at an altitude .8L I I I -I I I of 50 feet (15 m) and the terminal condition 0 400 800 1200 1600 m was taken from the vertical velocity meas- Horizontal distance, ft ured at touchdown. The calculated flare

u L.- u

0 m 400 600 path is very close to the measured path.

Horizontal distance, m The measured load factor varies consider- Figure 3-5.- Measured and calculated flare path for ably above and below the constant value used fixed-geometry supersonic transport configuration.

in the calculated flare; however, the constant factor represents a good average value.

On the basis of this result and other results not shown herein, it is believed that the constant-load-factor method of computing landing-flare parameters (ref. 1) will give a good estimate of the supersonic transport flare characteristics if a load factor of the order of 1.05 is used.

From figure 3-5 it can be seen that the maximum load factor used during the flare or pulsating variation of load is much greater than the average value. The oscillatory factor shown here is typical in magnitude of all the landings recorded during this inves- tigation. (See table 3-1 for comparison of maximum load factor values.)

The curve showing speed loss in figure 3-5 was based on a drag-lift ratio for free air but the use of the value for f u l l ground effect gave less than 0 . 5 knot difference between that based on free air.

Speed-Thrust Stability Limiting values of speed-thrust stability.- The fixed-geometry SST is expected to fly on the back side of the thrust-required curve during the landing approach; conse- quently, there has been much discussion concerning a possible criterion to define a toler- able level of speed-thrust stability. The piloting problems associated with back side operation are significant when flying under flight-path constraint such as during an instrument approach (ref. 2 ) . Some investigations (such as ref. 3) have indicated that it is desirable to have stable speed-thrust characteristics (front side) for an instrument approach, whereas some other work (for example, ref. 4) has indicated that a certain amount of instability can be tolerated.

w

a T The parameter has been used as a measure of speed-thrust stability.

av

For the aircraft characteristics investigated in reference 5 a speed-thrust level of

a T w

= -0.0012 per knot (unstable) degraded the longitudinal control characteristics

av

sufficiently to be unacceptable for normal operation but acceptable for emergency opera- tion during instrument approaches.

Demonstration of speed-thrust instability.- The time histories of airspeed, flight path, throttle position, and pitch attitude in figure 3-6 illustrate the differences between positive and negative speed-thrust stability. The data are for the basic fixed-geometry

a T w

configuration with the normal unstable value of -0.0024 per knot in one case

av

a T w

(solid lines) and a stable value of 0.0006 per knot in the other case (dash lines).

av

A 100-foot (30.5 m) vertical offset of the glide slope was used as a precision and repeatable task in evaluating the effect of speed-thrust stability variation. For the task illustrated in figure 3-6 the pilot attempted to fly to the glide-slope offset, stabilize, and return without changing the original throttle setting.

The results shown in figure 3-6 for the configuration with stable speed-thrust char- acteristics indicate that large flight-path changes can be made without changing power.

The result of not adding power is a reasonable exchange of altitude and airspeed - that is, the 100-foot (30.5 m) increase shown was attained while the airspeed dropped approxi- mately 10 knots. With positive speed-thrust stability the airspeed would be expected to return to the original value after stabilizing on the new glide slope (at the original rate of descent). For the configuration with unstable speed-thrust characteristics, the air- speed dropped very rapidly following the attempt t o increase the flight-path angle.

ll5L

--

-. M124 / knot

dV Altitude

L

100 ft

t

- d 5- Y ._ a -5- 5 .e -10 I I I I I I 0 10 20 30 40 50 Time after offset, sec Figure 3-6.- Time histories showing glide-slope offset flown with two fixed-geometry supersonic transport configurations illustrating the effects of speed-thrust stability.

Except for one oscillation, the offset flight path could not be maintained and the attitude had to be decreased and power added to keep the airplane from stalling. The examples presented demonstrate the difference between the two values of speed-thrust stability and are not representative of normal operation where the pilot would use the throttle as required and would not be expected to let the airspeed drop for 20 seconds without taking some corrective action.

Effects of speed-thrust stability on supersonic transport configurations.- The speed-thrust stability of the SST configurations of this investigation is indicated by the slope of the thrust-required curves shown in part 2. The fixed-geometry configuration = -0.0024 per simulates unstable speed-thrust characteristics with a value of

av

value of knot, whereas the variable -geometry configuration had a stable

av

a brief investigation w a s conducted with the fixed-geometry 0.0005 per knot. In addition, configuration having a C value which resulted in positive speed-thrust stability

(y = 0.0006 per knot . The results and opinions of this investigation conducted

with experimental test pilots apply only to the conditions of these tests - that is, large

thrust margins, quick engine response, calm air, no emergencies, and no abnormal cockpit distractions.

For the fixed-geometry configuration where all other characteristics were identical

a T w

except the value of CD the change in from -0.0024 per knot to 0.0006 per

Q! av

knot resulted in an improved Cooper pilot rating of 1/2 to 1 rating number based on speed control characteristics during an instrument approach. The results of reference 5 indicate about the same change in pilot rating for less than one-half of this change in the parameter a (T/W).

av

According to the pilot's comments the negative speed-thrust stability

(F = -0.0024 per knot of the fixed-geometry configuration did not have any serious

)

as compared with the effect of positive sta- effect on the instrument approach capability

bility ( * = 0.0006 per knot of the same configuration or with the effect of positive

)

stability (F = 0.0005 per knot of the variable-geometry configuration.

Most of the comments were similar to the thought expressed by one pilot who in

comparing the unstable with the stable speed-thrust characteristics said ". . . larger

variations (in airspeed) occur due to the unstable thrust-velocity relationship and an increased requirement is placed on proper coordination between elevator and throttle .??

Most of the pilot comments relative to speed-thrust stability were based on evalua- tion of the fixed- and the variable-geometry configurations where in addition to speed- thrust stability other parameters were also different. One of the pilots indicated that for the variable-geometry configuration the glide path w a s primarily controlled by elevator and the throttle was used only when the approach was , . definitely high and fast or low and slow. Generally the pilots said that elevator w a s used f o r glide-path control and throttle for airspeed control with one pilot indicating an occa- '/' sional need for a reversal of this combination.

The control techniques used for the fixed-geometry configuration a = -0.0024

(w

per knot tended toward a mixture of techniques - that is, a combined or coordinated use

)

of elevator and throttle to control both airspeed and glide slope. Only two pilots reported a definite use of elevator to control glide path and throttle to control airspeed, and one of these indicated that a lack o f time precluded an evaluation of reverse or other tech- niques. The problem of flight-path control for this airplane w a s summed up very well by one pilot in comparing the fixed-geometry configuration with other airplanes that he had flown which operate on the back side of the thrust-required curve.

He said, "Other airplanes that I have flown on the back side require less attention to speed; you tend to control your speed with your nose attitude and your rate of sink with your throttle, although you can't divorce one from the other. I f you make an input one place, you have to make another to compensate for it. The airplane (fixed-geometry configuration) is, probably down to 300 feet using the flight director, not too difficult to control.

The majority of the problems come below that (altitude). It does not fly as well as other air- planes I have flown on the back side of the thrust required (curve)."

Although no serious problems were encountered during the instrument approaches with the fixed-geometry configuration as a result of speed-thrust instability, several pilots pointed out that the rapid speed or altitude loss during a turn could cause a prob- lem. For example, a rapid turn necessitated for an avoidance maneuver while near the ground at a low speed could result in a serious speed or altitude loss at a time when the pilot has little time to observe or correct for these changes.

It does appear that instrument approaches with a fixed-geometry airplane, such as that simulated during this investigation, could be managed if necessary, provided the a proficiency for instrument approach with this pilots had been trained and maintained type of airplane. Therefore, if some form of automatic speed control is to be used, there should be no necessity for a redundant system to provide for equipment failure.

CONCLUDING -MARKS A flight test investigation of the instrument approach and landing characteristics of simulated fixed- and variable -geometry supersonic transport configurations has indi- cated the following results: The maximum normal load factor used during instrument approaches with either configuration was generally between 1.15 and 1 . 2 with the highest value at 1.3.

The airplane attitude at touchdown w a s generally about 2 ' higher than during the approach for both configurations; however, it should be pointed out that for the fixed- geometry configuration only about one-third of the estimated ground effects were simulated.

Landing-flare paths computed on the basis of a constant load factor of 1.05 give a good approximation to measured flare paths.

The unstable speed-thrust characteristics of the fixed-geometry configuration caused no serious problems for the experimental test pilots during these simulated were conducted in calm air with no emergencies or abnor- instrument approaches which mal cockpit distractions.

For the fixed-geometry configuration, a change in speed-thrust characteristics from unstable to stable resulted in an improved pilot rating of about 1/2 to 1 rating number for the instrument approach task.

f

REFERENCES 1. F’usfeld, Robert D.: A Method of Calculating the Landing Flare Path of an Airplane.

Aeron. Eng. Rev., vol. 10, no. 2, Feb. 1951, pp. 25-30.

2 . Neumark, S . : Problems of Longitudinal Stability Below Minimum Drag Speed, and Theory of Stability Under Constraint. R. & M. No. 2983, Brit. A.R.C., 1957.

3. Lean, D.; and Eaton, R.: The Influence of Drag Characteristics on the Choice of Landing Approach Speeds. C.P. No. 433, Brit. A.R.C., 1959.

4. Staples, K. J . : Flight Measurements of the Influence of Speed Stability on the Landing Approach. AGARD Rep. 420, Jan. 1963.

5. Bray, Richard S.: A Piloted Simulator Study of Longitudinal Handling Qualities of Supersonic Transports in the Landing Maneuver. NASA TN D-2251, 1964.

TABLE 3-1.- PEAK VALUES OF ATTITUDE AND NORMAL LOAD FACTOR FOR INSTRUMENT APPROACHES WITH VARIOUS SUPERSONIC TRANSPORT CONFIGURATIONS ~ Glide slope Simulati Maximum flare Maximum a t ude increment Configuration Load factor attitude during a iroach, deg load factor touchdown Nose up Maximum Minimum Nose down Fixed geometry - 4 . 1 0.90 1 . 0 8 3 . 0 -2.0 1 . 2 0 2 . 0 .89 (basic) - . 7 . 8 2 1 . 1 0 2 . 8 0 . 9 1 -1.7 .78

8 8

- Fixed geometry - 1 . 5 1 . 1 6 0.90 (a) (a) -2.0 1 . 2 0 2 . 3 1 . 1 3 . 8 6 - 2 . 1 1 . 1 3 -84 1 . 1 8 2 . 6 -1.5 1 . 1 8 .77 1 . 3 2 3 . 0 -1.5 1 . 2 4 1 . 1 6 2 . 5 .78 Fixed geometry 2 . 5 -1.3 1.20 0 . 8 2 1 . 1 6 3 . 2 2 . 7 -3.7 1.27 .75 1 . 1 6 2 . 5 ((6 + A d augmented) 1 . 1 6 1 . 1 6 2 . 5 1 . 3 -2.0 .62 with degraded Dutch 1 . 7 -2.7 1.17 .83 (a) (a) roll damping and Cn i Variable geometry 3 . 5 - 2 . 5 1.14 0 . 8 6 1 . 5 -2.5 1.20 .72 (basic) . 7 - 3 . 1 1.16 .a2 4 . 3 -2.6 1 . 1 8 . 8 4 3 . 5 -3.2 1 . 3 0 3 2 -2.0 4 . 4 1 . 2 0 .79 Variable geometry 6 . 5 -4.0 1.16 0.84 1 . 7 -3.1 1.14 .91 (i and 6 augmented) 1.14 2 . 5 - 2 . 3 -89 2 . 0 - . 7 1.09 .89 2 . 2 -1.0 1 . 1 6 .87 - 1 . 5 1 . 3 0 2 . 3 .81 Variable geometry 2 . 0 -1.1 1 . 0 7 0 . 9 4 3 . 8 2 . 0 -1.1 1 . 0 7 3 . 4 . 9 1 (i and ( b + Am) 1 . 5 -1.2 1 . 1 2 .91 3 . 0 augmented) 2 . 5 -1.7 1 . 1 7 .eo 3 . 3 1 . 2 0 3 . 3 2 . 1 -2.3 .85 2 . 5 -1.5 1.23 .eo 3 . 0 Variable geometry -2.2 0 . 7 8 - 1 . 3 .89 ( b augmented) -2.3 .88 with aft c.g. -2.5 .80 -1.5 .88 3.0 -2.0 .90 3 . 2 -2.0 . 9 1 3 . 0 I 1 . 2 2 Variable geometry 2 . 1 -2.0 1 . 2 2 0 . 7 8 1 . 8 1 . 5 -2.0 1 . 2 2 3 2 1.15 2 . 0

(s a n d ( 6 + A a )

. 7 -2.0 1 . 1 2 .81 1 . 2 4 2 . 5 augmented) with 1 . 3 -2.2 1 . 0 8 .90 1 . 1 3 3 . 6 . 1 - 2 . 1 1 . 0 7 .89 1.11 2 . 8 aft c.g.

.5 - 2 . 9 1 . 1 2 .84 1.11 3 . 0 Variable geometry with 3 . 1 -2.7 1 . 2 2 0.72 1 . 2 4 3 . 5 3 . 0 -4.0 1 . 3 0 .76 (a) (a) degraded Dutch roll 1 . 2 2 1 . 2 0 2 . 5 2 . 5 -4.0 .72 damping and Cn Variable geometry 3 . 5 -1.0 1 . 2 8 0.82 1 . 1 6 4 . 5 2 . 7 - 1 . 5 1 . 1 4 .82 1 . 1 1 3 . 2 ((6 + A d augmented) 2 . 0 -1.2 1 . 1 4 .84 1 . 1 6 3 . 0 with degraded Dutch 1 . 1 - 2 . 1 1 . 1 7 . 8 5 1 . 1 3 3 . 2 1 . 5 - 2 . 5 1 . 1 7 .87 1 . 1 4 3 . 0 roll damping and Cn 1 . 0 -2.0 1 . 1 7 .80 (a) (a) i %o touchdown.

4. LONGITUDINAL HANDLING QUALITIES By William D. Grantham and Lee H. Person SUMMARY An in-flight simulation study has been made to determine the handling qualities of several supersonic transport configurations during the landing approach. This report discusses the longitudinal portion of the study. The longitudinal handling qualities of the variable-geometry and the fixed-geometry SST configurations were considered "atis*- / .._ tory because of the sluggish initial pitch response and the apparent low damping. The use of stability augmentqon and an increase in control gearing made the longitudinal charac- teristics of both configurations satisfactory.

INTRODUCTION This report presents the results of the longitudinal portion of the study which was undertaken to determine (1) the handling qualities of several SST configurations during the landing approach, and (2) possible criteria for low-speed handling qualities which would be applicable for the establishment of certification requirements for the SST.

are listed in part 1 of this publica- The basic configurations ,flown during the study tion and in general were a variable-geometry configuration with the wings in the forward (A = 20°), a variable-geometry configuration with the wings in the fully swept position position (A = '72O), and a fixed-geometry (delta wing) configuration. The pilot evaluation 2 of this publication; in general, these procedures procedures a r e discussed in part included evaluation of the aircraft (a) at altitude, (b) during a visual approach and landing, and (c) during several instrument approaches down to an altitude of approximately 200 feet from which the flare and landing was performed visually.

RESULTS AND DISCUSSION For the most part, the longitudinal characteristics of the various SST configurations simulated are presented and discussed in relation to pilot ratings and opinions. The individual Cooper pilot ratings and comments for each test condition are presented as All configurations were evaluated by a minimum of two pilots, with some perti- table 4-1.

nent configurations being evaluated by seven pilots; however, the average pilot rating presented throughout the discussion was taken as a n average of the two pilots that flew all configurations (pilots A and B).

Variable-Geometry Configuration Basic.- The average pilot rating of the longitudinal handling characteristics of the variable-geometry SST configuration was 4.1, with the objection being the sluggish pitch response and apparent low damping as evidenced by some overshoot in pitch attitude changes. The pilots reported that a large portion of the total effort was used to control the glide path and airspeed on ILS approaches.

Static longitudinal stability : The static longitudinal stability of this variable -

geometry configuration was considered by the pilots to be adequate. Plots showing the stick-fixed ( 6, against Ve) and stick-free (Fc against Ve) static stability a r e presented as figure 4-1. A s can be seen, the stick-fixed stability is approximately -0.099 deg/knot, and the stick-free stability is approximately -0.380 lbf/knot (1.69 N/knot).

This configuration was flown on the stable side (front side) of the thrust-required curve. The variation of thrust required with velocity a- aV was approximately +0.0005

(7 )

f o r this variable-geometry configuration and is discussed in detail in part 3 of this publication.

Dynamic longitudinal stability : The short period undamped natural frequency W n

< of this configuration a r e indicated in figure 4-2(a) and are com-

and damping ratio pared with some subsonic jet transports. A s can be seen, the damping ratio of this SST configuration compares favorably with the damping ratios of the indicated subsonic trans- ports, whereas the undamped frequency of this SST is lower than tho& indicated. for the various subsonic transports. Pilots a r e not aware of the magnitude of wn, however, but instead see the damped natural frequency (the damped period of the short period oscilla- tion). See figure 4-2(b) for a plot of damped frequency W D against 5 . The relative difference between the SST and the subsonic jet transports is more pronounced for damped frequency conditions than for the undamped frequency conditions. (It should be mentioned that the indicated short period characteristics of the subsonic transports a r e normally considered acceptable by pilots.) A s stated previously, the pilots objected to the dynamic longitudinal stability characteristics of this variable geometry SST configuration because of the sluggish initial pitch response and the apparent low damping. These pilot comments can best be explained by examining figure 4-3, which shows the aircraft pitch rate response to a n elevator pulse. This figure indicates a pitch rate time constant of approxi- mately 1.6 seconds, and also shows that the pitch rate continues to increase even during the release of the control. The long pitch rate time constant would appear to the pilot as sluggish initial response, and the integral of the pitch rate following control release I (shaded area) would appear to the pilot as an overshoot in pitch attitude (low pitch damping). These two characteristics of the short period dynamics forced the pilot to anticipate and to check the pitch motion during maneuvers to avoid overshooting the desired pitch attitude. During instrument approaches, precise attitude control was diffi- 'cult, thus the pilots tended to oscillate about the glide path - hunting for the desired glide path and airspeed.

Maneuvering characteristics : Longitudinal maneuverability was considered to be adequate for any normal situation encountered during the approaches. The longitudinal maneuvering stability in a wind-up turn, is shown in figure 4-4 as column deflection 6, and stick force F , as a function of normal acceleration n. (The value of 6c/n is approximately 0.90 deg/g, and Fc/n is approximately 33 lbf/g (147N/g); both were con- sidered by the pilots to be adequate.) It should be noted that the simulation was limited to approximately 1.3g and that the data at higher accelerations a r e not reliable.

Control: The pilots commented that the initial pitch response to column inputs was sluggish. This sluggish response, which was caused by the high pitch inertia, is illus- trated in figure 4-5. When compared with a present subsonic jet transport, the initial SST response is rather sluggish; however, the steady state response is considerably better than the subsonic transport. This figure also shows that, because of the sluggish pitch response of the SST, a longer time w a s required for small glide path changes, which of course made it difficult for the pilot to make quick and precise glide path corrections.

initial pitch In order to fly the airplane on an ILS approach, the pilot had to quicken the response to a more acceptable level by supplying a forcing function. This procedure involved the use of an increased initial input of the column, followed by a reverse input, in order to avoid overshooting the desired pitch attitude. (An illustrative example is shown as fig. 4-6.)

The pilots found the control and trim activity required to establish and hold a .

desired rate of descent and airspeed to be quite high during ILS approaches. This trim difficulty was due, in part, to the sluggish pitch response and the apparent low damping (low frequency short period).

The trim change with thrust was in the normal direction experienced with large subsonic jet transports (nose-up with increased thrust). Furthermore, the trim change

was small - which the pilots stated helped in stabilizing on the glide slope. The thrust

response was considered to be excellent. The pilots controlled the speed mainly with the throttle, and the speed control was considered adequate. Some of the possible factors contributing to good speed control were (1) excellent thrust response, (2) smooth air, and (3) the use of a sensitive airspeed indicator.

Landing characteristics : The flare characteristics were poor. As stated previ- ously, this variable-geometry configuration had sluggish pitch response and apparent low 1 1 . I, 1.11..

. . .

damping; these characteristics caused control problems during the instrument approach, but were most evident during the landing flare where the pilot was trying to arrive at a reasonably precise touchdown point with a reduced rate of descent and at a proper landing attitude. Most of the pilots felt that there was a tendency to overcontrol during the flare and occasionally rather severe cases of low frequency control pumping occurred. The flare time history of figure 4-7(a) is one example; this landing occurred on a clear calm day in the early stages of the flight test program. Even though pilot training may elimi- nate this type of oscillation, poor conditions such as turbulence and/or low visibility might produce dangerous situations during landings.

The control forces required for flare were considered acceptable in that a maxi- mum force of only about 10 lbf (44 N) was used.

The ground effects presented no problem insofar as the incremental pitching moment experienced. However, during the flare, the aircraft tended to llfloatlldown the runway.

(The pilots commented that this floating tendency seemed unrealistic when compared with present-day subsonic jet transports.)

Pitch rate e' augmentation.- The first longitudinal stability augmentation system

(hereinafter referred to as SAS) evaluated during the flight tests of the variable-geometry was a pitch rate damper which produced the frequency and damping characteris- concept tics shown in figure 4-8. (As can be seen, the frequency was increased approximately

50 percent, and < was increased from 0.672 to 0.940.) In addition to the SAS, the

elevator to column gearing was increased from -1.3 to -2.6. (This increase in elevator to column gearing was made in an effort to maintain the same Fc/g, and this change in con-

trol gearing appeared to the pilot as a reduction in speed stability 6c/Ve.) This 6 aug-

mentation was better than the unaugmented, but it still had several somewhat undesirable features. The average Cooper pilot rating of the longitudinal handling qualities with the e' SAS was 3.4, the objections being the still less than good pitch response and the deterio- ration of speed control.

Pitch rate plus alpha (e' + Aa) augmentation.- The second and most satisfactory Ion-

gitudinal SAS used during the flight tests of the variable-geometry concept was a pitch rate

plus angle of attack (6 + Aa) feedback system. (The elevator to column gearing was

increased to -5.2.) The average Cooper rating of this configuration was 2.5, compared with 4.1 for the unaugmented configuration and 3.4 for the 4 augmented configuration.

The effect of the (i + Aa) SAS on W n and 5 is also shown in figure 4-8 for comparison

with the unaugmented and the d augmented configurations.

Static longitudinal stability: The stick-fixed and stick-free static stability were ade- quate and very similar to that of the unaugmented configuration. Figure 4-9 presents plots

of GC and Fc against Ve for the basic and the (e' + Aa) augmented configurations. It

should be noted that the increase of the elevator to column gearing did not appear to the

pilot as a deterioration in speed stability when the (8 + Aa) augmentation was used

C m d brought about by the Aa o r static stabil- because of the increase in the effective ity SAS.

Dynamic longitudinal stability: The pitch damping was said to be very good; small o r large pitch attitude changes could be made without overshooting the desired angle.

was approximately the same as that for the unaugmented Actually, the damping ratio

configuration; however, the short period frequency was increased approximately 100 per -

cent, which appeared to the pilot as an increase in pitch damping. The damping parameter 2 < ~ n was approximately equal to 2.46 for this augmented configuration compared with 1.19 for the unaugmented configuration.

Maneuvering characteristics : The maneuvering capability w a s quite adequate; see figure 4-10 for plots of 6, and Fc against n (&/n = 11 deg/g and F& = 45 lbf/g (200 N/g)).

Control: All of the pilots agreed that the ( 6 + Aa) SAS appreciably improved the

fact, felt that the pitch response pitch response over the unaugmented configuration and, in was now quite satisfactory. An illustrative example of the response of an airplane with

the (6 + A o r ) SAS is compared with an unaugmented airplane in figure 4-11. This SAS

works thusly: The increased control gearing 6e/6c initially causes an increased pitch i / rate and angle of attack response, but as both pitch rate and angle of attack build up, the SAS, which is sensitive to both of these, washes out the increased elevator deflection. In this way the initial response is considerably improved without making the already ade- quate steady-state response overly sensitive. The pitch control sensitivity was generally thought to be good; however, a few pilots felt that it was possibly higher than desirable.

This configuration was very easy to trim and/or establish a desired rate of descent when flying the glide slope. The pilots commented that it was easy to change the rate of descent by approximately *lo0 ft/min (*0.5 m/sec), and then to stabilize again at the ori- ginal rate of descent.

