modeling, simulation, and flight characteristics of … · modeling, simulation, and flight...

,it

NASA Technical Memorandum 104236

Modeling, Simulation, and FlightCharacteristics of an AircraftDesigned to Fly at 100,000 Feet

Alex G. Sim

NASA Dryden Flight Research Facility, Edwards, California

m

1991

N/LRANational Aeronautics and

Space Administration

Dryden Flight Research FacilityEdwards, California 93523-0273

https://ntrs.nasa.gov/search.jsp?R=19910020839 2018-08-14T05:37:06+00:00Z

CONTENTS

ABSTRACT 1

INTRODUCTION

DESIGN CONSIDERATIONS

DESCRIPTION 3

AERODYNAMIC MODEL 3

Coefficients ................................................ 3

Lift.................................................. 3

Drag ................................................. 4

Pitchingmoment ........................................... 4

Derivatives ................................................. 5

Elevator and stabilizer conlrol effectiveness ............................. 5

Longitudinal damping ........................................ 5

Sideslip ............................................... 5Aileron ................................................ 6

Rudder ................................................ 6

Lateral-Directional Damping ..................................... 6Lateral-Directional Trim .......................................... 6

MASS AND PROPULSION MODELS 7

Mass Model ................................................ 7

Propulsion Model ............................................. 7

SIMULATION RESULTS 7

Approach .................................................. 7

Unaugmented Simulator .......................................... 8

Augmented Simulator ........................................... 9

Rapid Descent ............................................... 10Aileron Size ............................................. ., . . 11

CONCLUDING REMARKS 11

REFERENCES 12

FIGURES 13

lll

PRECEDING PAGE BLANK NOT FILMED

ABSTRACT

A manned real-time simulation of a conceptual vehicle, the stratoplane, was developed to study problems asso-

ciated with the flight characteristics of a large, lightweight vehicle. Mathematical models of the aerodynamics, mass

properties, and propulsion system were developed in support of the simulation and are presented. The simulation

was at first conducted without control augmentation to determine the needs for a control system. The unaugmented

flying qualities were dominated by lightly damped dutch roll oscillations. Constant pilot workloads were needed at

high altitudes. Control augmentation was investigated using basic feedbacks. For the longitudinal axis, flightpath

angle and pitch rate feedback were sufficient to damp the phugoid mode and to provide good flying qualities. In the

lateral-directional axis, bank angle, roll rate, and yaw rate feedbacks were sufficient to provide a safe vehicle with

acceptable handling qualities. Intentionally stalling the stratoplane to very high angles of attack (deep stall) was

investigated as a means to enable safe and rapid descent. It was concluded that the deep-stall maneuver is viable forthis class of vehicle.

INTRODUCTION

At the July 1989 Truckee Conference (Chambers, 1989), representatives of the atmospheric science community

affirmed the need for an atmospheric sampling aircraft capable of flight in the sU'atosphere at a minimum altitude

of 100,000 ft. Their mission requirements were for an aircraft to carry a 2500-1b payload for 6000 mi at an al-titude of 100,000 ft. Before the conference, a vehicle with similar mission capabilities had been studied by the

Lockheed Aircraft Corporation, Burbank, California (Chambers, 1990). "[he Lockheed vehicle was a twin-engine

configuration designed around reciprocating engines integrated with a three-stage turbocharger system. Each engine

was to produce 500 hp. In the total vehicle design, the area of high technical risk was the engine and turbocharger

development. Two aircraft could be developed and manufactured in approximately 4 years. At the time of theconference, the scientists considered the development time unattractive and lacked sufficient funds for this level of

alrcrat_ development.

After the Truckee Conference, the NASA Dryden Flight Research Facility responded to the need for a high-

altitude aircraft by continuing to study, advocate, and further develop the aircraft design through in-house and

contracted efforts. Engines have been identified that need only to be derated and can be used without extensive

development. Also, applicable turbo-machinery hardware have been identified that were originally developed for

the Teal Rain ('IRX) and its successor, the Condor (Henderson, 1990).