Landing characteristics: The flare characteristics of the (i + Aor) augmented

were quite satisfactory. (See figs. 4-7(a) and 4-7(b) for variable-geometry configuration

comparison of flare time histories between the unaugmented and the (4 + ha) augmented

configurations.) The pilots commented that the pitch response was good, that the attitude was precise, and that no tendency to oscillate in pitch occurred during the flare control when the (6 + Aa) SAS was engaged.

The ground effects were of no consequence except for the previously mentioned floating tendency, which affected touchdown accuracy. It should be mentioned, however, effects of ground could not be changed by longitudinal augmentation, it that although the appeared that the more precise control characteristics of the augmented configuration made it easier to overcome the floating tendency.

Aft center ___ of gravity - . . - (no -. augmentation).- - _ _ -____ For the present in-flight SST simulation program, the center of gravity has thus f a r been held constant at 46 percent C for the a static variable-geometry configuration. (This center of gravity location corresponds to margin of 9.75 percent E.) However, it w a s believed that it would be desirable to deter- of gravity position be altered mine what effects might be expected should the center appreciably. For the variable-geometry concept, the center of gravity location was moved aft to 52.75 percent C (static margin, 3 percent E ) .

It should be mentioned that when the center of gravity was varied, the -elevator to column gearing was not changed from the basic value and, therefore, Fc/g varied and may have had some effect on the pilot's evaluation.

axis of this aft center of gravity The average Cooper pilot rating of the longitudinal configuration was 5.1, the major objections being the low level of static stability and the low pitch damping.

Static longitudinal stability: The stick-fixed and stick-free static stability were low. (See fig. 4-12 for plots of 6, and Fc against Ve for the basic and the aft cen- ter of gravity locations.) This configuration was said to be very difficult to trim; it was easily excited in pitch and had a very slow or almost nonexisting tendency to return to the t r i m condition. One pilot (pilot A) made the following general comment: " 1 think that as a single-axis airplane you could handle it, but if you had anything else to do other than spend your time on the longitudinal axis, it would be very difficult to fly and probably unsafe."

Dynamic longitudinal stability: The damping in pitch was low, and the need for the pilot to supply the necessary damping made maneuvers, such as the flare, very difficult.

2 < ~ n is approximately the same for this aft center of gravity The damping parameter configuration as it was for the basic configuration ( 2 c ~ n = 1.21 and 1.19, respectively).

During the landing approach the airplane seemed to oscillate in both attitude and airspeed around the desired t r i m point.

Control: The pitch response was sluggish, however, once the pitching motion was started, the pitch rate and pitch rate per degree of column were very good. The control sensitivity, in t e r m s of angular velocity, seemed adequate but was actually masked by the slow initial response. The low level of static stability, the low pitch damping, and the large pitch trim change with thrust (resulting from the low static stability) all combined to make this configuration very difficult to trim.

Landing characteristics: The sluggish pitch response and the need for the pilot to supply the pitch damping made a precise flare to proper touchdown attitude quite diffi- cult - even in calm air.

Aft center of gravity; 4 + A a augmentation.- The variable geometry aft center of

gravity configuration (c.g. = 52.75 percent C; static margin, 3 percent C) was briefly flown

with the same SAS as was used for the basic variable-geometry configuration (4 + Aa) to

see if the previously used augmentation would also make a significant improvement in the longitudinal flying qualities of the aft center of gravity configuration. (The elevator t o column gearing was -5.2.) The average Cooper pilot rating of the longitudinal axis of this configuration was 3.0. The pilots commented that this configuration was not difficult t o trim, had a feeling of almost immediate pitch response, and had good pitch damping.

Because of the good response and damping, the flare capability was good with very little tendency to overshoot the desired attitude. Also, the touchdown accuracy w a s con- sidered to be good.

Fixed- Geometry Configuration Basic.- The average Cooper pilot rating of the longitudinal handling qualities of this fixed-geometry SST concept was 5.4. The objections to this configuration were sluggish pitch response, apparent low damping, and difficult airspeed control. The pilots stated that they spent over 50 percent of their time controlling the longitudinal axis of this configuration.

Static longitudinal stability : The pilots felt that the static stick-fixed and stick-free Sc/AV = -0.042 deg/knot and longitudinal stability were low; at an airspeed of 135 knots, Fc/AV = -0.168 lbf/knot (0.75 N/knot). (See fig. 4-13 for plots of €jC and Fc against Ve .) It should be mentioned that this low level of static stability (static margin, 3 per- cent E ) may very well be characteristic of some SST aircraft since higher levels of static stability may tend to compromise the cruise performance by increasing the supersonic trim drag.

This fixed-geometry configuration w a s more difficult to trim than the variable- geometry concept. It should be noted, however, that the variable -geometry configuration This trim was also difficult to trim when the center of gravity was moved rearward.

difficulty seemed to be associated with the apparent low damping and the lack of static stability and left the pilot hunting for correct attitude and airspeed.

This configuration had negative speed-thrust stability a- T/av = -0.0024, which also

W made it somewhat difficult to hold any given airspeed. This factor is discussed in detail in part 3 of this publication.

Dynamic longitudinal stability: The apparent low damping caused the pilots to have a tendency to overshoot any small pitch attitude changes. The reason for this apparent low damping is similar to that encountered with the variable-geometry SST configuration

and is discussed under that section. The short period damping ratio < was approxi-

mately 0.87, and the damping parameter 2<wn 1.3. (See fig. 4-2.)

In addition to the apparent low damping (the damping appeared low because of the low frequency of the short period mode), the initial pitch response was said by the pilots to be poor. The pilots had to overcontrol in order to obtain the desired response to a control input. An illustrative example was discussed earlier in regard to the response characteristics of the variable-geometry configuration. (See fig. 4-6.)

Maneuvering characteristics : The pilots felt that the longitudinal maneuver capa- bility was generally sufficient. Plots of 6, and Fc against load factor n a r e pre- 4-14. As can be seen, z z 11 deg/g and Fc/n = 50 lbf/g (220 N/g).

sented in figure Control: The initial pitch response, illustrated in figure 4-5, was considered to be sluggish by the pilots although they did think it may have been somewhat better than that of the variable-geometry configuration.

The sensitivity of the pitch control 6/tjC was poor. The maximum control power ..

emax was considered to be adequate (better than some present-day subsonic jet

transports).

For straight and level flight as well as for an established rate of descent, constant attention was required to maintain the desired attitude and airspeed. It was obvious to the pilot that this configuration had speed-thrust instability.

The thrust response of the fixed-geometry configuration was considered to be excel- lent. The trim change with thrust was in the normal direction and was quite mild.

Landing characteristics: Because of the sluggish pitch response, the flare charac- teristics were quite similar to those of the variable-geometry configuration, that is, poor.

The incremental pitching moment due to ground effects was bothersome to the pilots; a large nose-down pitching moment was experienced at an altitude of approximately 30 feet (9 m). ACm used to simulate the ground effect for this configu- (The values of ration are presented in part 2 of this publication.) However, it was stated by the pilots that this ground effect on pitch would not constitute a problem if the pitch response char- acteristics were improved. Since some difficulty was experienced in simulating ground effects on lift and drag for the fixed-geometry configuration (see discussion in part 2), the effects of the ground on lift and drag will not be discussed.

Pitch rate plus alpha augmentation.- . Since the (4 + Aa) augmentation was found to

be the best longitudinal SAS tested for the variable-geometry configuration, and since the longitudinal characteristics of the fixed-geometry configuration were quite similar to

those of the variable geometry, the (6 + Aa) was the only longitudinal SAS tested for the

simulated fixed-geometry SST.

Generally, the longitudinal axis of the (6 + Aa) augmented configuration was said to

be quite good and easy to fly, the only adverse comment being the constant attention 11111111111 1 1 1 1111111111111111111111111 II I I I I I II I I 111 II I I Ill required to control airspeed. The average Cooper pilot rating of this augmented configu- ration was 2.9, compared with 5 . 4 for the unaugmented configuration.

Static longitudinal stability: The stick-fixed and stick-free static stability were improved over the unaugmented configuration and appeared adequate. See figure 15-4 for plots of 6, and Fc against Ve for the augmented and unaugmented configurations.

Dynamic longitudinal stability: The pitch damping was considered to be very good

for the ( 6 + Aa) augmented configuration. Both small and large pitch attitude changes

could be made very easily with essentially no tendency to overshoot the desired angle.

The short period damping parameter 2<wn was 2 . 3 for this configuration compared with 1 . 3 for the unaugmented.

Maneuvering characteristics : The maneuvering capability was adequate and essen- tially unchanged from the unaugmented configuration. Plots of 6, and Fc against load factor n are shown in figure 4-16 comparing these parameters with and without augmentation.

was considerably improved over the unaugmented con- Control: The pitch response figuration and with augmentation was considered to be very good. The control sensitivity was adequate for any maneuver encountered during the approach and landing.

The augmented configuration was still somewhat difficult to trim, but was said to be less difficult than the unaugmented. The speed control was not good since it was obvious to the pilot that the fixed-geometry configuration was operating on the backside of the power required curve 8- aV = -0.0024). The ability to hold a desired trim speed

( G I

w a s better with the (4 + Aa) SAS, not because of any improvement in speed-thrust stability, but because of the ability to make small and precise pitch attitude corrections more easily.

Landing characteristics : The pitch response, damping, and attitude control during the flare were good. The augmentation eliminated the control-induced oscillations and thus improved the touchdown accuracy.

Ground effects produced no significant problems. A nose-down pitching moment was noticeable below 30 to 40 feet (9 to 12 m), but was easily controlled with the column.

Variable-Geometry Emergency Configuration (A = 7 2 ' ) The variable-geometry SST concept was tested briefly in the emergency landing configuration (A = '72'), the sole objective being to see whether a variable-geometry air- plane could be safely flown during the landing approach should the wings become inopera- tive when in the swept (cruise) position. Only one pilot (pilot A) flew this particular configuration and made no attempt to complete the landing approaches to touchdown. The pilot commented that this configuration was easier to trim and had better pitch response than either the basic variable-geometry (A = 20') or the basic fixed-geometry configura- tions. The pitch damping was similar to the other two basic (unaugmented) configurations, however, and was poor. It should be mentioned that this variable- geometry emergency configuration (A = 7 2 ' ) was flown at a simulated airspeed of 182 knots, compared with 135 knots for the fixed- and variable-geometry (A = 20') configurations.

The Cooper pilot rating of the longitudinal handling qualities of this variable- geometry emergency landing configuration (A = 72O) was 4.0, the major objection being the poor pitch damping characteristics. As stated previously, no landing approaches were completed to touchdown; however, the pilot stated that he believed he could have landed this configuration safely. No longitudinal stability augmentation was used on this configuration.

Longitudinal Criteria and Requirements For many years aerodynamicists have striven to establish adequate handling quali- ties criteria. Although various criteria have been developed and used, -it has been neces- sary to alter these periodically, because of the expansion of flight envelopes, the increase of airplane size, and the diversification of operational usage.

An often used longitudinal handling qualities criterion is the short period damping requirement appearing in the military specification of 1959, designated MIL-F-8785. This specification requires that the short period oscillation be damped to 1/10 amplitude in no more than 1 cycle, which is a minimum damping ratio of 0.34. However, this requirement applies only to cases where the short period frequency is greater than 0.167 cps (0.167 Hz) and gives no damping requirement for the lower frequency cases. (See fig. 4-17 for longitudinal short period damping requirements of MIL-F-8785.)

A plot, related to this requirement, of the short period frequency and damping ratio of the various SST configurations simulated during the present flight test program is presented as figure 4-18.

As mentioned previously, MIL-F-8785 gives no short period damping requirement f o r aircraft having short period frequencies as low as those for aircraft the size of an a plot of pilot rating against SST. To illustrate the effect of frequency on pilot opinion, short period undamped natural frequency is presented in figure 4- 19, with the various configurations simulated for the variable-geometry concept indicated. The figure indicates that, as the short period frequency varied from 0.10 to 0.30 cps (0.10 to 0.30 Hz), the ratings of the longitudinal characteristics varied from approximately 5 to 2.

In the past, various longitudinal stability and control requirement criteria have been suggested that involve the frequency and damping ratio of the longitudinal short period.

Several of these are briefly discussed in the following paragraphs.

Reference 1 used a short period requirement criterion which involved the damping

parameter 2 y ~ nand the short period natural frequency squared wn . This criterion

was developed for aircraft much smaller than the SST, however, and therefore was much too restrictive and is not presented in this report.

Reference 2, which presents the results of an extensive flight test program that was conducted to obtain data on the optimum and minimum acceptable longitudinal stability and control characteristics for fighter and bomber airplanes during cruise flight, also Figure 4-20 shows this developed a criterion for the longitudinal response and damping.

criterion as a plot of short period frequency fn against damping ratio c. Ratings for

the SST configurations simulated during the present study, as well as those for some present-day subsonic jet transports, are located in this figure. Although the pilot ratings for some of these configurations were satisfactory or acceptable, all of the configurations would be interpreted as being unacceptable on the basis of the criterion of reference 2, which, as mentioned previously, w a s developed for cruise flight conditions.

It has been proposed that the plot of fn against 5, as shown in figure 4-21, be used as a longitudinal requirement criterion. (Note that these boundaries a r e similar to those of ref. 2, presented in fig. 4-20.) Some subsonic transports and the simulated SST configurations are also located on this chart. In regard to this longitudinal requirements criterion, reference 3 stated the need to modify the boundaries for better agreement with flight test results. Figure 4-22 presents an estimate of the type of boundaries that might be drawn to indicate an area of acceptable longitudinal short period dynamics for low- speed operation of large aircraft similar to an SST. (The scales have been omitted from at this graph since the knowledge required to establish definite boundaries does not exist the present time.) This estimated boundary was presented and discussed in reference 3 and agrees with the results of the present SST simulation program in that it proceeds in the proper direction.

Reference 4 stated that factors other than wn and < should be considered when

attempting to establish longitudinal handling qualities criteria and pointed out one that is very significant, that is, the ability to change flight path with normal acceleration, which is related to By using this parameter and by recognizing that the pilot's mode of La.

control is not constant for all flight regimes, two criteria for satisfactory short period characteristics were developed that correlate well with current airplane experience, as well as with various simulation experiments. All of the configurations studied in this program fall in the class for which the criterion recommended in reference 4 was

developed and is expressed as a plot of La/Un against c. (It should be noted that the

W L ,

~, where C is measured per definition of La, as used in this case, is La = -

La mV radian and V is in ft/sec (m/sec).) This criterion is presented in figure 4-23, and several of the configurations studied during the present in-flight SST simulation program and some subsonic jet transports are located. Upon comparing the location of these SST configurations with the Cooper pilot ratings of the longitudinal handling qualities, pre- sented in table 4-2, it can be seen that this short period requirements criterion agrees with the results of the present SST simulation study. It should be noted, however, that because of the limited number of configurations flown during the present study, much is needed before it can be said that this criterion, or any other discussed in more work this report, can be said to be an adequate longitudinal stability and control requirement criterion.

CONCLUDING REMARKS Based on the results obtained during the in-flight simulation program the following remarks a r e made summarizing the longitudinal characteristics of the various configura- tions tested.

Variable Geometry The dynamic stability of the variable-geometry configuration was considered to be poor because of the low frequency of the longitudinal short period, which made the pitch damping appear low to the pilots. Although the damping ratio was quite good, the long period of the oscillation made the damping parameter 2 C ~ ntoo low. This low frequency oscillation made precise pitch control difficult during instrument approaches and also resulted in poor flare characteristics in that it caused the pilots to induce pitch oscilla- tions when trying to position the airplane for landing. The initial pitch response was sluggish which made it difficult to make quick and precise glide path corrections. The sluggish response also contributed to the previously mentioned poor flare characteristics.

Because of the apparent low damping and the sluggish initial pitch response, the longitudi- nal handling qualities of this variable-geometry configuration were considered unsatis- factory (average Cooper pilot rating of 4.1).

The use of stability augmentation, and an increase in control gearing made the lon- gitudinal characteristics of this variable- geometry configuration quite satisfactory (Cooper pilot rating of 2.5). The augmentation used was a combination of pitch rate and angle of attack which increased the frequency of the longitudinal short period and appeared to the pilot as improved pitch damping. The increased control gearing increased the initial pitch response and as the pitch rate and angle of attack built up, the augmentation system washed out the effects of increased elevator gearing. The use of stability aug- mentation also eliminated the tendency toward control-induced oscillations during the flare.

landing I I 1111II I 111 I I In tests where the static margin was changed from 9.75 percent F to 3 percent E, the flight characteristics became worse and the pilot ratings changed from 4 . 1 to 5 . 1 for the unaugmented condition and from 2 . 5 to 3 . 0 for the augmented condition.

Fixed Geometry The basic fixed-geometry configuration generally had the same low frequency of the longitudinal short period and sluggish initial pitch response problems as that discussed for the unaugmented variable-geometry configuration. The longitudinal flight character- istics of this configuration were also considered to be unsatisfactory (average Cooper pilot rating of 5 . 4 ) . During instrument approaches with this fixed-geometry configura- tion, speed thrust instability resulted in an excessive number of throttle adjustments to maintain airspeed. In the flare, the incremental nose-down pitching moments caused by ground effects were somewhat bothersome to the pilot.

The same augmentation system that was used on the variable-geometry configura- tion also made the longitudinal flight characteristics of the fixed-geometry configuration satisfactory (average Cooper pilot rating of 2.9).

Handling Qualities Criteria Several longitudinal handling qualities criteria, which have been used in the past and involve only short period frequency and damping ratio, were found inadequate t o pre- dict the pilot ratings obtained in this program. One previously published criterion which involves short period frequency, damping ratio, and an effective flight path response parameter agreed well with the results of the present investigation.

REFERENCES 1. Harper, Robert P., Jr. : Flight Evaluation of Various Longitudinal Handling Qualities in a Variable-Stability Jet Fighter. WADC Tech. Rept. 55-299, U.S. Air Force, July 1955.

2 . Chalk, Charles R. : Additional Flight Evaluations of Various Longitudinal Handling Qualities in a Variable-Stability Jet Fighter.

WADC Tech. Rept. 57-719, Part II, U.S. Air Force, July 1958.

3. Kehrer, William T . : Longitudinal Stability and Control of Large Supersonic Aircraft at Low Speeds. AIAA Paper No. 64-586, Aug. 1964.

4. Shomber, H. A.; and Gertsen, W. M.: Longitudinal Handling Qualities Criteria: An Evaluation. AIAA Paper No. 65-780, Nov. 1965.

TABLE 4-1.- PlLOT OPINION OF THE LONGITUDINAL HANDLlNG QUALlTIES OF THE VARIOUS SST CONFIGURATIONS SIMULATED Pilot rating of Control Configuration Pilot comments pilot longitudinal parameters bandling qualities Basic variable geometry A 1. pitch response is sluggish; there i s a 1 to 11 second lag in a response once the column 646. = -1.3 4.50 is moved.

(static margin = 9.75 percent E) ae/i = 0 2. Pitch control eensitivity is adequate.

6,/Aa = 0 3. A long time is required to trim for hands-off c o d t i o n .

4. Trim speed band o f 2 to 3 knots. Speed control in descent i s very good.

5. Thrust control response is very good. Trim changes with p e r a r e very light.

6. Phugoid oscillation was apparent with some small ~ a r i a t i o ~ in approach speed. Had to hunt glide path.

7. Maneuverability is adequate for any normal situation encountered during normal approaches.

8. Glide path control: When glide path control i s thrown in on top of the rate of descent and airspeed control the precise 0, a, and V must be sought to give the right glide path.

all the way down.

Seemed to be hunting 9. Attitude comrol on touchdown leaves a little bit to be desired.

10. Work level in the approach is very high an glide path.

11. The major reasons for downgrading this configuration a r e the sluggish pitch response, low damping, and the workload required on the glide path.

B 3.50 - 3.75 1 Basic variable geometry 1. The pitch response could be better but as long a s the pilot doesn't mind moving the column large amounts it is adequate.

(static margin = 9.75 percent E ) 2. The airplane is relatively hard to trim, but once trim is acquired it stays for a long period of time. It will fly 2 to 3 h o t s above or below trim speed. The airplane definitely has speed stability. It is not difficult to bold trim speed a8 long as e is kept Constant.

3. The phugoid oscillation is apparent in trying to find the glide slope. There a r e many O S C i l h t i O n S on the glide slope.

4. On small 0 changes, there is a tendency t o overshoot, and damping must be provided by the pilot. The pitch damping is low.

5. The maneuverability seemed t o be fine.

6. The ability to establish a desired rate of descent at altitude was amazingly easy, but more difficult when flying the glide path.

7. The major reasons for not giving this configuration a better rating were the pitch response characteristics and the lack of pitch damping.

Basic variable geometry 6& = -1.3 C 4.0 1. The aircraft is difficult to trim possibly because of the low control power and Sensitivity, low angular velocity damping, and high rate of trim actuator.

(Static margin = 9.75 percent F ) 6 4 = 0 2. The pitch response to large control inputs and pitch control sensitivity a r e adequate, but 6,/Aa = 0 more control sensitivity is preferred.

3. The ability to hold trim speed is acceptable but not good.

4. The short period damping is acceptable but not good; there is a tendency to overshoot when making small attitude changes which requires the pilot to reverse the control inplt in or to damp the pitch motion.

5. The longitudinal maneuverability seems adequate. The longitudinal forces are SatieLaCtorg but lower values of FJdc and FJg might be preferable. The breakout forces are a little high but satisfactory.

6. Glide path control is satisfactory and the ability t o establish a desired rate of descent is m.

7. Thrust control response is good for a jet engine; it is not as good as thrust reverser modulation. Trim change with thrust is in the normal d i r e a i m but a little high.

8. Pitch response and pitch damping during the flare are poor.

9. The touchdown accuracy i s poor because of too much float.

10. The ground effects produced a moderate nose-down pitch but w m readily W e d if the correction was started soon enough. The floating seems unrealistically prevalent.

Ill1 . . . _.

TABLE 4-1.- PIUJT OPINION OF TEE LONGITUDINALHANDLING QUALITIES OF THE VARIOUS SST CONFIGURATIONSSIMULATED - Contirmed Control Wot comments m e t e r s K e / K c = -1.3 3.5 - 4.0 1. It is not particularly difficult to trim at the desired speed.

2. Pitch response to large control inplts is satisfactory. Pitch response to small control (Static margin = 9.75 percent E Ke/i = 0 inplts is adequate for airwork, but marginal when close to the ground.

Ke/Am = 0 3. The damping in pitch was inadequately evaluated, but appeared to he too low. There w a a tendency to overshoot when attempting to stop a pitch motion.

4. The ability to hold desired airspeed was satisfactory. There was only a 2 to 3 hot variation in speed except when larger e r r o r s were plrposely introduced.

5. Some long p e r i d oscillation of airspeed was apparent hut was felt to be largely due to h i & throttle gearing which caused some overcontrolling with thrust initially. The trim was satisfactory, but probably somewhat larger than desirable. The change with thrust speed change with thrust l a g s the attitude change by several seconds.

6. The longitudinal maneuverability was quite adequate.

I. Glide path control was satisfactory. Glide path was controlled primarily with elevator.

Throttle was used only when glide path was definitely high and fast or low and slow.

8. U was not particularly difficult to maintain desired approach speed - only occasional thrust changes were required. The reasons for good speed control were good thrust response, small effect of maneuvering on speed, and relatively small attitude changes required with flight director.

9. The ability to control attitude during the flare is margoal. There a r e definite over- controlling tendencies.

10. No pitching tendencies due to ground effects were noticed, but there were very strong floating tendencies.

I 11. The touchdown accuracy is satisfactory if initial flared attitude is correct. An over- flare results in some extension of touchdown point.

12. The longitudhal characteristics of this configuration were rated 3.5 to 4.0, hut this rating would be 3 . 0 except for the flare problem.

Basic variable geometry 6e/Kc = -1.3 4.25 1. It is not unduly difficult to establish the desired speed within 11 to 2 Imots, but is a hit hard to hold in trim.

(Static m u g i n = 9.75 percent 3 6 4 ; = 0 2. Pitch response to large control inplts is satisfactory but there is some lag and a ae,fAu = 0 slight tendency to overshoot in 9 . Pitch response to small wntrol inplts i s too slow.

3. The pitch damping is fair. There is no excessive tendency to overshoot during atti- tude changes.

4. Trim change with thrust is excessive, but is in the proper direction (increased thrust results in nose-up). The s p e d change with thrust is obscured by attitude change - the speed change does have normal response U the attitude is held coastant.

5. The longitudinal maneuverability characteristics a r e normal.

6. The glide path cantrol is satisfactory. The glide path was controlled with elevator and the airspeed with throttle. There was no problem in m a i n W n g the desired approach speed within +4 to -3 Imots.

I 7. The flare and landing portion of the evaluation clearly shows h g in control of attitude when close to the ground. This l a g in pitch response near the ground forces the pilot to perform a mild ptsh-pdl in the flare. The attitude control k i n g the flare and the touchdown accuracy is not good because of this pitch response delay.

8. There was no apparent pitching or floating tendency due to ground eIfects.

TABLE 4-1.- PILOT OPINION OF TEE LONGITUDINAL HANDLING QUALITIES OF TFIE VARIOUS SST CONFIGURATIONS SlMULATED - Continued pilot rating of control Pild longitudinal Pilot oomments Configuration parameters haodling W t t e s Basic variable geometry 6& = - 1 . 3 F 3 plus 1. It waa very easy to trim a t the desired airspeed; there were some short period Oscillations however.

6=/6 = 0 (static margin = 9.75 percent E) 2. Pitch response to large control inp.ds is hatisiactory; there in some tendency to over- 6,/Am = 0 shnot pitch -Me Md to m0, but only on IFR. No problem on VFR. Pitch response to small control inplts was satisfactory on YFR (pitch time constant not -q%g). It was hard not to P I 0 small amplitudes when on IFR.

3. The phupid damping seemed about neutnl. The short period damped but caused some PI0 when IFR.

4. It waa very easy to hold desired speed and the change in speed with vlrust was very fast; - in fact, the engine W t response seem8 optimistic for an SST.

5. Glide path control r p a ~ satisfactory. The airspeed was controlled with throttle and glide path with elevator excel* when it seemed appropriate to me throttle for glide p t h . Both techniques a r e easily applicable.

6. The desired approach speed can be maintained within t3 lmds with normal attention because o f the excellent thrust response.

7. During the flare, the pitch response, pitch damping, and attitude control were satisfactory - but there wae a slight tendency to set up m0.

8. Ground effects: There was a slight Mae-down pitch, Lmt this may be due to thrust and speed decrease, There w a noticeable tendency to float and this floating tetldency at idle power and constant pitch attitude i s detrimental to the touchdown accuracy. This floating tendency seems unrealistic compared with present-day subsonic jet transports.

6d6, = -1.3 G 4.5 1. The aircraft -fairly difficult to trim.

Basic variable geometry 2. Pitch response to either large or small control inplte is not satisfactory. The airplane (Static margin = 9.75 percent E ) 6 4 6 = 0 responds sluggishly.

64Aa = 0 3. The pitch damping appeared to be satisfactory. No tendency to overshoot TMB noted during attitude changes.

4. The ability to maintain desired speed is poor. It requires careful monitoring of rate of eUmh and pitch attitude. The attention required to control speed i s high because there seems to be no apparent help from natural stability of aircraft.

5. The maneuvering forces are much too light, shmld be about twice 88 heavy per unit 6 or per g.

6. Glide path control is fair. The flight director and the good sensitivity of the airspeed indicator help considerably. The low static sticl-free stability detracts from what could he called good speed control.

7 . Flare: There was a tendency to pump the control which i s a eymptom of a too sluggish response In pitch. The pitch damping seemed adequate. Attitude control is poor, rela- tive to m a g changes, but once a change i s made it does hold attitude fairly well.

8. Some floating tendency was noticed near the ground.

9. The approach is 100 percent work level.

The pitch Variable geometry with = -2.6 A 3.5 1. Pitch response i s quicker than that for the unaugmented configuration.

damping is also better and has eliminated any tendency to overshoot small attitude pitch rate B 6e/6 = 1.46 changes and hae also eliminated any tendency of low frequency plmping of the controls augmentation during flare and landing.

64Aa = 0 (static margin = 9.75 percent E ) 2. The apparent static stability, 88 speed is displaced from trim condition, seems lower than that for the unaugmented variable sweep configuration.

3. The 6 SAS reduces the longitudinal workload from 80 percent to about 60 percent.

If an autospeed control were added to the system, it would reduce the workload of the pilot, on this axis, to probably SO percent.

4. There waa little difference between approaches with and withwt simulated grwnd effects. The 6 SAS has completely eliminated the pitch-down that waB MtiCed for the una-ented configurntion.

__ TABLE 4-1.- PILOT OPINION OF T E LONGITUDINAL HANDLING QUALITIES OF THE VARIOUS SST CONiTGUFlATIONSS D " T E D - Continued Pilot rating of Control Configuration P i l O l longitudinal Pilot comments parameter handling q d t i e variable geometry with 6e/ac = -2. B 3 . 2 5 1 . The 6 SAS has definitely increased the pitch damping and has slightly increased the pitch pitchrate B response. The major benefit is the increased damping which allows better attitude hold.

K e , 6 = 1.46 augmentation (If the airplane has poor pitch response the pilot can learn to live with it, it is just a matter of pilot anticipation and pilot lead time; whereas, poor damping makes the workload much be/Aa = 0 E 9.75 percent E (static margin higher all the time.)

6 SAS giving a good approach attitude hold, some static or speed stability is last; 2 . With the thus, the speed control is noticeably more difficult. The total task is still far easier.

3 . No ground effect on C , o r CD was noticed; bowever there was quite a bit of additional CL. On a wuple of approaches the aircraft descended to within 4 to 5 feet (1.2 to 1.5 m) of the runway and just ~ o r t of floated along.

Variable geometry with 6 4 6 = = - 5 . : A 3 . 0 1. The aircraft is not difficult to trim. It takes from 45 t o 60 seconds to trim it but it can be pitch rate plus trimmed and then it will stay essentially there a t a b u t +2 knots. The aircraft has positive 6,l = 1.46 angle of attack static stability. It feels nice through the trim position - it is fairly linear in the pull and feedback (i + A 4 push forces to slow and speed up the aircraft at about one pound per knot (4.4 N/knot) within 6 4 A u = 1 . 5 augmentation 10 knots on either side of trim speed.

2. Pitch control sensitivity is very high. The pitch rate per degree of column is higher than (static margin = 9.75 percent 5 desirable for landings, although it feels good in the air.

3 . The response to control input is satisfactory. Pitch rate and angle of attack response is very very rapid, occurring probably within a half second after the column input.

4. The pitch damping is very g o d . Small attitude changes are quite easy to make, with no tendency to overshoot and notendencytoward PIO.

5. The maneuverability is very good.

6. The glide path control is satisfactory. About 30 percent of the time was spent on the longi- tudinal anis (20 percent on speed and 10 percent on attitude t o follow the fight director). The ability to establish a desired rate of descent is fairly easy.

I. Thrust response is good. There is a light, but noticeable, trim change with thrust.

8. Flare control: The pitch response is snappy, almost too snappy for good landtngs. The pitch damping is excellent. Attitude control is very precise, but a very light touch is required on the column to prevent overcontrolling.

9. Touchdown accuracy seems very g o d - much better than the two previous variable-geometry configurations.

0. Generally, the longitudinal control is the best in the program to this paint. The response is good, no apparent lag in the pitch rate. The stick gearing and the gains on B and 01 are a little high. The longitudinal characteristics of this configuration a r e rated 3 . 0 , but the system could be optimized to a rating of 2 to 2 . 5 .

Variable geometry with ie/ac = -5.2 B 2.0 1. The aircraft is not difficult to trim. It is probably the easiest trimmable configuration pitch rate plus flown thus f a r in the program and the aircraft stays within 2 to 3 knots of trim which is Se/d = 1.46 angle Of attack about ail that could be expected of an airplane with such large inertia.

feedback (i + A C T ) )=/Am = 1 . 5 2. Pitch response to large or small control inputs is quite satisfactory. The control sensitivity augmentation is good.

:Static margin = 9.75 percent ?)

3 . Pitch damping is at a good level. Small and large attitude changes a r e easy to make without overshooting. There is no tendency toward PIO.

1. In the approach, glide path control is easy. It is easy to establish a desired rate of descent, but more important, it is easy to change it slightly - to take off a 100 ft/min (0.5 m/sec) or to add a hundred ft/min, briefly and then stabilize at the original rate of descent. The thrust control response is good.

5. The pitch response, pitch damping, and attitude control were good in the flare.

j. The 6d6, gearing is at a very good level, but lf the same gearing had to be used at high speeds, without a mechanical advantage change, a lower gearing might be desirable.

1. No ground effects whatsoever were noticed.

TAELF, 4-1.- PILOT OPRUON OF THE LONGITWD!NAL aANDLIN0 QUALITIES OF TEE VARIOUS SST CONFTGURATXONS SIMULATED - Contirmed P i l o t rating of control Configuration mot longlkdbal Mlot eammenta parameters handling @ties Variable geometry with be/& = -5.2 C 4.0 1. T h e aircraft can be trimmed satisfactorily. T h e chlef problem seems to be the high rate pitch rnte plus of the trim aduator.

6efi P 1.46 angle of attack 2. The pitch response to large control inputs is satisfactory and the pltch control sensitivity feedbock (6 + AU) &/Aa = 1.5 is very good.

nvgmentation 3. The pitch damping is good. When maldog small attitude changes, there is no tendency to (state margin = 9.75 percent E ) mrershmt, however, there is a strong terdency to ”spring back” alter the control i n @ is relaxed. This is mlldly undealnble.

1 . The mnnewernMUty is good.

5. On the approach, the glide path codrol is satisfactory nnd the ability to establleb a desired rate of desced is good. The speed control Is good U throttle is fixed. The t h w t control response is good and the trim change with thrust is eatMactory but a little high.

6. h the flare, the pitch response, the pitch damping, nnd the ability to control attitude are good.

1. The touchdown accuracy is poor because of excessive fioptbg.

8. The nose-down trim change due to ground effects is noLLced below 50 feet (15 m) M is of ~1 COMequenCe to the fioal lvdlng - provided correction for it is started promptly.

9 . The strong tendency for speed to decrease with normal use of controla following retardation nf thrdtle for glide path correction is bothersome. It seems strange that this happens con- sidering the prevalllng nose-down trim change with thrust reduction and the fact that the ha term i s In the augmentation.

10. The rating of the longitudinal eharaderlstics of thls configuration is 4.0 because o f the exeesslve floating tendency near the p d t h h rating wauld be 2.5 U not for the floating Variable geometry with 6e/6, = -5.2 D 1. It is not difficult to trim at the desired speed.

3.0 pitch rate plus 2. Pitch response is definitely improved over the unaugmented variable sweep configuration, Ke/ = 1.46 angle of attack but it is not tm sensitive for small corrections - and B control gearing a , / & , change feedbeck (6 + Am) wmld be required for pitch control. There la a tendency to overshwt during attitude KdAa = 1 . 5 augmentation changes became Of this high control sensitivity.

m a r 0 = 9.75 percent E) 3. It is relatively easy to hold desired speed within i.3 imots. Thia augmented configuration (statio is harder to control than the unaupnented configuration, however, because of the increased sensitivity io pitch control.

4. The trim change and speed change with thrust a r e the same as that for the wugmented configuration.

5. The longitudinal maneuverability is quite adequate.

6. The glide path control ~8 satisfactory a d there was no problem in maintaining the deslred approach speed. The increased pitch sensitivity was bothersome, however.

7. The attitude control during tlie landing flare is unsatisfactory but acceptable; the r e s p n s e is very good but there i s a tendency toward overcontrolling.

8. There waa a Bevere floating tendency due to ground effects if the aircraft was over- rotated duping the flare, but it was satisfactory U proper flare is executed and attitude is held for slight slnk rate.

9 . The touchdown accuracy was satisfactory except for the floating tendency. This Ls a technique problem rather than a control problem, however.

10. The longttudlnal characteristics of this augmented configurption were rated 3.0 I n spite of too high column se~itivity - as this would appear to be easily optimized.

Vaiinble geometry with 6 & = -5.2 E 3.25 1. It is eany to trim at the desired epeed in smooth air.

pitch rate Plus 2. Pitch response to either large or small control inplts is eatisfactory.

K e F = 1.46 angle of attnck 3. W c h damping is adequate, there is no tendency to overshoot duriog attitude changes.

feedbnck (6 +Ad 6e,Aa- 1.6 augmentation 4. The h” change with Ulrvst is almost discernible and the s p e d change with thrust is normal.

(static mnrgin = 9.75 percent E ) 5. Glide p t h control is satisfactory. It is quite easy to mplntain the desired approach speed within r 2 . 5 Imats, and the major reasw for good speed control is the good r e a p n s e to pitch attitude commnnis (control is predictable d repatable).

6. The pitch response and ability to control attihrde dvrlog the flare are good. T h e pitch dpmplng during the flare is adequate.

7. Very llttle pit- due to grouad effects wan noticed, but there deflnitely was a tendency t o float.

8. The t e a c h d m accuracy for thls augmented configuration is mme consistent tbm it was far the umugmented.

TABLE 4-1.- PILOT OPINION OF TEE UINCITIDINAL HANDLING QUALITIES OF THE VARIOUS SST C 0 " S SIMULATED - Continued C o b o l Pilot comments pPnmaers .. . _ _ _ . . .. - -~ - ._ -. ... -. .. ~ 1 2.6 bePC -5.2 1. The aircraft is mt difficult to trim but the phugold oscillation is mticeable.

2. The pitch response to either large o r small control inpts is eatisfactory and is noticeably 6 4 = 1.46 better thyl the unaugmented varinble sweep configmation.

aepa = 1.5 3. It Is very easy to hold the desired airspeed and there is na noticeable oscillation in speed.

4. The speed change with thrust seems to be quite fast. The trim change with thrust is in mrmpl direction.

5. Glide ppth control is satisfactory. Glide path was eontrolled e t h the elevator ami the airspeed with throttle - however, this technique could easily be reversed.

6. The desired speed ean be mnintained within i 3 knots with normal attention The rea80118 for good speed control are excellent thrust response, smooth air, and having a precise aspeed Micntor.

7. The pitch response and pitch &ping in the flare were very gcod. The pbugoid oscUlation was mticeable when trying to control the attitude, but this was no problem in smooth air.

8. Ground effects: There was a Mticeable tendency to f l a t , but no serious nose-down pitch.

This floating tendency is detrimental to the touchdown accuracy. This f l o a t i i tendency oceurring even at idle thrust is hard to believe.

G 4.0 = -5.2 1. The ability to trim t h i s configuration was the same a8 tbat for the unaugmented variable Variable geometry with sweep - fairly difficult.

pitch rate plus ae/ E 1.46 angle of attack 2. The pitch response is s t i l l sluggish but not aa sluggish as the unaugmented coofiguration.

feedbpclc (i + A d 6@a= 1.5 3. There was no tendency to overshoot during attitude changes and no tendency toward P I O ' s .

augmentation 4. Longitudinal maneuverability was the same aa that for the unaugmented configuration, but (static mar& = 9.75 percent E ) the glide paul control was a little bit easler.

5. In the flare, the pitch damping appeared to be good but the pitch response was sluggish which made it hard to change attitude precisely.

6. The floating tendency near the ground was noticeable and the touchdown accuracy was poor.

7. The ease with which the airplane ean be controlled on aFprcach and during the flare was similar in both augmented and unaugmented and it is questionable whether the dlfference, could be determined in anything but extremely smooth air. The longitudinal augmentation is f a r from o p t i " .

ae/ac = - 1 1. The aircraft is very difficult to trim longitudinally - it is easily excited in pitch and has Bnsic variable geometry 5.5 I * a very slow o r almost nonexisting tendency to return to the trimmed condition.

@ = O (static margin = 3 percent E ) 6, is in and the response 2. Pitch response to column innpas is very sluggish; however, once takes hold, the pitch rate and pitch rate p r degree of 6, seems adequate. However, the ae/Aa = 0 sensitivity is really maaked by the low response.

3. The pbugoid is apparent and appeared to be neutrally damped. The airplane seems to oscillate in both attitude and &speed around the desired t r i m point.

4. The static stability i s very light and is nonlinear tbrough the trim point.

5. Any pitch rate damping haa to be supplied by the pilot to prevent overshoot o f pitch attitude.

6. No tendency toward PI0 was mticed.

7. Glide path control is satisfactory, brt it requires about 80 p r c e n t of the pilot's attention, and 30 percent to control the a i r s p e d , which abmt M percent to control attitude with bC is done primarily w i t h throttle.

8. Thrust response Is good. The trim cbange with tbrust i s noticeable.

9. The slugglsb pitch response and need for pilot-supplied damping make the precise flare to proper attitude very difficult - even in calm air.

1 0 . As a single-axis airplane this configuration is ink but if the pilot had anything to do other TABLE 4-1.- PILOT OPINION O F THE UJNGiTUDINAL HANDLlNG QUALlllES O F THE VARIOUS SST CONFIGURATIONS SlMULATED - CodinUed wot rating of control Pilot comments wd longitudirvl Configuration p a n m e t e r s handling qualities 6.p. = -1.3 B 4 . 5 - 5 . 0 1. There is a definite lack o f static stability.

Basic variable geometry 2. The airplane has good pitch response, but low pitch damping.

[st.u. margin = 3 percent F ) 8e,6 = 0 3 . The trim change with thrust seemed quite hrge.

6,/Aa = 0 This is became of the very 4 . In the approach, the workload is very high on glide path ccntrol.

loss level of static stahuty and no apparent @ch rate damping. Another objeotion i q the seemingly very high pitch trim chpnge with paver. AU three combined to make the airplane very difficult to trim, to hold in a stabilized rate of descent, to hold in a stahilized attitude, and to hold at a stabilized speed. So what y m a r e doing is Just punping, p h i n g , and pllling - spending 70 to BO percent of the time on longihdinsl control going down the glide path.

5. The flight director on a configuration like ulis makes a world of difference.

A 3.5 1 . The aircraft is not difficult to trim but it requires patience to get hands-off cozdition (Speed Variable geometrg with 6& = -5.2 pitch rate plus bandof 2tO 3 hob).

= 1.46 angle of attack 2 . Control power is quite adequate.

feedback ( B + A C ~ 6e/Aa E 1 . 5 3 . Response is snappy - there is almost an immediate feeling o f g and hildup of ti with augmentation Control i n * .

(Static margin = 3 percent F ) 4 . Pitch rate and a a r e nicely damped - very l i t t l e tendency to overshoot small or large pitch attitude changes.

5. The 1 0 n g i t ” U maneuverability is good.

6. The glide path control is satisfactory and the ability to eetablish a desired rate d descent was very good.

7. Thrust response is adequate. The trim change with Ulrust is still noticeahle, hut much lighter than that for the unaugmented variable-geometry codtguration with ulis same 3 percent static margin.

8. In the flare, the pitch response, pitch damping, and attitude control a r e good.

9. There is some difficulty in controlling the airspeed during the approach.

B 2 . 5 1 . The aircraft is not difficult to trim. The pitch response and control sensitivity a r e good.

Variable geometry with 6 6 6 , = -5.2 pitch rate plus The pitch damping is quite good and the longiturunal maneuverability is good. Speed control ee/i = 1.46 is acceptable.

angle of attack feedback (4 + 6dAa = 1 . 5 2 . Glide paul control is satisfactory. The ability to estnbUeh a desired rate o f desced is good augmentation and the speed control 1ya8 quite easy. Actually, speed cmtrol in itself is proLnhly not much easier on this augmented configuration than it 1ya8 for the unnugmented, it is Jnst that abmt (static margin = 3 percent F) 50 percent more time ie available to devote to speed control. The trim ehpoge with Uvust is much less than on the unaugmented eonfkpration which again nukes attitude control much easier.

3 . This configuration has very very good response, probphly the best response to this pou.

Actuplly, it is preferable to give up some static stahiUty on an airplane in order to get more responee. All in all, it is a very fine coofigurntion.

1. The tmchdm accuracy is actually aIKNt as good as possible. It is Just a d e r d practice and education.

TABLE 4-1.- PILOT OPWON OF TEE LONGITUDINAL HANDLING QUALITIES OF THE VARIOUS SST CONRGURATIONS SIhWLATED - Continued Pilot rating of control Configuration Pilo longitcdinal Pilot comments parameters handling qualitic = - l . C A 6.0 I. The airplane is difficult to trim. Far straight and level flight, a s well a s for an established rate of descent, constant attention is required to airspeed and attitude control.

(Static margin = 2.45 percent i 6 $ = 0 2. Indicated a response i s sluggish and lags column inputs by about 1 second. Discernible 6@a= 0 changes in 0 appeared to lag column inputs by about 1 ; seconds.

3. Trim changes with thrust a r e pretty mild.

1 . Although the pbugoid mode is present, it is not obvious a t alkitude or on the glide slope and is not bothersome to the pilot.

5. A l o w pitch damping combined with the large inertia produces a tendency to overshoot any Small e change.

6. Glide path control is not t w difficult with the flight director, but it requires constant attention to airspeed and a great deal of throttle maniplation. The response to the throttle is fairly slow if the airspeed is off more than 3 to 4 knots.

7. During the landing flare, control of attitude and sink rate are very difficult, and there is a tendency to overdrive the controls.

8. The major reasons that this configuration was downgraded are l a & of static stability, lack of pitch damping, and very sluggish pitch response.

Basic fixed geometry 6e/5e/a, = -1.0 B 4.5 - 5.0 1. The aircraft is difficult to trim. It can be finally trimmed, but if left alone it seems to start to diverge after just a few seconds.

= 2.45 percent 5 6 d = O (static margin 2. The pitch response is slightly better than the umugnented variable-geometry configuration - 6,pa = 0 but, more is desirable.

3. Law speed stability: it is difficult to hold a desired s p e d and requires a large percentage of the workload.

4. On the glide slope, a 10 to 15 second period oscillation was noticed - Just constant nose-up and nose-down a t about that frequency.

5. The pitch damping is low. The aircraft would be hard to fly in turbulence. Small attitude changes are difficult because of this low damping.

5 . No tendency toward PI0 was noticed either in the flare or on the glide slope.

7. The ability to establish a desired rate of descent was fine a s long as there is no concern with calibrating that rate of descent with something else - e.g., the glide slope; but, when rate of descent is c u e d , e.g., from 500 to 450 ft/min (2.5 to 2 . 3 m/sec) for a few seconds and back to 500 ft/min, to correct a glide slope error, it is pretty rough.

3. The trim change with thrust does not seem to be a t quite as hi# a level a s the unaugmented variable-sweep configuration.

3. Ground effects cause a very abrupt nose-down pitch at an attitude of a h t 30 feet (9m). An increase in C , is also evidenced by the a d d i t i o d amount of power needed to maintain speed for stabilizing and sort of feeling around for the ground. No chaoge was noticed in CL, but, of course, this could be masked quite a bit by the other problems.

b . The flight director makes a great difference in flying the approaches - much more so than it did when flying the unaugmented variable geometry.

. The major reasons for downgrading this configuration are sluggish pitch response, low pitch damping, difficult speed control, and the nose-down pitch due to ground effects.

Basic fixed geometry 8 . p . = -1.0 C 5.0 . The aircraft is difficult to trim within *3 knots of desired speed.

1 . The pitch response to large control inputs i s satisfactory but the pitch control sensitivity (Static margin = 2.45 percent F) ,,/e = 0 is poor (very low).

,@a = 0 . It is difficult to hold desired trim s p e d within r3 h o t s and this is even worse in turm.

. Pitch response to small or normal control inputs is poor (low), the damping is low, and there is a strong tendency to overshoot when making small attitude changes.

. The longitudinal maneuverability is satisfactory.

.Glide path control: there were large pitch excursions.

. The ability to establish rate of descent seemed adequate but when changing throttle, excessive longitudinal controlling was required to maintain speed.

. The thrust control response is good but the trim change with thrust is high. This trim change dDes not really help get the MSB dawn - far instance, when you pull o f f power for a glide path correction, you have to force the nose down with the elevator control; and when you have established the desired speed after this maneuver, you are left holding a pll force excessive for trim.

. For the W g liare control, the very large trim change in ground effect posed the question of whether there was enough control io complete the landing. Pitch response is low for normal displacements of the column. T h e pitch damping is poor, therefore, attitude control is poor and the tendency is for large overshoots.

Touchdown accuracy is poor.

TABLE 4-1.- PILOT OPINION OF THE L O " A L HANDLING QUALITIES OF THE VAFUOUS SST c o m c u R A n o N s SIMUIATED - continued m o t rating of control C d i a t i o n wot comments mot longibxlinal parameters handling quauties - Baeic fixed geometry 6,/aC = -1.0 D 3.5 - 4.0 1. It is not difficult to trim at the desired speed once the pilot becomes accustomed to the pitch and airspeed sensitivity.

(Static margin = 2.45 percent 6 e b I 0 2. Pitch response to either large or small control inpis is satisfactory.

6./Aa = 0 3. The pitch damping appears to be adequate - there is some minor tendency to overshoot when maldng emaU attitude changes. No tendency toward PlO.