Many configurations have been studied, including variations in the number of engines, span, payload, and un-

manned flight. One configuration studied was a manned single-engine stratospheric engine demonstrator vehicle,

the stratoplane, and is the topic of this report. The stratoplane is a conceptual design of a large, manned single-engine

monoplane created to conduct atmospheric sampling and to validate its engine technology. The vehicle was designed

with a gross weight of 11,100 lb, a wing span of 180 t, and a propeller 30 ft in diameter. Conceptually, this vehicle

would be used to validate and refine the propulsiontechnology before building a full mission vehicle and to conduct

missions of lesser payload and range. The stratoplane would have an approximately 1000-1b payload capability and

a 3700-nmi range at an altitude of 100,000 ft.

A simulation study was conducted from May through October 1990 to address concerns over the stability, control,

and flying qualities of operating a manned stratoplane vehicle. The study was limited to quickly bounding the

concerns rather than conducting an indepth analysis. For example, a rigid body model was used for the aerodynamics

and specific aerodynamics terms were varied to bound the aeroelastic effects.

The final phase of the study used a real-time manned simulator with a fixed-base generic cockpit similar to the

type used in Smith, Schilling, and Wagner, 1989. Mathematical models of the aerodynamics, propulsion system

and mass properties were developed to support both the simulation and a linear control analysis. The aerodynamics

model was implemented to very high angles of attack to allow for evaluation of the controlled deep-stall maneuverused in Sim, 1990.

The scope of the control system evaluation was fimited to primarily the basic feedback of rates and attitudes.

This report documents the models used to build up the simulation, the vehicle flying qualities, and the resulting

control system.

DESIGN CONSIDERATIONS

The primary design of the stratoplane had been established (fig. 1) before undertaking this study of flight char-

acteristics. However, it is of interest to higlflight some of the generic characteristics of any vehicle designed to fly

at 100,000 fl that also has range and payload requirements.

For the atmospheric sampling instrumentation (payload) to work, subsonic flight must be maintained (Chambers,

1989). Thus, a limit speed of Mach 0.70 was chosen for the vehicle. This speed makes it convenient to use an unswept

wing and simplifies propeller design. The maximum wing loadings are driven into the 6- to 8-lb/ft 2 range (fig. 2),

while the maximum usable lift coefficients are generally within 4-0.2 of 1.0. Although it is possible to design for

lower lift coefficients, such designing drives up vehicle size and drives down cruise speed. Lowering cruise speed

usually lowers range, especially when considering upper atmosphere winds (fig. 3) or limits to endurance. Thus, it

is prudent to design for cruise at a high lift coefficient. Also, the lift coefficient for cruise must be slightly lower

(approximately 0.2) than maximum lift (stall) to allow for both gentle maneuvers and gusts.

For a vehicle the size of the stratoplane design, the mean-chord Reynolds number at 100,000 ft is near 600,000.

As shown later in the report, a good airfoil at this Reynolds number at Maeh 0.65 will have a maximum lift coefficient

in the 1.2 to 1.3 range. Thus, the usable lift coefficient for cruise will be near 1.0. At the low Mach number and

low-altitude flight condition, the available maximum lift coefficient will be in the 1.8 to 1.9 range. As a point of

reference, Mach 0.65 at 100,000 ft corresponds to a dynamic pressure of only 6.9 lb/ft 2 , a calibrated airspeed of

47 knots, and a true airspeed of 382 knots.

The aspect ratio is driven up by the need to fly with low drag at a high-lift coefficient to minimize the horsepower

required for cruise at 100,000 ft. For a propeller-driven engine, thrust is proportional to horsepower divided by true

velocity. At 100,000 ft, true velocity is approximately eight times its sea level value at the same dynamic pressure,

thus requiring at least eight times the horsepower. Vehicle drag increases with both the higher Mach number and

the lower Reynolds numbers associated with high-altitude flight and additionally increases the required horsepower.