4. It is somewhat difftcult to hold exact desired airepeed, but to hold wUhin t5 knots, M) problem. The airspeed varies r( to 5 knots from trim.

5. The trim change and speed change with thrust a r e satisfactory.

6. Longitudinal maneuverability: the FJg during wind-up turns is satisfactory. Moderate to large thrust adjustments a r e required at and beyond 3 0 ' banked turns to hold level flight.

Normally, however, the angle of kank would never be larger than 20' to 30'.

7. On the appraach, the glide path control is satisfactory and the s p e d control i s good although somewhat larger variations in speed occur with unatahle thrust-velocity relationship.

Because of this, an increased requirement is placed on proper coordination of throttle and elevator.

8. Guhjectively, the pitch sensitivity and damping appeared greater on ulis unaugmented fixed- geometry configuration than it did on the unaugmented variable-geometry confi-tton, thereby making the pitch response a p p e v better. The control of flight path and airspeed required somewhat increased pilot attention over the unaugmented variable geometry - this is particularly true in increased thrust adjustments and need for cmrdinating those with pitch corrections.

Basic fixed geometry 6& = -1.0 E 4.5 1. The ability to trim at the desired speed i s acceptable.

2. For large control tnplts, the lag in i n i t i a l response is apparent but the pitch attitude (Static margin = 2.45 percent K e / e = 0 overshoot appears to be too great - about 2-1/2' in roller coaster maneuver where 1 ' 6,pa = 0 would be expected. The pitch response to small control inplts i s slow Lmt sure.

3. The pitch damping is no problem, but the tendency to overshoot during attitude changes is quite evident. There i s definitely a tendency toward PlO, especially in maneuvering turns, where the long short period is apparent.

4. The ability to hold desired speed depends on the degree of concentratton; seems easier of airspeed seems very small in static condition but is under the hood. The oscillation apparent in maneuvering. The trim change with thrust is apparent but not objectiomhle.

The speed change with thrust is secondary. The primary effect seems to be on rate of climb. A direct connection behveen throttle and rate of climb is evident. which may be a desirable feature if properly used in IFR approaches.

5. Longitudinal maneuverability: a fair amount of concentration is required because of tendency to slow down, to pitch m s e dawn, and to oscillate in airspeed. The Fc/g does not have a steady feel because of these tendencies.

6. When trying to control glide path one must concentrate on attitude, but not excessively.

The ability to maintain the desired approach speed requires an appreciahle amount of work, especially in rough air.

7. During the flare, pitch response, pitch damping, and attitude control a r e adequate.

8. No apparent pitching or floating tendency near the ground was noticed. The touchdown acnvacy was rather consistent. However, the rather rapid a i r s p e d decrease in ground effect is undesirable.

Basic fixed geometry 6& = -1.0 3.0 1. The ahiuty to trim at the desired s p e d is very good. No short period oscUlptions were I F noticed.

(Static margin = 2.45 percent 6 & = 0 2. Pitch response to either large or small control inplts is satisfactory. The phugoid to be neutrally damped and no short period PlO tendency was noticed.

6,/Aa = 0 seemed 3. It is easy to hold the desired speed within *5 knots, Lmt the pilot has to be alert. There is no apparent oscillation of airspeed. The trim change with Ulrust was very mild and in the p r o p r direction. The speed change with thrust did not seem to be as responsive as the unaugmented variable sweep c o o f i w t i o n , but the desired speed could be held to within 55 kmta easily.

4. The longitudinal maneuverability is goad.

5. Glide path control is satisfactory. A mixed technique of throttle for speed and elevator for attitude pnl vice versa was used. Either technique is s a ~ c t o r y .

6. In the flare, the pitch response and pitch damping were very good. Only small control applications were r-ed for which there was little lag in response. The attitude could be controlled very precisely in smooth air.

7. There were no apparent grouad effects and the touchdown accuracy d ulis ConfigVRtion is equal to current jet transports.

I I. I OF THE VAFUOUS SST CONFIGURATIONSSBWIATED - Contirmed mot rating of Control

Confirration - '04-

wot comments p v u n e t e r s huding qualities Basic lked gMmehq 8e,8c=-1.0 4.7 G 1. This wnft-tion would be somewhat "It to trim in anythhg but vary smooth e.

2. Eitch response wm sluggish for either luge or small m h o 1 Inplts. Tbc pitch dunping (StaUc margin = 2.45 percent Z) 6,/8 = 0 was good. There was M tedency to overshoot auring rttiMe chngee and m t d e n q 6.pa = 0 torprd mo.

3. It wae quite difficult to mpinhltl desired sped. No trim change with Uvust was detected, and speed changes with thrust were eptistretory.

4. L o n @ b U mnneuverablllty: The stick forces were too light, pad the npeed bleed-ofl due to drag, re- from the normal acceleration, SA8 excessive. ai# rates of descent occurred, as high as 1500 to 2000 i t / & (7.6 to 10.2 m/sec) In a 45' mnk, 5. Glide path control was saustactory. l t required a lot of attention to cootrol speed with throtlle. The reasons a r e the d n g variation due to the mrmal acceleration pial also the low static 1 0 - stability. The forces ami the momenb generated at c f t trim sped seemed quUe low in terms of esectiveness In try% to keep it on speed.

6. Because of excessive turbulence, landing evaluations were not attempted. The l o @ M cha+aeteristies of this uruugmented fixed-geometry configuraHon were g l e n a Cooper r a t i n g of 4.7, primarily because t h i s configuration was a llffle bit worse t h a n the uaugmented ppripble-geometry confkgrahlon.

Fixed geometry with a&.; -4.0 A 2.75 1. The aircraft was difficult to trim for hplds-aif Iught. There appears to be a trim speed pitch rate Plus bnnd of 3 to 4 knots on either side of the desired trim speed. Precise control of trim speed 6 & = 1.46 angle of attack requires constant attention.

feedim& (e' + A O ~ a e p n = 1.0 2. Very snappy pitch response. Gxd sensitivity rate which Is sdequate for any mmeuverlng augmentation e n c d e r e d In approach or l a n d i n g . Pltch damping is very good. Essentially m tendency to overshoot.

(Static margin = 2.45 p r c e n t a) 3. Iar.@tudinal lnaneuverabuuy Is good.

4. Glide ppth control is SatigIaetOry. Corrections can be ma& easily. There is M problem in estpblishing a desired d e of descent, ulnnlrs to good pitch codrol. Sped control requires much attention. Both speed and trim changes with Uvust are adequate.

5. Flare control is gmd; there is a slim tendency to bobble but thb is pilot induced and w d d disappear with learning. The pitch response, pitch dpmplng, and pttitude control during the flare are all good. The toueM0wn accuracy is improved over the uapugmented fixed-geometry configuration.

6. Generally, the longitudinal characteristics are very good. The smppy responee and excellent damping allow precision control of pitch at a greatly reduced work level.

There are M real adverse characteristics on this ads, other than the bothersome effect of flying on the backside of the thrust-required curve d having a very llght gradient with speed.

Fixed geometry with ae/bC = -4.0 1 3.0 1. The a i r u a f t is difficult to trim.

l B

pitch rate plw 2. pitch response is quite adequate. There is more response Uua would ever be needed.

6 , / e = 1.46 angle of attack Pitch control sensitMty was very good.

feedback (e' + ad 6 4 A a = 1.0 3. The difficulty of holding trim speed may possibly be improved somewhat over the u a u g - avgmentation mented fixed-geometry coofiguration - but mt because d the speed stability - it would (static margin = 2.45 percent E ) be bemuse of being able to make small attitude corrections.

4. Response to control hwts is quite eatisfactory. The pitch damping vsd very good - there was no tendency to overshoot and no tendency toward P I O .

5. The longbihd M n e v v e M t y is good. The ahlllty to estnhllsh desired rate of sint was fine.

6. The trim change with m e r is a t a very low level - very acce#nble.

7. A slight me-down pitching moment due to ground effects wan noticed around 30 to 40 Ieet (9 to 12 m), which was easily controlled with the column.

8. The longitudinal characteristics of thls canfiguRUon were given a pilot rating of 3.0, and the reanoo it is not even better is the d d l y unpleaeaat charaeteristies of speed control - but it flys so well in otber respects that the time is available to devote to speed control.

TABLE 4 - 1 . - PIIDT OPINION OF THE ION-INAL EANLIUNG QUAI.TPIES OF THE VARIOUS 66T CONFIGURATIONS SMULATED - COdhXd Pilot rating d CDnhOl Pllot comments Cantipration Pilot p u P m e t u s handling qualities 1. The aircraft is not " l t totrim.

F M geometry vith be/ac - -4.0 C 4 . 0

pitch rate plus 2 . Pitch response to large control i n & # is Satisfactory and pitch control sensitivity is good b& = 1 . 4 6 q l e of attack 5. The pltcb damping is good; s d altitude changes can be made with preelslon - no overahoat.

feedtack (e' + A C ~

6 p a - 1.0

4.lan@udi~I maneuverability is good.

-cribtion 5. Glide path control VPB mt easy be-e d the problem d controlling the speed. When thrust (static margin = 2 . 4 5 percent a) is reduced for glide path correction, speed immediately hlb. The altitude must be forced m e down with the elevator to regrin sped. Aa speed is rcmed, 8 plll force is m a repuLed at the origlrd speed - the nose is down and more altitude thnn desired is bst, and the a p e d canlimes to go up. As a @-up is made to correct h c k to glide path, some t h w t is nddcd to ~tahillze an the desired path, but the speed staye up, thus requiring pmther t h w t reduction, and probpbly inltinting a similar chain d events a @ . CbPRc- teristics d the "hckslde" o f the thrust-required m e probably play a put in the sequence, but the initiating factor seems to be the large iDerUa which prevents the pirplpne from respondiog to a ebvlge in pltehing momeds quickly emugh.

6. The flare wntrol pitch response wp8 g a d , pltch damping SPB good. and a w e codrol wan good.

7. The tmchdown accuracy is pmr because of excessive floating tendency awed by ground effects. It is possible that I n the achml case the nose-down trlm change will tend to offset the floating tendencies, depending on pilot techniques. The major reason the longitudinal characteristics of thls conflgurrtIon were not given a better d i n g is because of thin excessive finking tendency near the grataa; the Cmper rating would be 2 . 0 otherwise.

be,%c = - 4 . 0 D 3 . 5 1. It is not seri-ly difficult to trim at the desired speed once the pilot becomes acmtomed Fixed geometry with to the pitch and airspeed sensitivities.

pitch rate plus 6 d = 1.46 angle of attack 2 . Pitch response to large control inpb h satisfactory. Response to small control inplts feedtnck (6 + ~ o ) Kepa = 1 . 0 IS good.

augmentation 5. Pitch damplng is good - there I s little or no tedency to overshoot during attitude changes.

(static margin = 2 . 4 5 percent 5) 4. The ability to hold desired sped is about the same p8 that for the unaugmented fixed- geometry eonfiguration - relatively easy to hold within *5 M a .

5. Trim change with thrust is small. Speed sensitivity to pitch attitude is high 8. The 1ongltudI"al maneuverability is the same an thak for the unaugmented fixed-geometry configuration - satisfactory.

7. Glide peth wntrol is satisfactory and I s controlled by coordinated use of throttle a d elevator.

8. In the flare, pitch damping and attitude control are good. €'Itch response is improved over the vnaugmeuied fixed-geometry configuration.

9. The pitching and floating tendencies due to grwnd effects are mlmr, and the touchdown accuracy is satisfactory.

10. The Increased pitch response and even better damping helps the wntrol during the flare.

1. It is relatively easy to trim at the desired speed.

Fixed geometry with a,,%, i. - 4 . 0 E 3.5 pitch rate plus 2 . Pitch response to either large or smnU control i n p t s is sati~factory.

K e ) = 1.48 mgle of attack 5. The pttch damping is good. There is w tendency to oaershoot during attitude chnnges feedback (6 +ad bdAa = 1 . 0 and no tendency h a r d PIO.

aupentntion 4. Definite wncentntion is requtred to hold desired speed.

(Static margin = 2.45 percent T ) 5. rim change with thrust is not too apparent. Speed chvlge with thrust is more appprent and seems normal.

6. The lon@udbal maneuserability is greatly improved over the unaugmented fixed- geometry configunllon. T h e augmented configuraUon seems &mer and a p p w s to have a slightly hi@er F g .

7. GUde path wntrol is Sztiefactory. The glide path PPB eontrolled with elevator and airspeed with throttle. It PPB mt difficult t o m a l r h h desired approach speed, but required concentrated effort. The better pitch respome ud better apmping helped speed wdral.

8 . In the flare, the pitch response, pitch dpmplng, and a w e control are good. No pitching or flaaung tendency due to g r d effects wan noticed, but airsped does decrease Rther rapidly in the flare. The hcMm accuracy PPB rather consistent.

TABLE 4-1.- mLQT OPINION OF THE LONGITUDINAL H A N D m G QUALITIES OF THE VARIOUS SST CONFIGURATIONS SIMULATED - Concluded PUOt rating of Control Pilot 1OngitudtIlal Pilot commente Configuration parameters handling qualities Ke/ac = -4.0 F 3 . 0 1. No essential change in the longitudinal characteristics was noticed between the unaugmented Fixed geometry with pitch rate plus and augmented fixed-geometry configurations.

ae/ = 1.46 angle of attack 2. The longitudinal characteristics were rated the same for augmentation on and off.

feedtack (i + ad ae/Aa = 1.0 augmentation (Static margin = 2.45 percent F ) 3 . 0 1. The aircraft i s still fairly difficult to t r i m .

Fhed geometry with 6e/6c = -4.0 0 pitch rate p l w 2. The pitch response was considerably improved over the unaugmented fhed-geometry a , $ = 1.46 angle of attack configuration. There i s still a M e b i t o f lag, but it looked Uke more elevator power, feedback (6 + ad Kepa = 1.0 shorter time constant, a d in general would be a more controllable configuration.

augmentation 3. Pitch damping appeared lower than the unaugmented because the response was faster; however, for step inplts the amount of pitch attitude springback was actually about like (static margin = 2.45 percent F ) we see i n large present day airplanes. There was some small tendency to overshmt or undershoot when making small attitude changes.

4. The speed change with thrust was satisfactory. There was no trim change with thrust.

TABLE 4-2.- LONGITUDINAL SHORT PERIOD CHARACTERISTICS OF THE VARIOUS SST CONFIGURATIONS TESTED Damped Undamped Configuration

c

period, sec fn, CPS Variable geometry, 9.6 0.141 0.672 1.190 basic Variable geometry, 14.1 0.208 0.940 2.456

B ,augmentation

Variable geometry, 5.1 0.278 0.705 2.462

( 6 + ACY) augmentation

Variable geometry, 30.0 0.102 0.945 1.211 aft c.g.

unaugmented Variable geometry, 5.9 0.206 0.755 1.953 aft c.g.

(6 + ACY) augmentation

~ Fixed geometry, 16.8 0.120 0.869 1.310 basic Fixed geometry, 7.1 0.231 0.793 2.301

(4 + ACY) augmentation

t

1.6 1.2 .a -0 -V e . 4 r- .- c

.-

.. .

E 0

r

-

0 V

-. 4

-

c c

-. 8

-1.2 I. I - > l-- _ L . I .

-1.6 120 125 130 135 140 145 150 Equivalent airspeed, V,, knots

-

- 4 0

-

- .- - 0

-

-20 - I Equivalent airspeed, V,, knots Figure 4-1.- Indication of stick-fixed and stick-free static stability for simulated unaugmented basic variable-geometry SST configuration.

2.0 0 Variablegeometry SST El Fixedgeometry SST 0 Some subsonic jet transports L8 L6 1 . 4 a E i 1 . 2 f 1 . 0 F

-

.a 2.

.6 k 5 .4 . 2 I I 1 1 1 1 1 1 1 1 0 . I .Z . 3 .4 . 5 .6 . 7 .a . 9 1.0

Short period damping ratio, 4

(a) W n against 5 .

1 . 6 0 Variable geometry SST 0 Ill Fixed geometry SST 1 . 4 0 Some subsonic jet transports

: 1 . 2

??

b J 1 . 0 5 a .

a .a

-

z

. 6

x

k 5 .4 c VI . 2 I I 1 1 1 1 1 1 1 1 0 1 .2 . 3 .4 . 5 . 6 . 7 . 8 . 9 1.0 Short period damping ratio, 4 (b) WD against 2 , .

Figure 4 - 2 . - Comparison of short period frequency (undamped and damped) and damping ratio of unaugmented variable- and fixed-geometry SST configurations with some present-day subsonic jet transports.

OD

3 r

I

I

I I

i

-1 I V T i m e of control release

I

-2 i

I

i

I

I

I

r I

I

~

0 1

2 3 4 5 6 7 8 9 10 11

Time, sec Figure 4-3.- Aircraft pitch rate response to an elevator pulse.

1.0 1.1 1.2 1.3 1.4 1.5

Load factor, n, g

z

12 -

..

ILU

8 -

I= E

-

4 -

l o

0 -

1.0 1.1 1.2 1.3 1.4 1.5

Load factor, n, g

Figure 4-4.- Indication of longitudinal maneuvering stability in a wind-up turn for simulated basic unaugmented variable-geometry SST configuration.

I

1 1 1 1 l l 1 1 1 1 1 l 1 l l l l l l Ill11 I I I I 11l11l111 I I1 I I I I I I 1 I I I1 1.2

.8

. 4 Variable-geometry SST

-. 4

----

Fixed-geometry SST

----

Subsonic jet transport 1.6 1.2

/

.8

. 4 -1 I . - I 1 2 3 Time, sec Figure 4-5.- Comparison of pitch response for the variable- and fixed-geometry SST configurations with a subsonic jet transport.

Normal pilot input

---- Modified pilot input

t

g ' Z

o s

4 r

0 1 2 3

Time, sec

Figure 4-6.- Example of pilot over-controlling i n order to compensate for inherent sluggish pitch response.

a3 80- 80- E

- 20 E

a=

-

- - 2 0 d

a= u a- a- -u c a- .- 3 c c

.- -

- 3 40-

- z 40- 1OZ

c .- -10 Q: c a a -

- a -

0 - 0 0 0 s a-

-

m

- 5 c m S m -5 : [ , I I I I .- c e v -5 " 7 1 1 _I

.-

a 1 6 12 8 4 0 1 6 12 8 4 0 Time before touchdown, sec Time before touchdown, sec (b) Augmented (a) Unaugmented Figure 4-7.- Landing time histories for the variable-geometry supersonic transport configuration showing effects of (6 + ha) SAS.

111 I I I l l I I I I 2.0 1.8 1 . t P i t c h r a t e p l u s alpha

( e + & >

augment at ion

1.4

P i t c h r a t e 1.2

(6 )

augmentation 1. c - 8 \ Variable ge 3met r y Fixed geometry .6 I

.4

I 1 .6 -7 .8 - 9 1.0 1.1 Short-period damping r a t i o , 5 Figure 4-8.- Effect of 6 and (6 + Aa) SAS on the short period damping ratio and natural frequency.

1 . 6 1 . 2 .8 . 4

-. 4

-. 8

-1.2 - 4 L -~ J ~- -.L-- ~ - 1 I I -1.6 .

120 125 130 135 140 145 150 E q u i v a l e n t airspeed, Ve, k n o t s - c - Basic --+I--- (;,+ma) E z L

I 20 u b

U a- a- - V ? L

- -

S - 0 c

- -

- V

-

-20 - -40 Enllivalent a i r w e e d , Vp, k n o t s Figure 4-9.- Comparison of stick-fixed and stick-free static stability for the variable-geometry, basic and (6 + Aa) augmented, configurations.

1.0 1 . 1 1.2 1.3 1.4 1.5 Load factor, n, g z c E !

L

- 12 U

V U a i - W - 2 40 .c Y- c E c

-

-

I I I -I m 1.0 1.1 1 . 2 . 1.3 1.4 1.5 Load factor, n, g Figure 4-10.- Comparison o f longitudinal maneuvering stability in a wind-up turn for variable-geometry, basic and (6 + Aa) augmented, configurations.

Augmentat ion Basic

-----

E i - t - & Time Figure 4-11.- Indication of initial response for basic and augmented configurations for a step column input.

-

1 . 2 - .8 5?

U L I o

-

.4 ! = E - 0 V

- -

- . 4 e c c V -

-. 8

- -1.2 I I .. I 1 . - - I . -2 -1.6 ' 120 1 2 5 130 1 3 5 140 145 150 Equivalent airspeed, Ve, knots - s - Basic - - +-- Aft c.g.

8l \

c

=

V LL. - -

- - -*-- - -

- 2 0 =

- - _ _ _ _

W - L L . L c W - V L

I

- 0 . E O l c

-

- -20

-

j -40

120 125 130 135 140 145 150 Equivalent airspeed, Ve, knots Figure 4-12.- Comparison of stick-fixed and stick-free static stability for the variable-geometry, basic and aft center of gravity, configurations.

1llllll1l11l1lll11l I IIIIIIIII Ill I 1 1 1 I Ill I I I

r E .

-

V

-

e

i-.

c I I

0 -. 4

120 125 130 135 140 145 Equivalent airspeed, Ve, knots 40 z

-

Y- V LL E W - - V L L 20 E P a - c E +

-

c E

O s

-

-4 -20 I. 1 . . ... 1 I . . I . . i I -8 120 125 130 135 140 145 150 Equivalent airspeed, Ve, knots Figure 4-13.- Indication of stick-fixed and stick-free static stability for the simulated fixed-geometry SST configuration.

p

! I 1 .I 1.2 1.3 1.4 1.5 1.0 1.1 Load factor, n, g

. . I : 0 .. 1 . 1 I _ . u

1.0 1.1 1.2 1.3 1.4 1.5 Load factor, n, g Figure 4-14.- Indication o f longitudinal maneuvering stability in a wind-up turn for simulated fixed-geometry SST configuration.

' 1 1 1 1 1 1 1 1 . 1 . 1 1 11111111 I I I 111111III I I . ~ I I 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 I I I I 1 1 1 1 1 1 1 1 120 125 130 135 140 145 150 Equivalent airspeed, V knots e '

. - e - Basic

---D--- ( e + &)

-

-

-

z

-

.. 20

V U - a- 0 - 0 c E

-

-

-4 - -20

-

I . . . .I 1 I -8

-

-40 120 125 130 135 140 145 150 Equivalent airspeed, Ve, knots Figure 4-15.- Comparison of stick-fixed and stick-free static stability for fixed-geometry, basic and (6 + Aa), configurations.

/ ' 0

d , '

/

;/ 2'''"

_1 1 I I 1 .

1.5 1.0 1.1 1.2 1.3 1.4 Load factor, n, g z

-

n u * U V LL

60 $

a- E 12 .c

-

I= E c E

-

8 0 I . I 1 1 - -I 1.5 1.0 1 . '1 1.2 1.3 1.4 Load factor, n, g Figure 4-16.- Comparison of longitudinal maneuvering stability i n a wind-up turn for fixed-geometry, basic and (6 + Aa), configurations.

1.2 v) 8 1 . 0 M IL-F-8785 t

-

8 .8 Unacceptable region c m S B

I -

- .6 c m c .4 .- I a , n .

5 .2

c m I I I I I I 1 I .1 .2 .3 .4 .5 .6 .7 .8

Short period damping ratio, <

Figure 4-17.- Longitudinal short period damping requirements of MIL-F-8785.

.30 o P R = 2.5 v) M IL-F-8785 n MI L-F-8785 V .Unacceptable region .25 c

-

Satisfactory region BPR = 2.9 PR = 3.4 5 c & P R = 3.0 5 .20 K . . ~~ - m 2 .15 c m c -0 .- 2 . l o m n I I I I I I I .. , - 1

- 1 .2 . 3 .4 . ; - 6 . 7 .8 .9 1.0

Short period damping ratio, 4

Figure 4-18.- Location of short period frequency and damping ratio of various SST configurations simulated on damping requirements specification chart of MIL-F-8785.

.. -

I c I

0 P i l o t A

P i l o t B P i l o t r a t i n g (Cooper s c a l e ) 0 0 E l 0 .10 .20 30 Short periad n a t u r a l frequency, fn, CPS Figure 4-19.- Variation of pilot ratings of longitudinal characteristics of variable-geometry configurations simulated w i t h short period natural frequency.

W UI 0 Variable geometry (basic) 0 Variable geometry (0 )

O V a r i a b l e -geometry ( e + h)

b Fixed geometry (basic)

0 Fixed geometry ( 6 -t &)

Present -day j e t transports

I

1.0 m PI W c\ d .8 4 - 1 .6 \

I-

\

t

\ Acceptable t o poor , \

.4

\ / /

.---

-P Unacceptable k o P R = 2.5 PR = 2.9 m OpR = 3.4 .2

= 4.1

PR = 5.4 J I ~ . . I L . . I I I- - . . - -. -- - _ 1 2.0

.2 - 3 .4 .6 .8 1.0

Short period damping r a t i o , 5 Figure 4-20.- Longitudinal short period criterion of reference 2.

- I 1 11111111111 1 1 1 1 1 1 1 1 1 1 1 1 Normal operation (satisfactory region)

E mergency

operat ion

/

0 Variable geometry (basic)

0 Variable geometry ( e + ha)

b Fixed geometry (basic)

L3 Fixed geometry (6 + ha)

U ! ? Z I l Present-day jet transports

I 1 I I I I I I

. 4 .6 .8 1.0 1.2 1.4

0 . 2

Short period damping ratio, <

Figure 4-21.- Longitudinal short period criterion of reference 3.

/

Type of boundaries commonly

proposed f o r large transports

o r bomber aircraft

Note: Airplane frequency and

damping must fall w i t h i n

bounded area t o be

accept ab1e

I '

I I

'K Estimated shaping of proper boundaries

I

f o r large transports

I

Short period damping ratio, <

Figure 4-22.- Estimated reshaping of an existing longitudinal short period criterion (reshaping as proposed i n ref. 4).

Variable geometry (basic) Variable geometry ( 0 ) Variable geometry (6 -t h) Variable geametry ( a r t c.g.)

Variable geometry (aft c.g.; 6 + h) IFixed geometry (basic) Fixed geometry (0 + h) Two present j e t transports PR = 5.4 b + m = 5 . 1 I 1 pR = 4.1 I:

/q$ PR = 3.0

/ 1m = 2 . 9

I

8 F ' R = 2.5

._ I . ! ! . I . 1 ! - I .!. . I 2.0 0 .2 - 3 .4 .6 .8 1.0 Short period damping r a t i o , ( Figure 4-23.- Longitudinal short period criteria of reference 4.

IIIIIII 5. LATERAL-DIRE CTIONAL HANDLING QUALITIES By Robert E. Shanks, Samuel A. Morello, and Jere B. Cobb SUMMARY An investigation w a s made t o determine the lateral-directional handling qualities of several super sonic transport configurations during the landing approach. The basic variable-geometry and fixed-geometry configurations were found to have satisfactory handling characteristics. Although the characteristics of the variable -geometry emergency-landing configuration (wings in sweptback position) w e r e considered to be unsatisfactory, it w a s determined that it would be possible to make landings with this configuration.

INTRODUCTION As pointed out in the introduction, part 1, the proposed supersonic transports differ markedly from current large subsonic jet transports and might therefore be expected to have different handling qualities. For instance, it w a s suggested in reference 1 that changes in pilot techniques may be required because o f the large difference between the relatively high lateral control sensitivity and the reduced control response in yaw of such configurations. The t e s t s discussed in this part were made to determine the effects of supersonic transport characteristics on the lateral-directional handling qualities.

RESULTS AND DISCUSSION The lateral-directional results of the flight tests a r e summarized in tables 5-1 and 5-2 and in figures 5-1 to 5-17.

Table 5-1 presents qualitative results in the form of pilot ratings and comments for the flight-test configurations. The numerical ratings assigned each configuration by the pilots a r e based on the Cooper pilot evaluation scale (presented in table 2-3). All the configurations, except one, were evaluated by a minimum of two pilots, some configurations being evaluated by seven pilots; however, the average pilot ratings presented throughout the discussion were taken as an average of the ratings of pilots A and B. In addition to the overall rating for each configuration, pilots A and B also assigned ratings based only on the Dutch roll characteristics for correlation with the criteria dealing particularly with the characteristics of the Dutch roll oscillation.

The data of table 5-2 and figures 5-1 to 5-9 present quantitative results obtained from flight tests. The data presented in figures 5-1 to 5-3 are taken from reference 2, but it should be noted that the sideslip angles have been reduced by 20 percent. This reduction approximates the sidewash correction to the angles indicated by the sideslip vane, whereas, in reference 2 this correction has not been applied.

Basic Variable-Geometry Configuration The basic variable-geometry configuration represents a supersonic transport of the at minimum sweep angle in the normal variable-wing-sweep concept with the wings landing configuration. In general, the lateral-directional characteristics of the basic variable-geometry configuration were good and w e r e characterized by good Dutch roll damping, approximately neutral spiral stability, good static directional stability, positive effective dihedral, good roll response and roll damping, and low workload. Although some adverse sideslip (sideslip right with right roll) was noted in performing heading changes, it w a s felt to be of small consequence and the heading-change precision was found to be good. The basic variable-geometry configuration was given an overall Cooper pilot opinion rating of 3.0 (see table 5-1) and one pilot (pilot C) gave it an unusually good rating of 2.

Static lateral-directional . . - - _ _ stability characteristics. ~- - - The static lateral-directional stability characteristics obtained from flight in steady sideslip a r e shown in figure 5-l(a).

The pilot comments of table 5-1 correlated with the data of this figure in that they showed this configuration to have good directional stability and satisfactory positive effective dihedral.

Dynamic lateral-directional stability - characteristics.- ~ .- The measured character - _ _ _ . _ -.

istics of the Dutch roll oscillation following a rudder input a r e illustrated in figure 5-2(a).

The characteristics of the Dutch roll oscillation of the basic variable-geometry configu- ration were obtained from these data and a r e shown in figure 5-6 in comparison with those o f current large jet transports. The data of figure 5-6 and table 5-2 show that the oscilla- is fairly well damped, but has a low natural frequency or long period (9.6 seconds).

tion The rather long period of the Dutch roll oscillation apparently w a s not objectionable to the pilots. Pilot D, however, did note that although the damping of the lateral oscilla- tion appeared to be satisfactory in t e r m s of cycles to damp to half-amplitude, the actual time to damp to half-amplitude was greater than desirable because of the long period of the oscillation. From observation of a sensitive sideslip indicator, he also noted a tendency to induce the Dutch roll oscillation but otherwise he was not conscious of it because the motion was slow and gentle. The increased damping for this configuration compared with that of current jet transports apparently compensates for the undesirable effect of low frequency. For example, the comparisons of table 5-2 show that the time t o damp to half-amplitude for this configuration is appreciably less than those a€ several representative large subsonic jet transports. Figure 5-6 also shows fairly good damping

and a ratio of roll angle to sidewise velocity @be (which is also a measure of roll

angle t o sideslip angle) which is about the same as those for current large subsonic jet The pilots generally considered the damping of the Dutch roll oscillation to transports.

be good and most of the pilots stated that the motion w a s predominately yawing.

The basic variable-geometry configuration had essentially neutral spiral stability with a calculated time to double amplitude of approximately 240 seconds. The roll- subsidence mode was heavily damped with a roll time constant of about 0.5 second. The flight-test results are in agreement with these predicted results since the pilots found this configuration to have neutral or slightly divergent spiral stability and good roll damping. (See table 5-1.)

Lateral control ~~ characteristics. - The variations of maximum rolling velocity and

maximum rolling acceleration with wheel deflection as obtained from rudder -fixed wheel step and wheel step reversal maneuvers, respectively, a r e shown in figure 5-3(a) for the basic variable -geometry configuration. The pilots generally felt that the roll response characteristics of this configuration were good but did note a small amount of adverse sideslip on turn entries. Heading-change precision w a s considered to be good in spite of the adverse sideslip when these maneuvers were made with only lateral control (no rudder ) .

Landing approach.- The lateral-directional control activity and airplane displace- ments for a typical approach and landing a r e presented for the basic variable-geometry configuration in figure 5-4. These time histories show small localizer deviations, normal bank and sideslip angles, and a low level of wheel and rudder pedal activity for the basic configuration.

Augmented Variable -Geometry Configuration Although the Dutch roll damping was considered to be good for the basic configura- tion, the tests were conducted in smooth air and it was believed that greater damping would be desirable for more severe conditions. Therefore, a sideslip rate damper, as described in references 2 and 3, was used to increase the effective value of the parameter Cn from 0 to 0.086 to increase the damping ratio from 0.18 for the basic

b

configuration to 0.28 for the augmented configuration with practically no change in the other lateral-directional characteristics.

Generally, there was no signficant change in the characteristics of the augmented configuration compared with the basic configuration other than a small increase in adverse sideslip during heading changes. Although the reason for this sideslip increase was not apparent, it was not felt to be important enough t o warrant further flight testing

lllllIlllI I I I l l

to determine the exact cause since the handling qualities of the configuration were still considered to be good. The increase in Dutch roll damping was not significant since it had been well damped for the basic condition. The workload w a s found to be low and the pilot ratings for this configuration averaged 3.25 compared with 3.0 for the basic config- uration, mainly because of the slight loss in heading-change precision due to adverse sideslip of the augmented configuration.

Static lateral-directional stability characteristics. . - - Since the augmentation did not

affect the static stability characteristics, they are the same as for the basic variable- geometry configuration shown in figure 5-l(a).

Dynamic lateral stability characteristics. _ _ ~ - The characteristics of the Dutch roll

oscillation are illustrated in figure 5-2(b). The records show the oscillation to be well damped but the period is slightly longer than that for the basic configuration. The data of figure 5-6 show the damping to be good and the ratio of roll angle to side velocity to be lower than that for the basic configuration. Although the damping w a s higher, the pilots did not note any appreciable differences due to the improved Dutch roll characteristics on the handling qualities of the airplane, probably because the stability of the basic configu- ration was good and because the effects of the augmentation would probably have been more beneficial in rough air than in the relatively smooth air in which the flight tests were conducted.

The spiral and roll-subsidence modes were virtually the same as those for the basic configuration as shown in table 5-2.

Lateral control characteristics. - The roll-rate and roll acceleration characteris-

tics of the augmented configuration were not noticeably affected by the sideslip rate damper and were essentially the same as those for the basic configuration. In the flight evaluations, however, two of the three pilots who evaluated the augmented configuration noted that the heading-change response was not quite as good as that for the basic config- uration because of a slight increase in heading lag; this heading-lag increase (there had been a slight heading lag noted for the basic configuration) was attributed to a larger adverse sideslip in the turns. There is no apparent explanation for the increase in side- slip resulting from the augmentation because the sideslip rate damper should have reduced the sideslip angle rather than increased it. In spite of the increased heading lag, however, the pilots who noted it felt that the precision in the turns was still good and the average Cooper rating was only downgraded from 3.0 for the basic configuration to 3.25.

Landing approach.- The lateral-directional control activity and airplane displace- ments for a typical approach are presented for the augmented variable-geometry config- uration in figure 5-4. The wheel displacements appeared to be a little smaller for the augmented configuration than for the basic configuration but otherwise the records for the two cases a r e generally the same. The workload w a s found to be low.

Degraded Variable-Geometry Configuration In order to broaden the scope of the investigation, the lateral directional character- istics of both the variable -geometry and the fixed-geometry configuration were degraded to determine the effect, if any, on the handling qualities of airplanes of the size of these supersonic transport configurations. To obtain the desired characteristics, the damping of the Dutch roll oscillation w a s reduced and the adverse yaw due to rolling velocity w a s increased.

A sideslip rate damper w a s used to increase the effective value of Cn

b

(from 0 to -0.1204) and the rolling moment due to rolling velocity parameter was made more negative (from -0.023 to -0.076); as a result, the Dutch roll damping ratio was reduced to 0.05 and the desired degree of adverse yaw was obtained.

The pilots found that the principal results of the degraded characteristics compared with the basic variable -geometry configuration were a moderate increase in workload (especially when close to touchdown) and a reduction in heading-change precision. This configuration w a s given a Cooper rating of about 4.4.

Static lateral-directional stability characteristics. - The degradation affected only

the Dutch roll characteristics; thus, the static stability characteristics are the same as those for the basic variable-geometry configuration. (See fig. 5-1(a).)

Dynamic ~ ~~~~ lateral stability characteristics.- The characteristics of the Dutch roll oscillation for the degraded variable-geometry configuration a r e shown in figure 5-2(c).

The damping is seen to be low and the period of the oscillation practically unchanged from the basic configuration. The pilots stated that the oscillation w a s easily excited by abrupt or moderate lateral control and seemed to be present most of the time. The data of figure 5-6 show that this configuration has damping and a ratio of roll angle to side velocity similar to present-day large subsonic jet transports but, like the basic configu- ration, has a lower frequency.

The spiral stability appeared to be neutral to the pilots and the roll-subsidence time The calculated values for these modes a r e presented in constant was still desirably low.

table 5-2 and show no significant differences from the basic configuration.

Lateral control characteristics. - The variations of maximum roll rate and maxi-

mum roll acceleration with wheel position show that the degradation had virtually no effect on the acceleration but reduced the peak roll rate per degree wheel deflection to about 80 percent of that for the basic configuration (compare roll rates in figs. 5-3(a) and 5-3(b)), probably because of the increased adverse sideslip in combination with the fairly high dihedral effect. In addition, the roll r a t e s were found to be oscillatory because of the effects of the lightly damped Dutch roll oscillation. The lateral con- trol power was judged t o be more than adequate but there was an appreciable heading lag due to the larger adverse yaw due to rolling velocity of the degraded configuration.

The precision in heading changes was relatively poor and coordination with the rudder was found to be difficult.

Landing approach. - Typical time histories of control activity and airplane displace-

ments for landing approach are presented in figure 5-4. These records show more rudder activity and larger sideslip displacements over most of the approach for the degraded than for the basic configuration but otherwise there were no significant differ- ences. The increase in rudder activity is evidence of the difficulty in coordinating the rudder and wheel noted by both pilots.

In general, the workload w a s not found to be high in the approach, although it was for the basic variable-geometry configuration. The workload was not higher than that higher because throughout most of the approach, the pilots preferred to concentrate on keeping on the localizer and they permitted the airplane to oscillate, knowing that the Dutch roll motion was stable even if lightly damped. Only when the airplane neared touchdown w a s the Dutch roll motion closely controlled by the pilot to assure a good landing. The workload in this configuration, however, would probably be very sensitive t o turbulence and would require more attention throughout the entire approach, especially when the comfort of the passengers is a consideration.

Variable -Geometry (Emergency Landing) Configuration The variable-geometry configuration with wings at maximum or cruise sweep angle is an emergency landing configuration and, as such, w a s investigated only briefly to deter- mine whether it would be flyable. Only one pilot flew and evaluated this configuration.

In general, the lateral-directional characteristics of the variable -geometry (emergency landing) configuration were unsatisfactory because of the weak roll control, large positive dihedral, and low Dutch roll damping. This configuration, however, w a s considered to be acceptable for emergency operation and w a s given a rating of 5.5.

Static lateral-directional -. -. - - . stability characteristics. - The lateral-directional static stability characteristics obtained from flight in steady sideslip are shown in figure 5-l(b).

The largest difference in static characteristics between the variable-geometry (emergency landing) and the variable -geometry configurations previously described is the much lower lateral control effectiveness of emergency landing configuration. Other- wise, the data of figure 5-l(b) show that this configuration has good directional stability and positive effective dihedral. Qualitatively, the pilot found the directional stability to be fairly high; the effective dihedral, very high; and the lateral control, weak.

Dynamic lateral-directional stability ~ characteristics.- The characteristics of the Dutch roll oscillation of the variable-geometry (emergency landing) configuration are illustrated in figure 5-2(d). The frequency and damping characteristics of the Dutch roll oscillation are shown in figure 5-7. The oscillation has fairly low damping ( 5 = 0 . 0 9 ) , a high frequency (cod = 1.4), and a high ratio of roll angle t o sideslip velocity relative to the other configurations investigated and to current large jet tran'sports. 'The Dutch roll damping appeared to be low to the pilot, and there was a tendency with normal frequency of control inputs to sustain the oscillation rather than to dampen it. Based only on the Dutch roll characteristics, the pilot assigned a rating of 4.5 to this configuration.

The spiral stability of the variable-geometry (emergency landing) configuration w a s high probably because of the high effective dihedral. The roll-subsidence mode was not heavily damped and had a roll-time constant of 1.7 seconds. The pilot could not estimate the roll damping because of the low damping of the Dutch roll oscillation. The high ratio of roll t o yaw of the Dutch roll oscillation is also attributed to the low roll damping in addition to the high effective dihedral and to some extent to the appreciable decrease in the roll inertia.

~~ Lateral control characteristics. - The variations of maximum rolling velocity and

maximum rolling acceleration with wheel deflection are shown in figure 5-3(c) for the variable-geometry (emergency landing) configuration. The roll rate per degree wheel displacement is about the same as that f o r the basic variable-geometry configuration but the roll acceleration per degree wheel displacement of the cruise configuration is only about one-third that of the basic variable-geometry configuration. The responses to a wheel step control for the two configurations (basic and emergency landing) a r e compared in figure 5-8(a) to illustrate the poor response characteristics of the cruise .configuration.

The response to a loo wheel input of the basic variable-geometry configuration shows a

steady increase in roll angle, a maximum roll rate of about 3 O per second initially which reduces gradually to about 2' per second in the interval shown, and a moderately large resulting sideslip angle. The variable-geometry (emergency landing) configuration, however, reaches approximately the same maximum roll r a t e initially but then the roll rate decreases to about zero because of the influence of the high effective dihedral; the oscillatory character of the motion is attributed to the fact that the Dutch roll mode was excited during the maneuver. This record also illustrates the pilot's comment that con- tinuous wheel displacement was required to maintain the desired roll angle. (See table 5-1.) Because of the high effective dihedral of this configuration, the rudder was found to provide good roll control and could be used in conjunction with the wheel for this purpose (see fig. 5-8(b)); this combination might not always provide satisfactory roll con- trol as will be pointed out under the discussion of the landing approach.

Landing approach. - The lateral-directional control activity and airplane displace-

ments for a typical approach and landing are presented for the variable-geometry (emergency landing) configuration in figure 5-4. These time histories show relatively small localizer deviations, small bank and sideslip angles, a low level of rudder pedal activity but that a relatively high level of wheel activity and fairly large wheel displace- ments were used to keep the airplane displacements small.

The evaluation pilot observed that the combination of high effective dihedral and weak lateral control of this configuration could result in a dangerous condition near the ground in a cross-wind landing. In such a situation, the approach would probably be made in a crabbed attitude which requires a rudder control just before touchdown t o aline the airplane with the run- way. The resulting sideslip and effective dihedral will cause the airplane to roll, but with weak lateral control, it would be difficult t o hold the wings level for the landing. To illustrate the effect of high roll response to changes in sideslip, the records of roll angle and sideslip angle following a rudder pulse are compared in figure 5-8(b) for the two basic variable-geometry configurations. These records show the roll response of the variable -geometry (emergency landing) configuration t o be about twice that of the variable -geometry configuration although the sideslip angles a r e approximately the same.

Basic Fixed-Geometry Configuration The basic fixed-geometry configuration represents a concept in which the basic airplane geometry is the same for all flight conditions. In general, the lateral-directional characteristics of the basic fixed-geometry configuration w e r e good and were character - ized by good Dutch roll damping, good directional stability, positive effective dihedral, and the lateral-directional workload in the approach and flare was low. On the other hand, the low damping of the roll mode required a little extra care to make precise heading changes. A little sideslip w a s also noted in turns but was probably not signifi- cant. The average pilot rating f o r the basic fixed-geometry configuration was 3.5. (See table 5-1.)

Static lateral-directional -- stability characteristics.- - The lateral-directional static stability characteristics obtained from flight in steady sideslip a r e shown in figure 5-l(c).

Qualitatively, the pilots' comments in table 5-1 indicate good agreement with these data in that they stated that the directional stability was good and the effective dihedral was positive. Several of the pilots commented that the effective dihedral w a s high and they attributed the rather high sensitivity to rolling motions to the dihedral effect when rudder w a s used to keep the wings level.

Dynamic lateral-directional - _.

- stability characteristics.- The characteristics of the Dutch roll oscillation of the basic fixed-geometry configuration a r e illustrated in fig- ure 5-2(e). The frequency and damping characteristics of the Dutch roll oscillation of ~ . . - - _ _... --..a. 1.11111--.11.1 I I . I I . I I . 1m.11.11. I, 1 . 1 1 1 1 1 I 1 1 1 1 1 . 1 1 1 I II I I 11.11.1111 I 111 I111 1 1 1 1 1 1 I1111111 I - this configuration a r e shown in figure 5-9. The Dutch roll oscillation of the basic fixed- geometry configuration is well damped, and this configuration has a ratio of roll angle to sideslip velocity and an undamped natural frequency representative of current

@be

large subsonic jet transports. Although the actual damped period for this supersonic transport configuration is a little longer (about 33 percent) than that for the subsonic transports, no unfavorable pilot comments were made about the longer period.

Although the spiral mode of the basic fixed-geometry configuration should have been divergent (see table 5-2), the flight records show that it was convergent. In general, the pilots noted that the spiral stability w a s either neutral or positive; pilot C thought that the positive spiral stability w a s good but pilot G indicated a preference for neutral spiral stability.

The principal pilot criticism of this configuration resulted from the roll mode damping which most of the pilots felt to be a little low. It was not a serious deficiency, however, because the average of the pilot ratings for this configuration was 3.5. For the roll-time constant of 0.8 (see table 5-2), this evaluation is consistent with the results summarized in reference 4, which a r e shown in figure 5-10. It should be pointed out, however, that the curves shown in figure 5-10 a r e for smaller airplanes and may not be strictly applicable for super sonic transport configurations .

_ _ ~ - control characteristics. - The variations of rolling velocity and rolling

Lateral acceleration with wheel deflection are shown in figure 5-3(d) for the basic fixed-geometry configuration. Comparison of the data of figures 5-3(a) and 5-3(d) shows that the roll acceleration was a little higher and the roll rate per degree wheel deflection about twice as high for the basic fixed-geometry configuration as it w a s for the basic variable- geometry configuration. Most of the pilots liked the high initial response and roll rate but several thought it might be a little too sensitive. This result combined with the low damping in roll mentioned previously produced a tendency for the pilots to overshoot in the turn maneuvers which required a little extra attention to the controls. Several of the pilots also noted some heading lag and adverse sideslip in turns made with aileron-alone contr 01.

Landing approach. - The lateral-directional control activity and airplane displace-

ments for a typical approach and landing a r e presented for the basic fixed-geometry con- figuration in figure 5-5. These time histories show small roll and sideslip angles and a low level of wheel and rudder pedal activity for this configuration. Although the localizer command signal w a s rather large over most of the approach, it represents an angular deviation and the airplane w a s actually converging on the runway throughout the approach.

It was therefore a reflection of pilot technique rather than an indication of difficulty in tracking.

Fixed-Geometry Augmented Configuration The purpose of the lateral augmentation used on the fixed-geometry configuration was to improve the roll damping and thus to eliminate or reduce the tendency to over- shoot in turn maneuvers, Ground-based simulator results indicated that a 50-percent increase in the damping-in-roll parameter C would be desirable and this increased

%

damping in roll was used in the airplane. The increased damping in roll was accompa-

nied by an increase in Cn . The effect of this augmentation was to improve the heading

change precision compared with the basic configuration; accordingly, the average rating was improved to 2.75 from 3.5 for the basic configuration.

Static lateral-directional ~ stability characteristics.- - The increased roll damping did not affect the static stability characteristics and they a r e the same as those presented in figure 5-l(c) for the basic fixed-geometry configuration.

Dynamic lateral stability characteristics.- The characteristics of the Dutch roll oscillation a r e illustrated in figure 5-2(f). Comparison of the records of figure 5-2(f) with those of the basic fixed-geometry configuration in figure 5-2(e) shows very little difference between the Dutch roll characteristics of the two configurations. The data of figure 5-9 show that the measured damping was somewhat higher for the basic configu- ration than that for the augmented configuration but both were at a high level of damping.

The augmentation also reduced the ratio of roll angle to side velocity and the undamped natural frequency about 10 percent.

Although the predicted spiral instability w a s reduced somewhat by the augmenta- tion, it w a s essentially neutral and the effect would be negligible. As in the case of the

basic fixed-geometry configuration, the actual spiral stability was positive; the conver -

gence, however, was slower for the augmented configuration. The intended effect on the roll-subsidence mode w a s achieved and the roll-time constant was reduced from 0.80 sec- ond for the basic configuration to 0.57 second for the augmented configuration. (See table 5-2.)

Lateral control characteristics.- As expected, the roll rate was appreciably affected by the improved roll damping and reduced t o about 75 percent of that for the (See fig. 5-3(e).) The roll rate, however, w a s still basic fixed-geometry configuration.

considered to be good and the tendency to overshoot or undershoot in turns was elimi- nated, at least in the evaluation of two of the pilots who flew this configuration. On the other hand, although pilot C was aware of a slightly reduced roll rate, he could detect very little difference between the augmented and basic fixed-geometry configurations.