Also, the difficulties of designing an efficient propeller for the high altitudes also increase the required horsepower.

The total propulsion system weight required to fly at 100,000 ft is approximately five times as heavy for the same

horsepower as a reciprocating engine designed to operate at sea level. This additional weight further increases the

horsepower required for straight and level flight. For a gasoline engine, fuel used and heat generated are approxi-

mately proportional to horsepower. Cooling the propulsion system can become a major source of drag and, thus, a

requirement for additional horsepower. A low-drag technique must be used to dissipate the engine and turbocharger

heat. Conceptually, the cores of the engine and intercooler radiators must be cowled so that the cooling air is slowed

before passing through the core. Then the heated air is accelerated before exiting the cowl. Using this technique,

only a slight drag penalty is incurred to cool the engine and intercoolers. An early application of this low-drag

cooling technique was incorporated into the F-51 World War H fighter.

The vehicle structure envisioned for this study is primarily all carbon fiber sandwich with a nomex honeycomb

core. To meet mission goals, the wing structural weight must be near 1.2 lb/ft 2 of wing area, which is approximately

the same as the Voyager (Norris, 1988) and Condor (Henderson, 1990) aircraft. This factor is significant as these

aircraft have been flown with higher aspect ratios than the proposed stratoplane. Although care must be taken tominimize structural weight, this class of lightweight structure has been achieved using current technology. Thefuselage and tail surfaces must be designed in a similar lightweight manner.

It is envisionedthatthe cockpit will be designed for a pressure-suited pilot, but will provide pressurization to

the 35,000-fl level as an emergency backup to the pressure suit. The pilot and his equipment are estimated to add

500 lb to the vehicle weighL The main landing gear is located close to the center of gravity and the noscwheel wouldbe steerable.

DESCRIPTION

The stratoplane (fig. 1) is a conceptual design of a large, manned, singie-engine monoplane designed to con-

duct atmospheric sampling and to validate its engine technology. The vehicle was designed with a gross weight of

11,100 ib and a wing span of 180 ft that gave an aspect ratio of 20.9. The 30-ft-diameter propeller is two-bladed and

was clutched into the horizontal position during either a towed takeoff or a power-off landing.

The wings have payload pods 30 fi out on each wing. A box spar and D-tube leading edge are the primary

structure of the wing. The inboard box spar contains 4000 lb of fuel. The baseline airfoils are those used for the

"Ames wing" in Kennelly et al., 1990. The airfoils are of supercriticai profile with a Mach number design of 0.70 and

have good low Reynolds number characteristics. The wing thickness varies from 14 percent at the root to 12 percent

at the tip.

The horizontal tail shown is a T-tail with capability to deflect through large angles to give good deep-stall capa-

bility. The conventional tail location was also investigated and is still viable should deep-stalled flight not be needed.

The vertical tail shown is the original tail design that was used for the initial simulation. A vertical tail scaled up by

50 percent in area was later investigated and became the baseline tail for this study.

The identified engine is a gasoline-powered, ainminum-block, liquid-cooled V-8 Thunder aircraft engine (Jane _,

1984) that was derated to 600 hp and fitted with a two-speed, clutched gearbox. Level cruise at 100,000 ft and

maximum gross weight required most of the 600 hp. The engine radiator is contained in a large pod in the aft fuselage.

A three-stage turbocharger system is used with each four-cylinder side of the engine. Intercoolers are located under

each inboard wing. The total weight for the engine, turbomachinery, and cooling system is approximately 2200 lb.

AERODYNAMIC MODEL

The approach taken to generate the aerodynamic model was to first compute the low angle-of-attack data using

both the asymmetric vortex lattice program (Lamar and Gloss, 1975) and handbook methods CO.S. Air Force, 1978;

Etkin, 1967). Effects of low Reynolds number were included in the determination of maximum lift and profile drag.