This result suggests that a further increase in roll damping might be beneficial.

Landing approach.- The lateral-directional control activity and airplane displace- ments for typical approach are presented for the augmented fixed-geometry configuration in figure 5-5. In general, the control inputs given for the augmented configuration during the approach a r e fewer, the sideslip displacements a r e about the same, the roll displace- ments a r e more frequent, and localizer tracking is better than those for the basic fixed- geometry configuration. The workload was generally considered t o be low in the approach and heading-change precision was improved over the basic configuration.

Degraded Fixed-Geometry Configuration The fixed-geometry configuration was degraded in the same manner and to the same as the variable-geometry configuration and with essentially the same effect on the degree handling qualities, namely, a little higher workload and lower precision in making heading changes than for the basic configuration. Both pilots gave the degraded fixed-geometry configuration a rating of 4 . 5 .

Static lateral-directional stability - _ _ characteristics.- The degradation did not affect the static characteristics; thus they are the same as those for the basic fixed-geometry configuration which is given in figure 5-l(c).

Dynamic lateral stability characteristics.- The characteristics of the Dutch roll oscillation for the degraded fixed-geometry configuration a r e shown in figure 5-2(g).

The damping is low and the period of the oscillation about 0.6 second shorter than the period of the basic configuration. The pilots reported l e s s of a tendency to excite the oscillation when using normal controls for this configuration than for the degraded variable-geometry configuration, although rapid lateral controls did cause the airplane to oscillate.

The spiral mode appeared to be neutral or slightly convergent; this result is about the same as that for the basic configuration which w a s slightly convergent. The pilots indicated that the roll damping w a s not quite as good as that for the basic configuration and the calculated roll time constant shown in table 5-2 is about 10 percent higher for the degraded configuration.

Lateral control characteristics.- The roll-acceleration characteristics for the degraded fixed-geometry configuration a r e virtually the same as for the basic fixed- geometry configuration but the roll rate is appreciably higher as shown by comparison of figures 5-3(d) with 5-3(f). This result is evidence of the lower roll damping which was previously noted in the discussion of the roll-subsidence mode. The principal objection to the lateral control was the lack of precision in making heading changes and the appre- ciable adverse yaw associated with rapid heading changes.

Landing approach.- .. The lateral-directional control activity and airplane displace- ments for a typical approach are presented f o r the degraded fixed-geometry configura- It should be noted that random rudder inputs shown in figure 5-5(c) at tion in figure 5-5.

about 25, 65, and 93 seconds were given by the safety pilot to simulate gust disturbances to help the evaluation pilot t o assess the handling qualities of the condition. Except in response to these rudder inputs, the records show little difference from the other fixed- geometry configurations. The resulting work level for the degraded configuration was, however, a little higher than that for the basic configuration.

REQUIREMENTS AND CRITERIA Because the number of tests w a s limited and no parametric studies were made, criteria could not be established for supersonic transports in the landing approach.

From the results of the flight tests, however, it was determined which configurations were satisfactory and which were not, and these results are compared with existing c r i - t e r i a and with data relating pilot rating and various Dutch roll stability o r roll-control characteristics. The pilot ratings used in figures 5-11 to 5-13 were estimated by the pilots by considering only the Dutch roll characteristics and are the average of the ratings of pilots A and B as given in table 5-1. This procedure was followed to make a direct comparison with the other data in these figures which correlate the Dutch roll oscillation characteristics with handling qualities.

Variation of Dutch - roll - damping . with - . rolling . parameter.- _ _ Figure 5-11 presents the existing lateral directional damping requirements defined in the military specifications of reference 5 by the reciprocal of the cycles required for the Dutch roll oscillation t o damp to half-amplitude and the roll-to-side velocity ratio @/ve. The Dutch roll charac- teristics of the supersonic transport configurations of this program and of current large subsonic jet transports are compared with the requirements of figure 5-11. The pilot evaluations for the various supersonic transport configurations appear to be in good agreement with the boundaries shown in the figure. All but two of the supersonic trans- port configurations had the low ratios of roll to side velocity representative of the current jet transports; the degraded fixed geometry and variable-geometry (emergency landing) configurations have higher values and the corresponding ratios of roll to sideslip angle

( ; = 1.6 and 2.5, respectively are above the value of 1.5 suggested as acceptable for the

landing approach in reference 6.

The Dutch roll damping and Variation of Dutch - _ _ roll frequency -_ with - damping _. - - ratio.- - frequency characteristics of the configurations tested are compared in figure 5-12 with the lateral oscillation criteria proposed as a revision to the existing specifications of reference 5. Although the results of the flight t e s t s for the fixed-geometry configurations generally are in agreement with the criteria of figure 5-12, the results for variable- geometry configurations a r e not consistent with the criteria.

The basic and augmented variable-geometry configurations were found to have good handling qualities (pilot evaluation ratings of 2.4 and 2.1, respectively, based on Dutch roll characteristics) but are located in the unacceptable region shown in figure 5-12. It is evident that the low frequencies of the lateral oscillations rather than the damping ratios are responsible for their locations with respect to the boundaries.

Variation of pilot ~~ rating with damping.- Pilot rating has recently been related to the damping parameter in several papers. (For example, see refs. 7 to 9.) The characteristics of the test configurations of this investigation are compared with data 7 and 8 in figure 5-13. The test points agree with the reference data from references except for the variable -geometry (emergency landing) configuration.

Several factors may have contributed to the rather poor pilot rating for this con- figuration. First, the Dutch roll oscillation always seemed to be present because it was excited by almost any control input as well as by external disturbances and, because of the relatively short period, there was a tendency for the pilot to sustain the oscillation rather than damp it. Second, the high ratio of roll to sideslip of the Dutch roll oscilla- tion made it more objectionable. Finally, this configuration had several poor stability and control characteristics; thus it was the most difficult configuration in which to eval- uate the Dutch roll independently of the other lateral-directional characteristics.

Spiral stability characteristics. - The only requirement given in the existing mili-

tary specifications of reference 5 for the power-approach condition is that if the spiral motion is divergent, the rate of divergence shall not be so great that after a small dis- turbance in bank with controls fixed, the bank angle is doubled in less than 20 seconds The calculated data of table 5-2 show that all the in the power-approach condition.

supersonic transport configurations met the requirement. The fixed-geometry con- figuration was actually slightly convergent and, as such, it would be considered to be satisfactory according to the military specifications.

The calculated values of the spiral damping a r e compared with the boundaries taken from reference 10 and presented in figure 5-14. All the configurations are shown to be satisfactory according to the boundaries of figure 5-14. With one exception, the variable -geometry (emergency landing) configuration, the spiral stability characteristics were found to be satisfactory by the evaluation pilots; the strong spiral stability of the variable -geometry (emergency landing) configuration, resulting from the high effective dihedral, caused poor lateral control characteristics in that bank angle could not be

maintained without holding continuous, or even increasing, wheel displacement. Refer -

ence 10, however, suggests that T or T2 should be greater than 14 seconds for 1/2 satisfactory spiral stability characteristics; if this recommendation is applied t o the

variable-geometry (emergency landing) configuration T = 12 seconds , i t would only

( 1/2 )

be considered acceptable.

Cross-coupling characteristics.- Figure 5-15 is taken from reference 7 and w @/wd shown relates pilot rating with the aileron yaw parameter Values for w&/wd' in figure 5-15 were calculated by using the derivatives given in reference 2 and table 2-2 and the approximate expressions presented in the appendix of reference 11.

The characteristics of these three basic configurations are in good agreement with the variation shown by the band representing the results of previous investigations. All configurations had values less than 1.0 and thus had unfavorable yaw due to aileron since all had positive effective dihedral. (See ref. 12.) In spite of the large increases in yaw- to-roll moments of inertia (3 to 4 times) of the supersonic transport compared with the subsonic transports, no unusual roll-yaw coupling effects were noticed for the super- sonic transport configurations at the approach speeds of these tests.

Reference 12 relates pilot opinion to the aileron coupling parameter Nba and the

< as shown in figure 5-16. The configurations of the present

Dutch roll damping ratio investigation shown in the figure agree fairly well with the data of reference 12. Again, all the values of the aileron coupling parameter N i a a r e positive and indicate adverse aileron yaw.

Roll-response characteristics. - _ - - The variation of pilot rating with roll-time con-

- stant as shown in reference 4 is presented in figure 5-10 and compared with the corre- sponding characteristics for the supersonic transport configuration of this investigation.

The two curves in figure 5-10 represent fairings of test points from several investiga- tions using ground-based simulators and from flight tests. Although the curves repre- sent results for fighter and reentry vehicles, the data of this investigation a r e generally in agreement with the trends of figure 5-10. The pilot evaluation-rating number for the variable-geometry (emergency landing) configuration appears to be high according to the criteria suggested by the reference curves, but the evaluation was influenced not only by as previously discussed, and by the long roll-time constant but also by the low roll power, the adverse aileron yaw characteristics indicated in figures 5-15 and 5-16.

The roll-response characteristics of both the basic fixed-geometry and variable- geometry configurations were considered to be good and typical roll-response records a r e shown in figure 5-17. The variable-geometry curve w a s taken directly from test records and the fixed-geometry curve was extrapolated from flight records of a response to a 10' wheel input because the simulation for this case w a s only valid up to wheel dis- placements of 15'. Examination of the roll responses of these configurations shows that the roll performance of these configurations exceeded the minimums indicated in refer- ence 4 from an analysis of available data for large airplanes in approach conditions.

The two criteria offered are: (1) the time to bank to 30' of about 3.5 seconds seems to be the maximum acceptable and values below about 3.0 seconds a r e considered satisfactory; and (2) that the minimum acceptable roll rate can apparently be as low as 12' per second.

The basic variable-geometry configuration should be considered satisfactory since it reached a bank angle of 30' in 3.0 seconds. The basic fixed-geometry configuration required a little longer, 3.2 seconds, to reach a bank angle of 30' and would be acceptable and also not f a r from satisfactory. The roll rates of 13' per second for the variable- geometry configuration and 20' per second for the fixed-geometry configuration shown in figure 5-17 are both above the minimum acceptable rate of 12' per second and it should be pointed out that the actual supersonic transport airplane will probably have more con- trol power available than that provided by the 30' wheel deflection used to determine the roll rates for the simulated airplanes.

CONCLUDING REMARKS The lateral-directional results of the investigation of the low-speed handling qual- ities of three supersonic transport configurations are summarized by configuration.

Variable -Geometry Configuration The lateral-directional characteristics of the basic variable -geometry configuration were good and were characterized by adequate Dutch roll damping, good directional sta- bility, positive effective dihedral, good roll response, roll damping, heading-change pre- cision, and low workload; the relatively long period of the Dutch roll oscillation did not appear to be objectionable to the pilots. The basic variable-geometry configuration w a s given an average Cooper pilot opinion rating of 3.0.

The lateral-directional characteristics of the variable -geometry configuration were degraded to determine the effect of such characteristics on the handling qualities of air- planes the size of these supersonic transport configurations. The characteristics were degraded by reducing the Dutch roll damping ratio from 0.18 for the basic configuration to 0.05 and increasing the adverse yaw due to rolling velocity. The principal results of the degraded characteristics were a moderate increase in workload, especially near touchdown, and a reduction in heading-change precision. This configuration w a s given an average Cooper rating of 4.5 compared with 3.0 for the basic configuration.

Variable -Geometry (Emergency Landing) Configuration The lateral-directional characteristics of the variable -geometry (emergency landing) configuration were unsatisfactory because of weak roll control, large positive dihedral effect, and low Dutch roll damping. This configuration was, however, considered to be acceptable for emergency operation and was given a Cooper rating of 5.5.

Fixed-Geometry Configuration The later a1 -directional character istic s of the basic f ixed- ge ometry configuration were good and were characterized by good Dutch roll damping, good directional stability, positive effective dihedral, and low workload in the approach although the damping in roll w a s low and a little extra c a r e w a s required t o make precise heading changes. This configuration w a s given an average Cooper rating of 3.5.

The lateral-directional augmentation consisted of a 50-percent increase in the t o eliminate the tendency to overshoot or undershoot damping -in-roll parameter

C Z 4

in turn maneuvers. The augmented fixed-geometry configuration w a s given an average Cooper rating of 2.8 on the basis of reduction in effort required to make precise heading changes.

The lateral-directional characteristics of the fixed-geometry configuration were degraded in the same way and to the same degree as w e r e those of the variable-geometry configuration with essentially the same effect on the handling qualities, namely, a mod- erate increase in workload over the basic configuration and a reduction in heading- change precision. This configuration was given an average Cooper rating of 4.5 compared with 3.5 for the basic fixed-geometry configuration.

REFERENCES Features of Large Transport Aircraft Affecting Control During 1. Pinsker, W. J. G.: Approach and Landing. AGARD Rept. 421, Jan. 1963.

Simulation of 2. Eldridge, W. M.; Condit, P. M.; Schwanz, R. C.; and Taylor, C. R.: Three Supersonic Transport Configurations With the Boeing 367 -80 In-Flight Dynamic Simulation Airplane. No. D6-10743 (Contract No. NAS1-4096), The Boeing Co., 1965. (NASA CR-66125.)

3. Mabli, R. A.; Carlson, J. W.; and Sickeler, R. 0.: Results of Handling Qualities Research for the C-5A. AIAA Paper No. 65-740, Am. Inst. Aeron. Astronaut., Nov. 1965.

4. Ashkenas, I. L.: A Study of Conventional Airplane Handling Qualities Requirements.

Roll Handling Qualities. AFFDL-TR-65-138, U.S. Air Force, Nov. 1965.

Part I.

Flying Qualities of Piloted Airplanes. Mil Specification F-8785 (ASG), 5. Anon.: Sept. 1, 1954.

6. Stapleford, Robert L.; Johnston, Donald E.; Teper, Gary L.; and Weir, David H.: Development of Satisfactory Lateral-Directional Handling Qualities in Landing Approach. NASA CR-239, 1965.

7. Ashkenas, I. L.: A Consolidation of Lateral-Directional Handling Qualities. AIAA Paper no. 65-314, Am. Inst. Aeron. Astronaut., July 1965.

8. Quigley, Hervey C.; Vomaske, Richard F.; and Innis, Robert C.: Lateral-Directional Augmentation Criteria for Jet Swept-Wing Transport Airplanes Operating at STOL Airspeeds. Conference on V/STOL and STOL Aircraft, NASA SP-116, 1966, pp. 295-310.

9. Ashkenas, I. L.: A Study of Conventional Airplane Handling Qualities Requirements.

Part 11.- Lateral-Directional Oscillatory Handling Qualities. AFFDL-TR-65-138, U.S. Air Force, Nov., 1965.

10. Bisgood, P. L.: A Review of Recent Handling Qualities Research and Its Application

t o Handling Problems of Large Aircraft. Part I - Observations on Handling

Problems and Their Study. Part I1 - Lateral-Directional Handling. Report

Aero 2688, Brit. R. A. E., June 1964.

11. Newell, F. D.: Criteria for Acceptable Representation of Airplane Dynamic Responses in Simulators Used for Pilot Training. NAVTRADEVCEN 1146-1, U . S . Navy, Oct.

1962.

The Effect of 12. Vomaske, Richard F.; Sadoff, Melvin; and Drinkwater, Fred J., 111: Lateral-Directional Control Coupling on Pilot Control of an Airplane As Determined in Flight and in a Fixed-Base Flight Simulator. NASA TN D-1141, 1961: I TABLE 5-1.- SUMMARY O F FLIGHT-TEST CONFIGURATIONS AND RESULTS -- -- Lateral-directional pilot ratings Barameters Pilot's comments Pilot (Cooper scale) varied ___ -- -~ Variable-geometry configuration; r = 0.18 - - - ~ - - __-- __I- ~ - .

'n. = o A 3 . 0 Stability: P (a) Dutch roll oscillation seems to be mostly in yaw with little noticeable roll; appeared t o dam] ' = -0.022 in 1 to 11. cycles. No tendency to be excited in normal maneuvers.

"4 (b) Spiral stability seemed neutral.

(c) Positive directional stability.

Maneuverability: Roll damping high.

(a) Roll response is excellent.

(b) Heading-change precision within about lo. Slight headirig lag at low r a t e s of roll became noticeably large at high r a t e s of roll.

Work level in the approach and flare is low.

Stability: B 3 . 0 (a) Dutch roll damping is very good, appears to be almost all yaw and little roll, 6/@ = 3 to 4; damped in 1 to 1 1 cycles. No tendency to be excited in normal maneuvers.

(b) Spiral stability was neutral or slightly divergent.

(c) Very good directional stability and positive effective dihedral.

Maneuverability: (a) Roll rate very good; maximum wheel not used because r a t e s available with small inputs were adequate. Roll damping quite acceptable.

(b) Heading-change precision acceptable, about 1 ' to 2 ' . Slight heading lag noted. High adverse sideslip noted in turns.

Work level in the approach and flare is minor.

C 2 . 0 Stability: (a) Dutch roll damping very good; rolling is the predominant motion but is not excessive.

Tendency toward excitation in normal maneuvers is very small.

(b) Spiral stability not noted.

(c) Good directional stability and effective dihedral appeared positive and normal; not easy to keep wings level with rudder alone because o f long response time.

Maneuverability: (a) Roll response appears to be good with no detectable adverse yaw but some adverse side- slip; however, the behavior of the airplane following development of sideslip was good.'

Roll damping appeared tobe good.

(b) Heading-change precision was good; the adverse sideslip was no problem.

Work level in the approach and flare low.

. __ . _ _ - ~. ~ . -~ - - ~- - TABLE 5-1.- SUMMARY O F FLIGHT-TEST CONFIGURATIONS AND RESULTS -Continued Lateral-directional pilot ratings Parameters Pilot (Cooper scale) Pilot's comments varied Overall Variable-geometry configuration; < = 0.18 2 n . = o D 3 . 0 Stability: B (a) Dutch roll damping appeared to be marginal because, although the oscillation damps to l e s s :"$= -0.022 than one-half amplitude in 1 cycle, the period was relatively long (10 seconds) so the time to damp was longer than desired. The motion was predominantly yawing. There was a tendency to excite an oscillation but the pilot was relatively unconscious of it because of the long period and lack o f any side force in the cockpit.

(b) Directional stability was satisfactory and the effective dihedral was mildly positive and satisfactory.

Maneuverability: (a) Roll response: initial response was satisfactory and rate of roll was good. Roll damping was satisfactory.

(b) Heading-change precision was good; no appreciable heading lag although there was a definitc tendency to sideslip in maneuvers.

Work level in approach and flare was low.

E 3 . 0 Stability: (a) Dutch roll damping was good with a sideslip-to-roll ratio o f about 2. There was no tendenc to excite the oscillation in normal maneuvers.

(b) Spiral stability was neutral.

(c) Directional stability was fair to good and the effective dihedral was positive.

Maneuverability: (a) Roll response: initial response was good and rate o f roll satisfactory. Roll damping was fair to good.

@) Heading-change precision was within about 11". There appears to be no adverse heading change on the turn indicator but adverse sideslip noted on sideslip indicator.

Work level in the approach and flare was normal.

F 3 . 0 Stability: (a) Dutch roll damping was almost deadbeat, damping completely in l e s s than 1 cycle. The sideslip-to-roll ratio was about 1 and the roll lagged the yaw by 2 to 3 seconds. No tendenc to excite the oscillation in the very smooth air encountered on this flight.

(b) Directional stability: returned from 1 0 ' to trim properly; dihedral effects were positive to 1 0 ' o f sideslip but preferred l e s s dihedral than airplane bad.

Maneuverability: (a) Roll response: initial response was 6 ' to 7 ' in the first second and the rate of roll seemed to be at least 20' to 2 5 O per second. Roll damping permitted roll to precise bank angles.

(b) Heading-change precision was satisfactory; tendency to sideslip was relatively large but rudder response was so slow that no rudder given to coordinate turns. Slightly objection- able but did not seem to interfere with other tasks.

Work level in the approach and flare was very low laterally and none on the rudder.

--,. .... . .,, .. ... , .

TABLE 5-1.- SUMMARY OF FLIGHT-TEST CONFIGURATIONS AND RESULTS -Continued Lateral-directional

1 1 pilot ratings 1

Parameters Pilot (Cooper scale) Pilot's comments varied Variable-geometry configuration; 5 = 0.18 3 . 0 Stability: (a) Dutch roll dampilg was very good, 2 cycles to damp to zero roll rate.

Perceptible motion is primarily rolling and sideslip noted from sideslip indicator only. The tendency to excite the oscillation in normal maneuvers was very slight.

(b) Directional stability was low and the effective dihedral was satisfactory.

Maneuverability: (a) Roll response felt to be too high initially (roll control forces a r e a little low and the roll acceleration was too high for small aileron inputs). Care required to fly the airplane smoothly on turn entries. Some tendency to sideslip in maneuvers. Roll damping was good.

(b) Heading-change precision: wershoot or undershoot was the order of lo, Z O , or 3 ' depending on the turn entry rate and roll out rate as a result of the sideslip generated.

Work level is normal in the Fpproach but fairly high in the flare because of the high wheel activity resulting from overcontrolling.

-- - - . - _ _ Augmented variable-geometry configuration; improved Dutch roll damping: 5 = 0.28 - - - .- .

_ _ A Stability: 3 . 5

% = o.086

(a) Dutch roll damping was very high, oscillation disappeared in 1 cycle. The motion appears to be mostly yawing. No apparent tendency to excite the oscillation in normal maneuvers.

% = -o*022

(b) Spiral stability was neutral.

(c) Directional stability apparently lower than for basic configuration because of rather sluggish return from steady sideslip. Dihedral effect was positive.

Maneuverability: (a) Roll response was quick with a high initial rate and good roll damping.

(b) Heading changes a r e slower than f o r the basic configuration because of larger initial adversc sideslip and heading lag. Precision was good. Trace of adverse lateral acceleration in turn entries.

Work level was low with no appreciable difference from the unaugmented configuration. Lift-drag controllability h a s been downgraded because of the increased sideslip on turn entries and trace of adverse lateral acceleration. Tendency to fly sideslipped several degrees.

B 3 . 0 Could not tell any appreciable difference between this configuration and the basic configuration; especially during approach could not tell any difference.

C 2.5 stability: (a) Dutch roll damping was very good with the same roll-to-yaw ratio as f o r the basic configu- ration (predominantly roll). Tendency to excite the oscillation in normal maneuvers is very little.

(b) Directional stability and effective dihedral were both good; comment on holding wings level with rudder alone same as for 20' basic configuration.

Vlaneuverability : (a) RoII response was good both initial rate and maximum rate. The roll damping was good.

@) Heading-change response was good but not quite as good as for the basic configuration because there was a little tendency toward motions of the airplane in t e r m s of residual oscillation.

rhe work level in the approach was low.

_ - TABLE 5-1.- SUMMARY O F FLIGHT-TEST CONFIGURATIONS AND RESULTS - Continued ~ ~~~ Lateral-directional (Cooper scale) Pilot's comments = 0.05 Degraded variable-geometry configuration; low Dutch roll damping and higher adverse yaw; 0 to 4.5 Stability: 4.0 (a) Dutch roll damping was very low or neutral; oscillation appeared to have no damping; the rolling motion could be damped in about 1 1 cycles with lateral control and the yawing motion would damp in about 2 cycles more with control fixed. Initially, the roll-to-yaw ratio was about 2:l or 3:l decreasing to about 1:l after 3 Cycles. The oscillation was excited by any abrupt wheel input or by any wheel input of greater than 5 ' .

@) Spiral stability appeared to be neutral.

(c) Directional stability was somewhat masked by the Dutch roll oscillation but appeared to be no1 as stable as basic configuration. Dihedral effect was positive.

Maneuverability: (a) Roll response: initial rate is adequate with no apparent lag in buildup of roll rate and apparent small time constant. A very pronounced heading lag noted. The maximum roll rat€ oscillates because of the sideslip and dihedral effect.

@) Heading changes a r e difficult to make because of the heading lag (about 2 seconds) and the precision is only about 3 ' to 4 ' . Difficult to coordinate w i t h rudder, also.

Work level not found to be high in the approach, although not as low as for basic variable-sweep configuration.

4.5 1.5 to 5.0 Stability: (a) Dutch roll damping noted to be low but stable, with a roll-to-yaw ratio of about 2:l. The oscillation was easily excited in normal maneuvers and seemed to be present almost all the time.

(b) Spiral stability seemed to be neutral.

(c) It was fairly difficult to hold steady sideslip because of the Dutch roll oscillation. The dihedral effect seemed quite positive.

Maneuverability: (a) Roll response: the initial rate seemed to lag a bit; 2 0 ' wheel gives all the roll rate required.

Slight adverse yaw but quite a bit of adverse sideslip.

(b) Heading changes were very difficult t o make more precisely than 3 ' to 5 ' in a rapid turn; rudder was used but was difficult to coordinate.

Work level is very low at the start of the approach letting the airplane oscillate until close to the Work load increases accordingly and is greater than for the basic configuration ground to go VFR.

I _. . .. .

TABLE 5-1.- SUMMARY OF FLIGHT-TEST CONFIGURATIONS AND RESULTS - Continued ~- - Lateral-directional Parameters Pilot's comments varied Variable-geometry configuration (emergency landing); 5 = 0 . 1 7 5.5 4.5 Stability: (a) Dutch roll damping was low and the period was short enough for a tendency to sustain the oscillation rather than damp it. It is chiefly a rolling oscillation which is excited by almost any external disturbance or control input. Hard to estimate roll damping because o f Dutch roll presence all the time.

@) The spiral mode was extremely stable and because of this there was no divergence as the roll angle was always around zero in spite of the low Dutch roll damping.

(c) Directional stability seemed fairly high and there was a large positive dihedral effect.

Maneuverability: (a) Roll response to wheel was sluggish and oscillatory because of the influence o f the Dutch rol and because the strong spiral stability arrested the roll rate; continuous wheel was required to maintain a desired bank angle.

(b) Heading response: some heading lag noted the precision low because Dutch roll so easily excited and limits accuracy to 2' to 3 ' . The time for completion of a turn is longer than that for the basic variable-geometry configuration but not unacceptable.

(c) Because of the high effective dihedral, rudder generated a high roll rate which could be extremely dangerous near the ground in decrabbing from sideslip in a cross-wind landing.

Work level was high.

- ~ ~ .. ~ _ _ -.

Fixed-geometry configuration; C = 0.38 - ~ . - ~. . .

~ ~ ~.

3.5 2.5 Stability: ! z i = -0.04 (a) Dutch roll was well damped with a roll-to-yaw ratio of about 2 : l and no apparent tendency to !n. = -0.00 6 be excited in normal maneuvers.

' = o @) Spiral stability was positive.

"B (c) Directional stability and dihedral effect a r e positive.

Maneuverability : (a) Roll response: initial response is high, almost too sensitive in roll response; roll control power also is high. Roll damping is low and requires Some attention to control; tendency to overshoot.

@) Heading changes: small adverse sideslip.

Work level is low in the approach.

3.5 2 . 5 to 3.0 Stability: (a) Dutch roll damping did not seem as high as for the basic variable-geometry configuration.

The roII-to-yaw ratio was about 1 : l . There was a tendency toward excitation of the oscilla- tion in normal maneuvers.

@) Spiral mode seemed to be neutral or slightly divergent.

(c) Directional stability and effective dihedral were positive.

Maneuverability: (a) Roll response: initial response was good but roll damping was low.

@) Heading changes: small adverse yaw in turn entries.

Work level is low i n the approach.

- . . ~ TABLE 5-1.- SUMMARY OF FLIGHT-TEST CONFIGURATIONS AND RESULTS -Continued Lateral-directional Parameters Pilot's comments varied Fixed-geometry configuration; = 0.38 Stability: 3 . 5 : L 4 = -0-044 (a) Dutch roll damping was good with a fairly high roll-to-yaw ratio. Some tendency to over- control in roll on turn entries and recoveries.

= -0-005 (b) Spiral stability was positive and good.

(c) Directional stability seemed good and dihedral was positive to extent that it waa difficult to prevent an oscillation using rudder alone to keep wings level.

Maneuverability: (a) Roll response: initial response was adequate and maximum rate seemed higher than needed, but not objectionable. Roll damping seemed a little low.

(b) Heading change: adverse sideslip caused heading lag following bank but could be coordinated with a little rudder. Precision downrated a little because of tendency to overshoot or under- shoot in roll.

Work level was moderate in the approach.

3.5 Stability: (a) Dutch roll damping was good. Initially the motion is yawing and the roll-to-yaw ratio appear€ The rolling motion lags the yawing motion. There was a minor tendency to excite to be low.

the Dutch roll in normal maneuvers.

(b) Directional stability was good and the effective dihedral positive and higher than desired.

Maneuverability: The rate of roll (a) Roll response: initial response was very good but sensitivity was too high.

was very good but the roll damping was lower than desirable, but did not seem to cause any real problem. There was a mild adverse yaw.

(b) Heading-change response appeared satisfactory although the high lateral sensitivity was adverse. Some oscillatory tendencies during rapid entries and in roll-outs, probably due to high lateral sensitivity and high dihedral effect.

Work level in approach estimated to be satisfactory but no landings actually made because of high level o f turbulence.

3.5 Stability: (a) Dutch roll damping was fair and seemed to damp to 3 to 4 cycles; roll displacement is more apparent than sideslip, particularly in cross-wind landing, with a tendency to set up pilot- induced oscillation.

(b) Spiral stability was neutral and that was desirable.

(c) Directional stability was very good and the effective dihedral was positive.

Maneuverability: (a) Roll response: initial response was rapid and roll rate was excellent. Also noted that the roll-generating capability using rudder was good.

(b) Heading changes: a definite heading lag was noted - sizable roll angle established before airplane starts b n i n g . A large longitudinal deceleration noted in turns.

Work level was l e s s under the hood than visual, probably because of greater attention to the instruments.

TABLE 5-1.- SUMMARY OF FLIGHT-TEST CONFIGURATIONS AND RESULTS -Continued Lateral-directional pilot ratings Parameters Pilot's comments (Cooper scale) ?ilOt varied GiZ+G roll1 Fixed-geometry configuration; = 0 . 3 8 F 3 . 0 Stability: 3L i = -Om4 (a) Dutch roll damping appears to be deadbeat with a roll-to-yaw ratio of 1.5:l and did not notice very much lag between roll and yaw as had for the basic variable sweep. No tendency to

% = -O.OO5

- = o

excite Dutch roll oscillation in normal maneuvers.

%,j (b) Spiral stability was neutral in one direction and positive in the other indicating slightly positive.

(c) Directional stabil+.tywas g o d and dihedral effect was positive.

Maneuverability: (a) Roll response: initial response was quite satisfactory and the roll rate seemed to be above 2 0 ' per second.

(b) Heading response was good in turn entries with no oscillatory tendency in turns.

Work level was low, all on wheel and none on rudder.

Stability: G 4.0 (a) Dutch roll damping - no comment because of computer malfunction at the time this was being evaluated without pilot realizing that the simulation was in error.

(b) Spiral stability w a s positive but neutral preferred.

(c) Directional stability seemed low and the dihedral was moderate.

Maneuverability: (a) Roll response: extremely responsive initially. The roll accelerations were excessive and the rate of roll was too high for normal control inputs. Roll damping was fair.

(b) Heading changes: some lag in heading response on turn entry and exit and too much sideslip generated in steady turns. No oscillatory tendency noticed. Felt that precision might be poor in rough air due to overshoot and undershoot or springback in heading after turn.

Because of very gusty air near the ground, no landings were made; one simulated VFR approach was made from 3,000 feet to 1,500 feet altitude.

..

ked-geometry augmented configuration; increased damping in roll; p = 0 . 3 8 - . _ ~ _ _ _ _ ~ . ~.

Stability : 3 . 0 (a) Dutch roll oscillation w a s well damped with a roll-to-yaw ratio of about 1:l. The oscillation wi not excited by normal controls.

(b) Spiral stability was positive with a very slow convergence rate.

(c) Directional stability is high and the effective dihedral is positive.

Maneuverability: (a) Roll response w a s good and roll damping w a s high.

The pre- (b) Heading changes: there was a smaller heading lag than for the basic configuration.

cision was gwd, within 1 ' and there was no noticeable overshoot or undershoot.

Work level was low.

.. . - _ _ ~ ~~. . . .

. .

.. ..

TABLE 5-1.- SUMMARY OF FLIGHT-TEST CONFIGURATIONSAND RESULTS - Continued Lateral-directional Parameters Pilot Pilot's comments varied Fixed-geometry augmented configuration; increased damping in roll; < = 0 . 3 8 ~- B 2 . 5 2.0 to 2 . 5 Stability: :1 4 = -0.07a (a) Dutch roll damping was good with a roll-to-yaw ratio of 3 or 4 : l . No tendency noticed to exciting the oscillation by normal maneuvering control inputs.

:"i = - O e 0 l 5 @) Spiral stability seemed to be slightly positive.

(c) Directional stability was good and the effective dihedral was high.

Maneuverability: (a) Roll response: initial response was good and roll rate was much more than adequate. The roll damping was good.

(b) Heading changes could be made rapidly and with good precision.

Work level was very low.

C 3.5 Stability: (a) Dutch roll damping was good with a high roll to yaw ratio. There was little tendency to excite the oscillation by normal use of controls except moderate sideslip is developed.

@) Spiral stability was positive and good.

(c) Directional stability was same as for basic configuration - g o d and the effective dihedral positive.

Maneuverability: (a) Roll rate: initial rate was good; maximum rate was good; roll response seems less than for the basic configuration. Roll damping was somewhat low.

@) Heading changes were found to be slow developing because of the adverse yaw and adverse Small heading changes rated good but larger sideslip but did not pose much of a problem.

changes degraded some due to adverse yaw and tendency toward roll overshoot; requires normal use of rudder.

Work level is just moderate, use of rudder for coordination required.

TABLE 5-1.- SUMMARY O F FLIGHT-TEST CONFIGURATIONS AND RESULTS -Concluded Lateral-directional pilot ratings Parameters Pilot's comments (Cooper scale) Pilot varied ~ 1 1 1 -~ Fixed-geometry degraded configuration; lower Dutch roll damping and higher adverse yaw; 5 = 0 . 0 5 ~ _I = -0.044 Stability: 4 . 0 4.5 ' 1 4 to (a) Dutch roll damping obviously lower than for the basic configuration but can be damped by 5.0 wheel control with low level of work. No tendency for pilot-induced or sustained oscillation "4 = -0*035 in approach.

:"/3 = -0*138 @) No spiral divergence noted.

(c) No change in directional stability or effective dihedral noted from basic configuration.

Maneuverability: (a) Low roll damping apparent in tendency to bobble about a selected bank angle.

@) Heading-change precision is about +ZO because of the slight oscillation in heading on roll-outs Stability: 4.25 5.0 to (a) Dutch roll damping is low, converging a t the rate of about 1 ' per cycle. The oscillation 4.5 seemed to be predominantly yawing. There was a tendency to excite the oscillation in rapid wheel inputs but f o r small gradual inputs it was not bothersome.

(b) Spiral stability - appeared to be slightly divergent; only checked for a few seconds.

(c) Directional stability about the same as f o r the basic configuration and the effective dihedral was positive.

Maneuverability: (a) Roll rate: initial rate was very good and the maximum rate was not used but was obviously much more than required f o r normal maneuvering. Roll damping appeared to be lower than for the basic configuration making it difficult to roll rapidly and stabilize on a desired bank angle of roll rate.

@) Heading changes: despite difficulty in maldng precise heading changes at altitude, it was not very bothersome in the approaches. On rapid heading changes appreciable adverse yaw was noted.

Work level in the approach was a little higher than for the basic configuration but as the airplane approaches the ground the peripheral cues increase and the control task becomes easier.

TABLE 5-2.- SUMMARY O F THE LATERAL DIRECTIONAL STABILITY O F TEST CONFIGURATIONS AND OF THE CURRENT LARGE SUBSONIC JET TRANSPORTS Spiral Roll Dutch roll oscillation mode mode Configuration wd, rad/sec T2, sec ~ 'alculated leasured :alculate :alculated Measured Calculated deasured alculated leasured :alculatec Ieasured alculated ~ rariable geometry 5.88 5.93 1.72 241.3 0.48 9.6 10.18 0.657 0.628 0.18 0.186 1.63 rariable geometry, augmente 10.54 3.98 3.96 2.64 238.3 1 0 . 5 .622 .621 .28 2 8 2 2 . 6 6 .48 9 . 3 rariable geometry, degraded 9.08 3 7 5 .692 .051 20.6 19.63 .45 .46 275.2 .49 .05 7 . 0 pixed geometry 8.37 .995 3 1 1 .40 .381 1.74 2.24 4.02 3.74 51.9 .80 7 . 5 pixed geometry, augmented 8.18 3355 .e29 .345 .379 2.42 2.21 3.10 3.70 75.8 .57 pixed geometry, degraded 14.1 14.14 67.3 .89 6 . 4 6.40 ,983 .982 .05 .050 .45 ,453 4.5 rariable geometry 5.15 1.40 1.24 .09 ,172 5 . 5 3.25 .82 1.58 -12.3 1.7 (emergency landing) :"=rent jet transport 6.1 1.03 .07 9 . 6 .635 @) (b) (b) (b) @) (b) @) 1.6 .80 .06 14.5 (b) .525 (b) (b) :urrent jet transport B~ (b) (b) (b) (b) aUnaugmented.

NO^ available.

. . . . .

-. ,..-. -

.06 + .04

a

a " , .02 0 0 -.02 -1 +) %

s

-.Ob -2 -.06 M a" 10 -10

c

-20 -2 -4 J -6 -15 -10 -5 0 5 10 15 S i d e s l i p angle, deg (a) Variable-geometry configuration.

Figure 5-1.- Static lateral-directional stability characteristics of test configurations.

.06 .04 .02 -.02 -.Oh M a M ., 2 20 .rl * .rl v) R -20 c, %I

2 -40

k c, -60 V -80 M a ., a , d

-4 -. 1

rl d ffi l ! I I l I l l 1 1 -6 -5 - 4 - 3 -2 -1 O 1 2 3 4 5 6 Sideslip angle, deg (b) Variable-geometry (emergency landing) configuration.

Figure 5 - 1 . - Continued.

I

- ....... ._ ..

.06 .Oh

p

. d .02 a; cl -.02 -1 -.Oh -2 -.06 %I 2 10

5 -10

J -20 /..ld-..

............ ......... ".- .

. . I . . . . . ..I J 1 0 15 -10 -5 0 5 -15 S i d e s l i p angle, deg (c) Fixed-geometry configuration.

Figure 5-1.- Concluded.

a C -A -L 1 ° h I I I * I I . I _-I- 0 5 10 15 20 25 30 35 Time. sec (a) Variable-geometry configuration.

Figure 5-2.- Dutch roll characteristics of test configurations.

13 1 4 r

-2 L

2 - 1 - Y, deglsec 0.

- -1

- 2 0 r

I I I I I I I 0 5 10 15 20 25 30 Time, sec (b) Variable-geometry augmented configuration.

Figure 5-2.- Continued.

6 - 4 - 2 -

-

0 - - -2 - -4 - -6 _I -1 --I--.l -..I. - . I I 0 5 1 0 1 5 20 25 30 Time. sec ( c ) Variable-geometry degraded configuration.

Figure 5-2.- Continued.

I- -

Br, deg I I I I I I I 0 5 10 15 20 25 30 Time, sec (d) Variable-geometry (emergency landing) configuration.

Figure 5-2.- Continued.

ib I

-1 L I 1 1 I I I 0 5 10 15 m 25 Time, sec (e) Fixed-geometry configuration.

Figure 5-2.- Continued.

13 5 I

lllll1ll1 l l l 1 l l 1 1 1 1 1 1 1 1 1 l 1 l 1 l 1 1 I 1 I I I 1

r

3 r -1 I . 4 . -d(_--. I I I 5 IO 1 5 20 25 30 Time, sec (f) Fixed-geometry augmented configuration.

Figure 5-2.- Continued.

4, deg/sec

4, degfsec

-3 t

5 r

Time, sec (g) Fixed geometry, degraded configuration.

Figure 5-2.- Conc~udeo.

. . . __I_- - . . .

-30

I I I - 1 1 . I . ... 1

-40 -30 -20 -io o io 20 30 40 L e f t Right Contra1 wheel p a s i t i o n , deg (a) Basic variable-geometry configuration.

Figure 5-3.- Lateral control characteristics of the test configuration.

--- C s l c u l a t ed

0 Flight t e s t V a , U 2 10 .--.

a ’

ho a ,

I 6k0

e .\ 0 a , E l l ’ % k

0 8 ’

rl -1c rl I ffi -2c 1 - ~ I 1 1 ~ I I J -30 -20 -10 0 10 20 30 L e f t Right Control wheel p o s i t i o n , deg (b) Degraded variable-geometry configuration.

Figure 5-3.- Continued.

1 1 l l 1 1 1 l I I I II I1 1lIIIll1111lIIIlIIIlIIIlIIIIllll I I 1 I1 I I I I 1 I1 I I I .\ a, -P cd k rl -10 rl ffi -2 0 -3 0 Tu

r

E f 0

.rl

*

cd k rY -10 a, : -2 0

i 20 I - 30 I 4 ’ 6 1 50 I

-3 0

-50 -40 -30 -20 -io 10

Left Right C o n t r o l wheel position, deg (c) Variable-geometry (emergency landing) configuration.

Figure 5-3.- Continued.

I

Left R i g h t Control wheel position, deg (d) Basic fixed-geometry configuration..

Figure 5-3.- Continued.

14 1 I al 20 -r2 M a , .d

. . 10

-10 -2 0 -3 0

-30 -20 -10 0 io 20 30 40

-40

L e f t Right Control wheel position, deg (e) Fixed-geometry augmented configuration.

Figure 5-3.- Continued.

I , - 1 1 1 1 1 I, I n 1--.11111111111111 I II 1 . 1 1 1 . 1 1 1 1 1 1 1 1111111111 I II I111 I I 1 . 1 1 1 1 11111=1 1-11

I

---

Ccllculated & F l i g h t t e s t

-

-

-

0 -

-

-10

-

-20

-

-30 I 1 I I 1 I I -30 -20 -10 0 10 20 30 Left Right Control wheel p o s i t i o n , deg (f) Degraded fixed-geometry configuration.

Figure 5-3.- Concluded.

I -P m Emergency landing k a , rl 50 N a - 0

.d C d *

0 0 .rl

J u cog -5O

4 1 I 1 I I I I I I I 8 I 0 10 20

30 40 50 60 70 80 90 loo iio 120

Time, sec (a) Localizer command signal.

Figure 5-4.- Typical landing-approach time histories of the variable-geometry configurations, * o ) Basic 0 , .

-20 2 0 . Augmented A -20 ' 20 Degraded

t

-20 '

. Emergency landing

1 I I I I L I t 1 I I I

60 70 80 90 100 110 120 0 1 0 20 30 40 50

Time, see (b) Control wheel position.

Figure 5-4.- Continued.

)--r b P r l ; Q) cd .d ,Basic .01 a 1 - - * 0 0 , -1 .01

Augmented 1

- -.01

t

L d .rl a

-

-. --

- . L o 0 ’

I

-1 - --.01

lbEmergency landing

- -

w - ~~~ V j o -

) -

-1

- -.01

.

I 1 I 1 I I I 1 I 1 I I 0 1 0

20 30 40 5 0 60

90 1 0 0 110 120 70 80 Time, s e e ( c ) Rudder pedal position.

Figure 5-4.- Continued.

rl a, 1 0 1 Basic

4 1 0 1 ’ Augmented

cdM o * - - - - v -A

rl.3

rl -10’ pi ffi I I I I I I I A I 1 I I I

& -

0 10 20 30 40 50 60 70 80 90 100 110 120 Time, sec (d) Roll angle.

Figure 5-4.- Continued.

CI a , .\ a J ri

2)

cd ,Augmented c4 .rl M r l a , m - a , rl Emergency landing c d Pi M daJ m a a , d .rl u1 I I 1 1 I I I I I I I 1 - 0 10 20 30 40 50 60 70 80 90 100 iio 120 Time, see (e) Sideslip angle.

Figure 5-4.- Concluded.

(0 k .P Basic a , N a - 0 I I I I I 1 I I I I I I 0 1 0 20 30 40 50 60 70 PO 90 100 110 120 130 140 150 Time, see (a) Localizer command signal.

Figure 5-5.- Typical landing-approach time histories of t h e fixed-geometry configurations.

Time, see (b) Control wheel position.

Figure 5-5.- Continued.

Basic . 0 1 E.

a] 1 , a L+o ( U T

7 r . y

A n n

- -- 0 “ E g

z.2 0 -

a J 4 J a -11 -.01 I 3 K a Augmented . 0 1 s g.

P Y K s ? 1 ’ - k . 4 0 , 0 P p I a .rl I3a

2: -lB

-.01 K 0 1 0 20 30 40 50 60 70 80 90 1 0 0 110 1 2 0 130 1 4 0 150 Time, s e c (c) Rudder pedal position.

Figure 5-5.- Continued.

(d) ~ o l l angle.

Figure 5-5.- Continued.

M 5 1 Degraded o , - A m a l w - - b / v a,d 28 - 5 , mtd I I I 1 1 1 I I I I I I I 0 1 0 20 30 40 50 60 70 80 90 100 110 120 130 140 150 T i m e , sec (e) Sideslip angle.

Figure 5-5.- Concluded.

IIIII

$ 1 . 0

$

k 2.4 w 0 e 2 . 1 r l d +I, - 5 d 0 .1 .2 .3 Dutch roll damping r a t i o , 5 Basic variable geometry Augmented variable geametry €I3 Degrade$ variable geometry

Current large j e t transports - unaugmented

3 - @ 2.1 2 '

0 2.4

1 -

I _1 .: -... ;... I ... J

.1 .2 .3 -4 deg sec

Rolling parameter, @/ve, -

f t 1 I . I - t - 1 0 * 33 .66 .99 1.32 deg sec Rolling parameter, @/ve, Figure 5-6.- Dutch roll oscillation characteristics of t h e variable-geometry supersonic transport configurations. Numbers adjacent t o symbols a r e t h e pilot rating (based o n t h e Dutch roll characteristics) for t h e particular configuration.

V

-

W 1.5 -?

cd El 4.5 k 0 .1 - 2 Dutch r o l l damping r a t i o , 5 0 Variable geometry (emergency landing)

Current large j e t transports - unaugmented

- 4.5

Cl - P I 1. ~ 1 - 1 I I'

.1 .2 - 3 .4 .5

Rolling parameter, $/ve, deg sec ft I I 1 I I 1 0 33 .66 99 1.32 1.65 Rolling parameter, @/v,, deg m Figure 5-7.- Dutch roll oscillation characteristics of the variable-geometry (emergency landing) configuration.

Numbers adjacent to the symbols are the pilot rating (based on the Dutch roll characteristics) for the configuration.

Basic v a r i a b l e geometry

h o - _ - _ _ Variable geometry (emergency landing)

a "

- 4 cd Pi .rl d

/--- ------- 1 LzLA/--

m

.- _ _ c e '

aJ a *rl m n M d .r(

x

0 2 4 6 8 10 12 14 m R Time, sec (a) Wheel step.

Figure 5-8.- Comparison of the responses of the basic variable-geometry and of the variable-geometry (emergency landing) configurations to lateral and directional control from flight records.

Basic variable geometry - - - - Variable geometry (emergency landing) k a ,

a l'F /-\

k a , a a h o

Z %

L I I - I 1 _ _ I 1 - 1 I 0 2 4 6 1 0 12 14 16 1 8 i 2 : 4 Time, sec (b) Rudder pulse.

Figure 5-8.- Concluded.

!..

A Basic fixed geometry A Augmented fixed geometry A Degraded fixed geometry

Current large j e t transports - unaugmented

a

4.5 A 2.6 A 1 2 . 4 .5

I

0 1 I I I -1 .1 .2 .3 .4 Dutch roll damping r a t i o , 5 4 - A 1 2 . 6 A 2.4 3 -

-

c, - 2 - 1 - I I I I I I 0 .1 .2 .3 .4 .5 Rolling parameter, $/ve, deg ft I 1 I I I ~- .L 0 .33 66 .99 1.32 1.65 Rolling parameter, $/ve, deg sec Figure 5-9.- Dutch roll oscillation characteristics of the fixed-geometry configurations. Numbers adjacent to symbols are the pilot rating (based on the Dutch roll characteristics) for the particular configuration.

Basic variable ge3metry Basic fixed geametry

A

A Augmented fixed g e m e t r y

0 Variable geDmetry (emergency landing)

- - -

Reference 4 (applicable t o configurations having good Dutch roll damping and low e m i t a t ion by ailerons ) P i l o t . r a t i n g

A

/-

I I I 1 1 1 I I I I I 1 1 1 4 I I I 1 1 1 I I I I I 1 1 1 4 1 t -I..- - 1 ~

.1 .2 - 5 1.0 2. 3 . 4.5. 10. .1 .2 - 5 1.0 2. 3 . 4.5. 10.