Data at high angles of attack were estimated using a combination of handbook techniques and the flight results and

wind tunnel trends in Sim, 1990. The high angle-of-attack data were included to investigate the deep-stall maneuver

as a way of rapidly and safely descending from high altitudes.

The resulting aerodynamic model ranged from an angle of attack of-6 ° to 90 ° and from Mach 0 to 0.70. The

angle of attack was digitized every 2 ° from -6 ° to 20 ° and every 10° from 20 ° to 90 °. Data were generated for

Mach 0, 0.40, 0.60, and 0.70. Moment data in the aero model are referenced to the quarter chord of the mean

aerodynamic chord.

Coefficients

Lift

Although specific airfoils CKennelly et al., 1990) have been identified for the stratoplane, data for these airfoils are

not available over a wide Mach and Reynolds number range. Thus, data from these airfoils were combined with data

3

from other sources (McCroskey, 1987; McGhee, Beasley, and Whitcomb, 1979; Althaus, 1980; and Miley, 1982)

and used to generate a plot of trends in maximum lift coefficient as a function of Mach and Reynolds number (fig. 4).

This same plot is shown with lines of maximum and minimum speed for the stratoplane superimposed on it (fig. 5).

Note that there is only a small Reynolds number variation for any given Mach number within the envelope bounded

by the speed lines. This allowed the maximum lift coefficient to be modeled as a function of Mach number in the

simulation with the Reynolds number effects implicitly included in the Mach number data. Also, the stratoplane

flight envelope is well above the subedtical Reynolds numbers (generally below 200,000) where strong viscouseffects dominate the flow.

The lift coefficient (fig. 6) was a result of several steps. First, a representative value for the angle of attackat zero lift was chosen based on airfoil camber and trends of similar configurations. Next, the asymmetric vortex

lattice program was used to calculate a value for the lift curve slope (at low angles of attack) based only on the

planar planform. Then, a hand fairing was drawn between the lift curve slope and the maximum rift coefficient.

Finally, a hand fairing was used between the maximum lift coefficient and the data estimated for high angles of

attack. The angle of attack corresponding to maximum lift coefficient (for each Mach number) was used as the

division between the low angle-of-attack calculations from the asymmetric vortex lattice program and the higher

angie-of-attack estimates for the remaining coefficients and derivatives.

Drag

For the low angles of attack, the estimation of total drag coefficient was separated into the profile and induced

components. A genetic plot of airfoil profile drag as a function of Reynolds number (fig. 7) was generated from many

sources in a manner similar to figure 4. Using the relationship between Reynolds and Mach numbers from figure 5,

the data from fig. 7 were used to generate increments in profile drag with Mach number. These increments were then

added to an estimated baseline profile drag value of 0.0181. Note that the stratoplane operates in a Reynolds number

range that avoids the large profile drag penalties that occur with Reynolds numbers less than 200,000. The induced

drag per unit iift-cx_fficient squared was calculated using the asymmetric vortex lattice program for each Mach

number. The profile and the induced terms were then combined to obtain total drag as a function of lift coefficient.

Using the lift coefficient data in figure 6, the data were arranged as a function of angle of attack and faired into the

high angle-of-attack estimates to give the final drag data in figure 8.

The lift-to-drag ratio was computed using the data in figures 6 and 8 and showed a maximum value of 29 (fig. 9).

Because a clean sailplane of similar aspect ratio is expected to have a maximum value near 36, the stratoplane value of

29 was judged reasonable considering the payload pods and the need to provide cooling. The lift and drag coefficients

were translated to normal force coefficient (fig. 10) and axial force coefficient (fig. 11) for the simulation model.

Pitching moment

The slope of the pitching moment coefficient per unit normal force coefficient (static stability parameter) was

computed using the asymmetric vortex lattice program and gave a neutral point of approximately 48 percent of the

mean chord at low angles of attack. For the high angles of attack, it was judged that the neutral point would move

back to approximately 60 percent of the mean chord. This shift in neutral point was included in the static stability

parameter for the high angle-of-attack estimates. A representative value for the pitching moment at zero lift was

chosen to give a trim value near a lift coefficient of 0.5. An airplane designer can vary this value through choice

of airfoil camber, wing incidence, wing twist, and tail incidence. The pitching moment coefficients (fig. 12) were

computed by multiplying the static stability parameter by the normal force coefficient from figure 10.

4

Derivalives

Elevator and stabilizer control effectiveness

The primary longitudinal control was defined as the elevator. Trim and transition to and from deep-stalled flight

was accomplished using the stabilizer which, on the simulator, was controlled using a separate side lever. In deep

stall, the stabilizer is somewhat aligned with the free-stream air (with attached flow) allowing both control surfacesto maintain reasonable control effectiveness. The low angle-of-attack values for pitching moment coefficient and

normal force coefficient derivatives were computed using the asymmetric vortex lattice program. Again, these

values were faired into the high angle-of-attack estimates. Trends in the axial force derivatives were taken primarily

from the high angle-of-attack estimates. The resulting derivatives are presented in figures 13(a), (b), and (c) for the

elevator and figures 14(a), (b), and (c) for the stabilizer.

Longitudinal damping

The damping in pitching moment coefficient due to pitch rate (fig. 15(a)) and the normal force coefficient due

to pitch rate (fig. 15Co)) were both calculated using the asymmetric vortex lattice program. Because these damping

terms primarily result from the stabilizer and because the stabilizer maintains attached flow at high angles of attack

(in deep stall), the damping derivatives were held constant over angle of attack.

Sideslip

The low angle-of-attack values for yawing moment coefficient and side force coefficient due to sideslip deriva-

tives were calculated using the asymmetric vortex lattice program with a side view planform panel model and the

original small vertical tail. Data for high angles of attack were taken from the high angle-of-attack estimates. The

yawing moment cx_ffieient due to sideslip derivative (fig. 16) represents the directional stability and is lower than

normally desired for an unaugmented vehicle. The side force due to sideslip derivative is presented in figure 17.

The baseline vertical tail, with 50 percent more area (large tail), was modeled using the multiplication factors from

table 1. These factors were obtained using the ratio of the asymmetric vortex lattice program calculations of panel

models with the two tail sizes.

Table 1. Vertical tail-size multipliers.*

Derivative Multiplier

Change in side force coefficient

due to angle of sideslip 1.31

Change in yawing moment coefficient

due to angle of sideslip 1.43

Increment in rolling moment coefficient

due to rudder 2.23

Increment in side force coefficient

due to rudder 1.89

Increment in yawing moment coefficientdue to rudder 1.82

Yaw damping due to yaw rate parameter 1.50

*Used to convert from original tail to 50 _wcent larger tail area.

The rolling moment coefficient due to sideslip is commonly called the effective dihedral derivative because it isstrongly influenced by wing dihedral. Although elastic bending properties of the wing are not yet well defined, the

lightweight structure of the stratoplane will give considerable wing dihedral. For this study, a curvalinear dihedral

with a 10" tangent line at the wingtip was used. Handbook calculations 03.5. Air Force, 1978) show that a rigid

wing with this dihedral will produce the low angle-of-attack values shown in figure 18(a). The high angle-of-attack

values were from the high angie-of-attack estimates. However, aeroelastic relief in sailplanes is known to alleviate

the effective dihedral. Considering the scope of this study and the indeterminate aeroelastics, it was decided to

hound the effective dihedral estimate from 25 percent of this value to the full value. Both values were studied in the

simulation, with the baseline value the 25 percent value (fig. 18(b)).