TRS sec Figure 5-10.- Variation of pilot rating with roll time constant.

Tolerable

----

Unacceptable ~ ~ l ~ unacceptable

I I 1 I I 0 .1 .2 .3 .4 .5 .6 deg sec Rolling parameter, $ / ve, ft .

I I I 1 I I I .99 1.32 1.65 1.98

0 - 33

deg sec Rolling parameter, $ / ve, Figure 5-11.- Comparison of the Dutch roll characteristics of the test configurations with the existing military specifications. Numbers adjacent to the symbols refer to the pilot rating (based on the Dutch roll characteristics) for the particular configuration.

. - I Basic variable geametry Augmented variable ge ome t r y

e 3

Degraded variable geametry Basic fixed geometry Augmented fixed geometry

Degraded fixed geametry H

Variable geometry (emergency landing)

Current large j e t transports - unaugmented

I

I

I

Acceptable -

\

damper off

\

\ Satisfactmy Unacceptable . _- I - 1 I I I .1 .2 .3 .4 .5 Dutch roll damping r a t i o , 5 Figure 5-12.- Comparison of t h e Dutch roll characteristics of the test configurations with the prcposed revised military specifications.

Numbers adjacent to t h e symbols refer to the pilot rating (based on the Dutch roll characteristics) for the particular configuration.

Basic v a r i a b l e geometry Augmented v a r i a b l e geometry Degraded v a r i a b l e geometry Basic f i x e d geometry Augmented f i x e d geometry Degraded f i x e d geometry Variable geometry (emergency landing) Reference 7 (for low Dutch r o l l frequencies) . . . .

. . . . .

. . . . . .

. . . . .

. . . . . .

. . . . .

. . . . . .

Reference 8 (simulator r e s u l t s ) . . .

........

...........

k, .. ..

...

....

P i l o t r a t i n g 0 .1 .2 .3 .4 rad - se c Figure 5-13.- Variation of pilot rating w i t h D u t c h r o l l viscous damping parameter.

Basic v a r i a b l e geDmetry Augmented variable ge 3me t ry (33 Degraded variable ge omet ry

e

Basic fixed geametry

a

Augmented fixed geDmetry

A

Degraded fixed geometry

A

Variable g e m e t r y (emergency landing) S a t i s f a c t ory 5 1 0 20 m 20 1 0 5 - T ,

T2 -

Unstable Stable Figure 5-14.- Comparison of the calculated spiral stability characteristics with the criteria presented in reference 10.

1 0

0 Basic v a r i a b l e geometry

Basic f i x e d geometry

A

0 . Variable geometry (emergency landing)

...--.*...- ,.........

Rcference 7 ( 0 . 0 8 4 6K0.17) ..........

-......

I

.4 .5 .6 .7 .8 .g 1.0 1.1 1.2 1 . 3 1.4 1.5

Figure 5-15.- Variation of pilot rating with r o l l coupling parameter.

~2.b' constant p i l o t rating l i n e from reference 12 ' 7 , E z z z S E -

- ____ - - ' !

-2 _ _ - ~ L

-____ - -

\ , , @ Augmented variable geometry - -4 1 I - 8 Degraded variable geometry \\ I I I A Basic fixed geometry (emergency landing ) -6 -

-.3 -.2 -. 1 0 .1 .2 .3 .4 .5

Figure 5-16.- Variation of pilot rating with aileron coupling parameter and Dutch r o l l damping ratio. Numbers adjacent t o symbols refer to pilot rating for t h e particular configuration.

I - - - - - Variable geometry Fixed geometry I -10 -2 0 -3 0 -4 0 Time, sec Figure 5-17.- Typical roll responses for basic fixed- and variable-geometry configurations.

6. AN EVALUATION OF PILOT WORKLOAD By Samuel A. Morello and Albert W. Hall SUMMARY Correlation between pilot rating and the physical effort required to control the air- craft during a n instrument approach is presented.

The physical effort required to operate the control column was a large enough por- tion of the total longitudinal workload to be used as documentation of the pilot's rating of the longitudinal characteristics. The wheel and rudder control effort did not correlate with the pilot's opinion of the lateral-directional characteristics.

INTRODUCTION Quite often in describing various aircraft configurations and flying tasks, the pilot's evaluation is expressed as a pilot rating number, based on a system such as that described in reference 1. The pilot bases this rating on the workload or ease with which the air- craft is controlled, the precision with which the aircraft performs the task o r responds to the pilot input, a comparison of these characteristics'with those from previous experi- ence, and an extrapolation of the expected aircraft behavior in critical situations.

Pilot ratings a r e sometimes questioned because pilot opinion varies with the pilot's experience and background and with the evaluation tasks involved. It is believed that there is a need for documentation o r verification of pilot opinion, not as a substitute for, but as a supplement to, pilot rating.

In this part is presented the correlation between pilot rating and the physical work required to operate the airplane controls during the instrument approaches. In addition, the variation of flight-path deviations with pilot rating is discussed.

DATA REDUCTION Pilot Work For this evaluation, the pilot work was based on the physical definition of work

which is I F ds where F is the force and s is the distance through which the force

acts. The pilot was assumed t o be working only when he moved the controls in opposition t o the spring-loaded feel system; therefore, no work w a s being done when the springs were returning the controls to the center position. The center, or zero force control, position could be adjusted by the pilot through a simulated t r i m system s o that, after the pilot trimmed the airplane f o r the approach speed, the zero-force-control position was very close to the average position for each approach. The workload was evaluated from the data recorded during the simulated instrument approaches between the time when the airplane was well established on the glide slope and the time for initiation of the landing flare.

In a few approaches the time period during which the work was evaluated varied significantly because the variation of conditions during glide slope capture affected the time required for the airplane t o become well established on the glide slope. In order to compare the work data on an equivalent basis, the work determined for each approach was multiplied by the ratio of the time required for a typical approach (126 sec) to the actual time of the particular approach.

Column work.- The column work was determined from the time history of column angular displacement since both the control force and the distance the column traveled at the position of the pilot's hands a r e functions of control column displacement. For a given control movement, the force was taken to be the average of the control force at the initial position and the force at the final position. The control column forces for the workload computation were based on a breakout force of 4.5 pounds (20.0 newtons) in each direction and a gradient of 4 pounds (17.8 newtons) per degree of column deflection.

A control movement was defined as the motion away from the center position until the direction of motion was reversed towards the center position. After a reversal of control direction, the next motion away from the center position was treated as another control movement. The total column work for an approach was the sum of the work for each control movement.

Wheel work.- Wheel work was determined in a manner similar to the column work from the time history of wheel angular displacement. The wheel forces were based on a breakout force of 2 pounds (8.9 newtons) in each direction and a gradient of 0.16 pound (0.71 newton) per degree of wheel movement.

Rudder pedal work.- Rudder pedal work was determined in the same manner as column and wheel work except that the time history of rudder pedal displacement was expressed in inches (meters) of travel. The rudder pedal forces were based on an ll-pound (49 newton) breakout force and a gradient of 20 pounds per inch (35 newtons per centimeter) of pedal movement.

Throttle work.- ~~ Throttle work differed from the work of the other controls in that the force required to move the throttle from any position in either direction was about 1/2 pound (2.2 newtons); therefore, work was required for all throttle movement. The time history of simulated SST throttle motion (in degrees) was used to determine the The work was taken to be the product of total throttle movement during an throttle work.

approach and the 1/2-pound (2.2 newton) force.

Flight- Path Deviations Flight-path deviations were determined from data recorded by the tracking radar unit which provided ILS type of information for the landing approach tests described in part 2 of this paper. The data recording was begun when the airplane first crossed the glide slope and ended when the airplane was about 200 feet (61 meters) above the ground.

These data were used to determine the r m s deviations from the glide slope for the approaches.

RESULTS AND DISCUSSION Measurements of control motion and force a r e relatively easy to obtain during flight investigations. These quantities can be presented in various forms to represent part of the pilot effort required to perform a given task. Reference 2, for example, shows good correlation between pilot rating and total control movement during an instrument approach.

Control motion could also be represented as a root-mean-square value. The present investigation combines control motion and control force to give a measure of pilot effort in t e r m s of work in foot-pounds (newton-meters). An indication of the relation between the control column movement and the column work can be seen in figure 6-1 for two instrument approaches. The column work for one approach is almost three times that of the other approach. From inspection of the time histories, the relative control displace- ments appear to have about the same relationship as the work levels. The control dis- placement for a given time interval near the end of the approach is much greater than that for the same time interval near the beginning of the approach. A measure of the work as defined here for an approach gives no indication of the variation of work for various por- tions of the approach.

Work 16.8ft-lb (22.8 N - m ) rad -. 09 I I I I I I I I I 90 80 70 60 50 40 30 20 10 0 Time before start of flare. sec Figure 6-1.- Control column time histories f o r two instrument approaches.

A summary of the control work and pilot ratings for the instrument approaches made during this investigation is given in table 6-1. The configurations listed in table 6-1 a r e described in detail in part 2.

Longitudinal Characteristics of pilot rating of the longitudinal characteristics with Column work.- The variation

---

column work is shown in figure 6-2 for the data of table 6-1. These data show a definite

trend of increasing work for increasing pilot rating - that is, the column work for the

approaches increases as the airplane characteristics deteriorate. For the basic variable- geometry configuration represented by the circular symbols, the work levels a r e higher for both pilots than indicated by the general trend of data for other configurations. These data represent the first configuration tested by each pilot and it is possible that the high work levels a r e representative of the early portion of the llpilotls learning phase." The possibility is indicated here that the measured work could be used to determine when the pilot's learning phase has been completed.

Differences between the two pilots are also indicated by the data shown in figure 6-2.

Although the same general trend is shown for each pilot, pilot A generally works harder Open symbols, pilot A Solid symbols, pilot B

1 2rL--

l0 4 a i2 16 hl 24

Column work, ff-lb M-

iz i 6 io-- -- -$4 is 32 0 4 8

Column work, N-m Figure 6-2.- Variation of pilot rating of longitudinal characteristics w i t h column work. Additional symbol identification is given in table 6-1.

7 - be combined directly with column work to give Open symbols, pilot A Solid symbols, pilot B a number representing total longitudinal work.

It can be seen in table 6-1 that the numbers for 6 - Cm G l Y) U throttle work are an order of magnitude lower A a L 0) c than that for the column work. According to U

5 5 -

-A U the opinion of the pilots, the throttle workload = V - m 0 0 was a much higher percentage of the total lon- a 4 - gitudinal workload than is indicated in table 6- 1.

z Therefore, the conversion of throttle motion to foot-pounds (newton-meters) of work did not provide a direct comparison with other control workloads having the same units of measurement.

From figure 6-3 it can be seen that there

I I I I 10 l I . 2

. 4 . 6 .8 L O is no consistent trend between throttle work Throttle work, ft-lb I I and pilot rating of the longitudinal characteris-

0 . 4 . ' I .2

Throttle work, N-m t i c s for the various configurations. Other fac- t o r s , such as mental effort and time required Figure 6-3.- Variation of pilot rating of longitudinal characteristics with throttle work. Additional symbol to operate the throttle, apparently are such a identification i s given in table 6 1 .

large part of the throttle workload that the physical effort required for throttle control is not proportional to the pilot's opinion of either the throttle workload or the rating of the longitudinal characteristics.

Flight-path deviations from glide slope.- The variations of flight path in the vertical plane a r e given in table 6-1 as root-mean-square deviations from the glide slope for each to expedite the presentation of these results, other methods of meas- approach. In order uring flight-path performance were not investigated; however, some of these methods could be more suitable than the r m s deviations. For example, deviations expressed in t e r m s of percent of glide slope altitude (angular deviation) could be used t o indicate the tighter flight-path control required as the airplane nears the ground. (An example of variation of airplane control effort as the airplane nears the ground is illustrated in fig- u r e 6-1 by the increased amplitude of control motions.) Another measure of glide-path

ll111ll1l I I I I I I I 1 I I 1 I I l l I1 I I I

control would be the e r r o r at the conclusion of the approach (point at which the pilot makes transition from instrument t o visual flight).

The variation of pilot rating of the longitudinal characteristics with r m s deviations of the airplane along the glide slope is given in figure 6-4. The r m s deviation from the glide slope has a value between 10 and 30 feet (3.05 and 9.14 meters) for all except four of the approaches shown in figure 6-4. Two approaches having a value outside this bound- a r y were made with the first configuration flown by pilot B. A s the longitudinal charac- teristics of the configurations deteriorate (increased pilot rating), each pilot tends to maintain glide-path control the same as, or better than, that for configurations with better longitudinal characteristics. This result agrees with previous observations that, as the piloting task becomes more difficult, the pilot tends t o increase his gain and continues t o perform with the same level of accuracy.

This statement indicates that there shauld be a correlation between work and flight- path accuracy if other variables such as configuration characteristics were held constant.

However, during this investigation too few approaches were made with a given configura- tion to determine the relation between work and flight-path accuracy.

Open symbols, pilot A Solid symbols, pilot B 6 - v) U .- c A A A VI .- L m c U m 5 - L m c U - m c.

.- 3 c 4 - .- m c 0

a

- L .- P 3 - c m L c 0 - .- a 2 - loL L _ - I _ . l J L- I I 8 16 24 32 40 48 56 64 Glide slope deviations, f t . I I

‘4 I i 2 16 0

Glide slope deviations, m Figure 6-4.- Variation of pilot rating of longitudinal characteristics w i t h rms glide slope deviations d u r i n g instrument approaches. Additional symbol identification is given in table 61.

Lateral-Directional Characteristics The variation of pilot rating of the lateral-directional characteristics with wheel work is shown in figure 6-5 for the approaches listed in table 6-1. The data indicate ,that for pilot B the trend is similar to that shown for the longitudinal characteristics (i.e., increased pilot rating is accompanied by increased work). This trend is not as evident for pilot A and in either case the data a r e rather widely scattered.

The rudder pedal work was added to the wheel work in an attempt to improve the correlation of work with pilot rating of the lateral-directional characteristics (fig. 6-6).

The addition of rudder pedal work did not appreciably change the correlation with pilot rating (see figs. 6-5 and 6-6), nor did it reduce the scatter between approaches with the same configuration and pilot. Just as for the throttle work, other factors which a r e not easily measured apparently constitute a large portion of the pilot's impression of lateral- directional work.

- Open symbols, pilot A Solid symbols, pilot B Y) U .- c v) .- & '5 c u m L m 0 B 8 . ..

c u

b' a a

c .- a 1 1 -

1 ; - ' 2 4 : 8 10 12

Wheel work, ft-lb I I -L .L -1 0 4 8 12 16 Wheel work, N-m Figure 6-5.- Variation of pilot rating of lateral-directional charac- teristics w i t h wheel work. Additional symbol identification i s given in table 6-1.

Open symbols, pilot A Solid symbols, pilot B v) U 1 - .e-.

v) .- k 0 4 P L m S V

-

m

4 "

! = 0 .- .e-.

0) L .- U - m L m - L m c .- c m L .e-.

- .- a

't

I I I I

4 2 A i ;r 10 12 14

Wheel and rudder pedal work, ft-lb I

0 6 k 1 ; i6

Wheel and rudder pedal work, N-m Figure 6-6.- Variation of pilot rating of lateral-directional characteristics with wheel and rudder pedal work. Additional symbol identification is given in table 6-1.

CONCLUDING REMARKS Part 6 of this publication has presented the results of a method of measuring pilot workload and compares these measurements with pilot opinion. The physical effort exerted by the pilot was expressed in foot-pounds (newton-meters) of work for the control column, wheel, rudder pedals, and throttle for simulated instrument approaches made during the in-flight simulation study of supersonic-transport configurations.

This exploratory study involving only a few approaches for each of several configu- rations did not furnish enough data to establish any firm conclusions; however, some ten- tative results and trends were indicated.

The physical effort required to move the control column during an instrument approach appears t o be a large enough portion of the total longitudinal workload to be used as documentation of the pilot's rating of the longitudinal characteristics.

The conversion of throttle motion to units of work did not provide a direct compari- son with other controls having the same units of measurement nor did the throttle work show any correlation with pilot rating of the longitudinal characteristics.

The physical effort required to move the wheel and rudder pedal controls during these instrument approaches did not correlate with pilot rating of the lateral-directional characteristics.

Langley Research Center, National Aeronautics and Space Administration, Langley Station, Hampton, Va., October 26, 1966, 720- 04-00- 06- 23.

REFERENCES 1. Cooper, George E.: Understanding and Interpreting Pilot Opinion. Aeron. Eng. Rev., vol. 16, no. 3, Mar. 1957, pp. 47-51, 56.

2. Klein, Richard H.; Archer, Richard B.; and Lew, Dan W.: Supersonic Transport Handling Characteristics During Approach and Landing Flight Regimes.

AFFDL-TR-65-227, U.S. Air Force, Dec. 1965.

I l l 1 ll1111ll111l111l11l11 I 1 llllIlllIllIllll1l1l111ll1 Ill I I I

I.

TABLE 6-1.- SUMMARY OF DATA FROM INSTRUMENT APPROACHES Pilot rating Configuration Pilc Lateral- Longitudinal directional characteristics characteristic

i

Variable geometry 15.5 4.72 14.48 19.63 5.26 1 7.13 2.36 3.20 0.320 0.43 4.5 3.0 (basic) 17.5 5 . 3 3 20.57 27.89 5 . 4 8 7.43 2 . 4 7 3.35 .415 .56 3.50 - 3.75 5 4 . 9 16.73 15.64 21.20 (a) .509 .68 3.0 (a) 34.8 10.61 21.25 28.81 (a) .309 .42 (a) Variable geometry 1 2 . 1 3.69 6.45 8.74 8.54 11.58 1.09 1.48 0.075 0.10 3 . 5 3 . 5 (6 and augmented) 27.7 8.44 7 . 0 6 9.57 4.76 6.45 1.94 2.63 (a) 24.0 7.32 4.62 6.26 3.15 4.27 1.19 1.61 .390 .53 B 11.2 3.41 3 . 7 6 5.09 2.86 3.88 1.41 1.91 .017 .02 3.25 3.0 4.51 16.2 4.94 6.11 4.85 6.58 1.21 1 . 6 4 207 .26 28.0 8.53 6.76 9.16 5.07 6.87 1.91 2.59 .310 .42 2 2 . 9 6.98 4.47 6.06 2.57 3 . 4 8 1 . 6 9 2.29 0.117 0.16 3.0 Variable geometry A 3.0 1 1 . 7 3.57 5.51 7.47 4.44 6.02 1.11 1.51 .120 .16 ((6 + A @ ) and b augmented) 1 8 . 6 5.67 6.01 8.15 5.26 7.13 2 . 0 4 2.76 .225 .31 B 15.8 4.82 4.02 5.45 2.00 2 . 7 1 1 . 9 2 2.60 .195 .26 22.4 6 . 8 3 4 . 8 3 6.55 3.25 4.77 1 . 4 2 1.93 248 .34 Variable geometry A 1 4 . 2 4 . 3 3 16.85 22.85 3.25 4.41 2 . 5 3 3.43 0.263 0.36 5.5 3.0 9 . 6 2 . 9 3 9.52 12.91 2 . 8 9 (b augmented) 3 . 9 2 1 . 4 9 2.02 .055 .07 18.0 5 . 4 9 11.89 16.12 3.03 4.12 1.13 1.53 .071 .lo with aft c.g.

B 1 0 . 5 3.20 10.36 14.05 3.30 4.47 1.47 1.99 .142 .19 4.5 - 5 . 0 1 4 . 5 4.42 12.25 16.61 3.73 5.06 2.57 3.48 .208 .28 1 1 3 . 8 4.21 11.77 15.96 4.00 5.42 2.09 2.83 .083 .ll Variable geometry A 1 9 . 2 5.85 7.53 10.21 4.77 6.47 2.63 3.56 0.458 0.62 3 . 5 ((6 +Aa)and 35.3 10.76 6.83 9.26 5.86 4.94 3.10 4.20 .225 .31 21.4 6.52 5 . 3 1 7.20 4.71 augmented) with 6.36 5.50 7.46 .083 .11 1 8 . 4 5.61 6.29 8.53 5.17 7.01 1.56 2.14 .067 .09 2 . 5 B 16.4 4.99 5.58 7.57 4.58 6.21 6.64 9.00 .095 .13 20.7 6.31 5 . 6 8 7 . 7 0 4.31, 5.84 1.81 2.45 .133 .18 A 17.7 5.39 4.27 5.79 3 . 5 8 4.85 1.74 2.36 0.308 0.42 4.0 - 4.5 degraded Dutch roll 21.9 6.67 8.47 11.48 4.91 6.66 1.64 2.22 .280 .38 14.4 4.39 8.34 11.31 8.97 12.16 1.23. 1.67 (a) B 37.0 11.28 7.34 9.95 8.95 12.13 .80 1.08 ,416 .56 4.5 21.2 6.46 7.27 9.86 8.65 11.73 3 . 8 4 5.21 ,575 .78 20.0 6.10 4.59 6.22 4.45 6.03 1.15 1.56 .200 .27 A 23.0 7.01 14.16 19.19 2 . 9 6 4.01 2.35 3.19 0.868 1.18 6.0 c4. 5 (basic) 16.2 4.94 11.58 15.70 3.86 5 . 2 3 2.77 3.76 .084 .ll 1 5 . 4 4.69 12.36 16.76 1.95 .ll 2.64 2.67 3.62 .083 B 17.7 5.39 6.24 8.46 4.49 6.09 1.30 1.76 .110 .15 4.5 - 5 . 0 ' 4 . 0 35.4 10.79 7.96 10.79 5.20 7.05 1.53 2.07 .237 .32 A 19.8 6.04 7 . 2 9 9.88 5.60 7 . 5 9 1.64 2.22 0.325 0.44 2.75 I 3.0 ((4 + Aa) and CL 22.6 6.89 7 . 6 4 10.36 3.98 5 . 4 0 1.47 1.99 .860 1.16 P B 25.0 7.62 4.65 6.30 3.54 4 . 8 0 1.22 1.65 .208 .28 3.0 2.5 augmented) 29.6 9.02 4.08 5.53 3.72 5.04 1.31 1.77 .192 .26 (a) 5.40 7.32 2.92 3.96 1.22 1.65 .176 .24 Fixed geometry A 22.5 6.86 5.84 7.92 9.47 0 0 0.457 0.62 12.84 4.5 - 5.0 ((6 + Aa) augmented) 27.0 8.23 5.19 7.04 6.36 8 . 6 2 0 0 .600 .81 with degraded Dutch roll B 25.6 7.80 6.45 8.74 4.73 6.41 0 0 .360 .49 4.25 - 4.50 dampingand C 2 8 . 4 8.66 3.52 4.77 4.04 5 . 4 7 0 0 .347 .47 n6 aData not obtained.

bRating not given for this configuration.

CThese approaches were made with unrealistic adverse yaw. See footnote on page 17.

L- 5221 NASA-Langley, 1967 - 2 l l 1 l l 1 l 1 1 1 l l 1 1 I I1 II I I1 I I I I l l "The aeronautical and space activities of the United States shall be

conducted so as t o contribute . . . f o the expansion of human Knowl-

edge of phenomena in the atmosphere and space. The Administration shall provide for the widest practicable and appropriate dissemination of information concerning its activities and the results tbereof."

-NATIONAL AERONAUTICS AND SPACE ACT OF 1958

NASA SCIENTIFIC AND TECHNICAL PUBLICATIONS

TECHNICAL REPORTS: Scient& and technical information considered important, complete, and a lasting contribution to existing knowldge.

TECHNICAL NOTES: Information less broad in scope but nevertheless of importance as a contribution to existing knowledge.

TECHNICAL MEMORANDUMS: Information receiving limited distribu tion because of preliminary data, security classification, or other reasons.

CONTRACTOR REPORTS: Scientific and technical information generated under a NASA'contract or grant and considered an important contribution to existing knowledge.

TECHNICAL TRANSLATIONS: Information published i n a foreign language considered to merit NASA distribution in English.

or of value to NASA SPECIAL PUBLICATIONS: Information derived from activities. Publications include conference proceedings, monographs, data compilations, handbooks, sourcebooks, and special bibliographies.

TECHNOLOGY UTILIZATION PUBLICATIONS: Information on tech- nology used by NASA that may be of particular interest in commercial and other non-aerospace applications. Publications include Tech Briefs, Technology Utilization Reports and Notes, and Technology Surveys.

Details on the availability of these publications may be obtained from: SCIENTIFIC AND TECHNICAL INFORMATION DIVISION NATIONAL AERONAUTICS AND SPACE ADMINISTRATION Washington, D.C. PO546