Aileron

The rolling moment coefficient due to aileron is the primary roll control derivative (fig. 19(a)). The derivative

was obtained using the asymmetric vortex lattice program calculations for low angles of attack and was combined

with the high angle-of-attack estimates. The yawing moment coefficient (fig. 19(b)) and corresponding sideforce

coefficient due to aileron (fig. 19(c)) were estimated based on the premise that a differential control scheme (more

up than down aileron) would be implemented. Thus, proverse yaw was attained at low angles of attack similar to

the data in Sim, 1990.

Rudder

Rolling moment coefficient due to rudder (fig. 20(a)), yawing moment coefficient due to rudder (fig. 200)),

and sideforce coefficient due to rudder (fig. 20(c)) were calculated using the asymmetric vortex lattice program

for low angles of attack and combined with the high angle-of-attack estimates. The 50-percent larger rudder that

accompanied the large, vertical tail was modeled in the simulation using the multiplication factors from table 1.

Lateral-Directional Damping

The change in rolling moment coefficient due to roll rate parameter, the primary roll damping derivative

(fig. 21(a)), was calculated at low angles of attack using the asymmetric vortex lattice program and combined with

the high angle-of-attack estimates. The change in yawing moment coefficient due to roll rate parameter (fig. 21 (b))

and the change in rolling moment coefficient due to yaw rate parameter (fig. 21(c)) were both estimated based on the

trends of Sim, 1990. The change in yawing moment coefficient due to yaw rate parameter (fig. 21(d)), the primary

damping in yaw derivative, was derived using the general levels from the asymmetric vortex lattice program and

the data in U.S. Air Force, 1978. The increase in yaw damping due to the large vertical tail was modeled in the

simulation using the multiplication factors from table 1.

Lateral-Directlonal Trim

A bias term in the rolling moment, yawing moment, and sideforce coefficients were included to model the

poststall. While stall is often treated as a symmetric maneuver, it rarely happens that way. Conditions such as

rigging, uncoordinated flight, and turn rate will cause one wing to stall first while the other wing remains attached.

This effect is accentuated by long wingspans. Once the second wing also stalls, a greater degree of symmetryis retained.

For the Schweizer SGS 1-36 sailplane in Sim, 1990, the trim biases were of greater magnitude than the available

control power during the poststall region. The values shown for the plots of rolling moment coefficient bias (fig. 22),

6

yawing moment coefficient bias (fig. 23), and side force coefficient bias (fig. 24) were multiplied by scale factors

within the simulation of typically -4-10.

MASS AND PROPULSION MODELS

Mass Model

For the mass properties, a weight and a primitive shape were assigned to 31 vehicle components. The vehicle

was defined as symmetric in planform. The moments of inertia and mass centroid locations were then generated

using a spreadsheet and are summarized in table 2. For the simulation, the longitudinal center of gravity was at

39 percent of the mean chord, while the neutral point was near 48 percent. The high vertical center of gravity is a

result of the wing dihedral. Much of the simulation study was conducted using a 2000-1b middle fuel weight, while

the full and empty conditions were given limited study.

Table 2. Mass properties.

Fuel weight, lb

Total vehicle weight, lb

Rolling moment of inertia, slug-ft 2

Pitching moment of inertia, slug-ft 2

Yawing moment of inertia, slug-ft 2

Roll-to-yaw cross product of inertia,

slug-ft 2

Center of gravity behind quarter

chord reference, ft

Center of gravity above thrust

centerline reference, ft

0 2,000 4,000

7,100 9,100 II,I00

132,200 136,900 141,600

36,840 36,990 37,150

162,500 167,300 172,000

4,620 4,680 4.740

1.32 1.32 1.32

1.13 1.03 0.97

Propulsion Model

The propulsion model included fuel flow, propeller efficiency, and propeller inertial effects. The throttle lever

and fuel flow was proportional to hp. The pilot could select either 250 rpm for low-altitude flight or 500 rpm for

high-altitude flight. The propeller efficiency (fig. 25) was calculated using the Hamilton-Standard data in Lan and

Roskam, 1980. An integrated propeller lift coefficient of 0.15 was chosen to minimize Mach number effects. The

500-rpm line is questionable at altitudes lower than 40,000 it as a result of the very fiat pitch needed at these altitudes.

Use of an intermediate propeller speed of approximately 375 rpm would benefit climb performance in the middle

altitudes. However, this speed was not considered significant for the simulation. The propulsion model used was

considered to be the minimum needed for a stability and control simulation.

SIMULATION RESULTS

Approach

Both low- and high'altitude nominal flight conditions (table 3) were chosen for evaluation. Hying qualities of

the unaugmented vehicle were first evaluated to find out whether the vehicle could be flown without a control system.

The maneuvers investigated included climb, descent, 20 ° bank turns, and recovery from control perturbations.

7

Table 3. Nominal flight conditions studied.

Reference name Low altitudeHigh altitude

Altitude, ft 10,000 100,000

Angle of attack, deg 5 3

Dynamic pressure, lb/ft 2 6.4 6.4

Fuel weight, lb 2,000 2,000Mach number 0.08 0.63

Reynolds number, per ft 4,240,000 66,900

True velocity, knots 50.6 367

Early investigations on the simulator showed the vehicle to be difficult to fly at any altitude as a result of near-neutral dutch roll oscillations. Thus, a vertical tall with 50 percent more area 0arge tall) was investigated and became

the baseline configuration for this simulation study. With the large tail, the vehicle could be flown without control

augmentation, but still exhibited very lightly damped dutch roll characteristics.

Linear analyses were conducted to provide insight into the dynamic modes of the vehicle and to substantiate

the control system design. The linear analysis results are displayed in classic root locus plots. The scope of the

control system evaluation was limited to primarily the basic feedback of rates and attitudes. Safe and rapid descend-

ing either from 100,000 It or through the turbulent middle altitudes was investigated by intentionally deep stalling

the stratoplane.

Unaugmented Simulator

At low altitudes, the vehicle showed good longitudinal short-period stability, but the phugoid mode was near

neutrally stable (fig. 26). High aspect ratio causes a decrease in phugoid damping and an increase in phugoid fre-

quency. Although the phugoid can be damped by the pilot, it does increase pilot workload as a result of the difficulty

to attain and hold longitudinal trim. Neither airspeed nor rate of climb could be precisely controlled.

At low altitudes, the vehicle showed good roll stability; however, the dutch roll mode was lightly damped and

the spiral mode was slightly unstable (figs. 27 and 28). The spiral mode can be damped by the pilot but requires some

attention. The loose dutch roll mode comes from generally low directional stability as a result of both a large fuselage

side area and a minimal-sized vertical tail. The larger tail improved the overall lateral-directional flying qualities

but still did not produce a good flying vehicle. The vehicle was difficult to fly and required constant pilot attention.

The stratoplane tended to fly at indicated airspeeds near 45 knots, which at low altitude is near 45 knots true.

With these low airspeeds and a loose dutch roll, the pilot must be careful to avoid excessive sideslip. With the small

tail, this task was difficult in the presence of larger dutch roll oscillations. Using the higher values of the effective

dihedral derivative in fig. 18(a) made the vehicle more difficult to control as a result of the increased tendency for

the pilot to couple into the sideslip excursions.

At 100,000 It, the true velocity is approximately eight times greater than at sea level for the same indicated veloc-

ity. Because the aerodynamic damping and force terms are divided by true velocity, they are significantly decreased

at high altitudes. The reduction in longitudinal damping results in a neutrally stable phugoid mode (fig. 29). This

results in a vehicle that can be safely damped but cannot be flown with precision because of the inability to stabilize

rate of climb. With full fuel, the flying tasks were more difficult as a result of the small margin between cruise lift

and stall.

At a 100,000-it altitude, the reduction in lateral-directional damping produced an unstable spiral and a barely

stable dutch roll mode that dominated the vehicle response (fig. 30). With the larger vertical tail, it was still possible

to manually fly the unaugmented vehicle, but only with high pilot workload. Piloting the airplane required constant

attention. Fortunately, the side force terms are also divided by true velocity, alleviating the tendency to generate

excessive sideslip. Unaugmented flight with the original small vertical tail was still possible, but only with very

high pilot workload. With either vertical tail, the effects of using higher Values of the effective dihedral derivative

in figure 18(a) were masked by the difficulty of maintaining control of the vehicle.

Augmented Simulator

The goal of the control system development task was to investigate the value of a relatively uncomplicated low-

gain system. Thus, with an exception, only gains were used in the feedback loops, and forward-loop compensation

was not used. A low set of gains was desired to minimize actuator rate and power requirements. As a result of the ease

of mechanization, the feedbacks were tried first on the simulation and then later analyzed using linear analysis. The

analysis results validated the control system design and cross-cbecked the simulation. Much of the understanding

of how the feedbacks affects particular modes comes from the linear analysis; however, the resulting flying qualities

come from the simulation.

In the longitudinal axis, the goal was to stabilize the phugoid at high altitudes and, preferably, eliminate its

second-order response. The goal was accomplished by feeding back flightpath angle to the elevator (fig. 31). At

low gains in the range of 0.2 to 0.3, the phugoid was stabilized, but with an underdamped response that resulted in a

minor tendency to search for trim. With gains of approximately 0.6, the phugoid became overdamped and produced

a good flying vehicle with little overshoot. At higher gains of approximately 0.8 and 0.9, the phugoid eigenvalues

were driven real, also producing a good flying vehicle with no tendency to overshoot.

However, the high flightpath feedback gains reduced the short-period damping. As a countermeasure, pitch rate

feedback was fed to the elevator (fig. 32), adding damping to the short period without significantly affecting the

phugoid. The resulting control system (fig. 33) provides good longitudinal flying qualities at all altitudes with the

same set of fixed gains. A minor degradation in flying qualities can be traded for lower gains.

The approach taken for the lateral-directional control system was to feedback the states (angles of bank and

sideslip, and roll and yaw rate) to the appropriate control (aileron or rudder). Positive spiral stability was achieved

by feeding the bank angle to the aileron to obtain a simple wings leveler. At low altitude this required a gain in the

0.3 to 0.4 range (fig. 34(a)), while at a high altitude only a gain between 0.04 and 0.1 sufficiently stabilized the spiral

(fig. 3403)). Higher gains at high altitude would drive the spiral too stable, destabilize the roll mode, and destabilize

the dutch roll mode. I_e roll rate was fed to the aileron (fig. 35) to increase roll damping and slightly improve flying

qualities. A roll rate gain of 0.3 was used at both high and low altitudes.

Also, yaw rate, with a washout filter, was fed to the rudder (fig. 36) to improve yaw damping. The yaw rate

feedback with a gain of 0.1 slightly improved the flying qualities. However, with other loops closed (not shown),

higher yaw rate gains degraded the flying qualities and made yaw rate an ineffective feedback. The unity time

constant chosen for the yaw rate washout filter is perhaps too large, but was not refined as a result of the ineffectual

nature of the yaw rate feedback.

The use of bank angle and the rates (fig. 37) greatly improved the lateral-directional flying qualities and produced

a safe flying vehicle at all altitudes with either the original or the larger vertical tail. However, the small sideslip

oscillations resulting from a loose dutch roll mode were not totally eliminated. Feeding back angle of sideslip to the

rudder (fig. 38(a)) was studied but did not produce significant improvement until unreasonably high gains greater

than 1 were used. Note that the angle of sideslip feedback loop was presented with the other loops closed as a result

of strong coupling with the bank angle feedback loop.

Variations in the effective dihedral derivative from the 25-percent baseline value (fig. 1803)) to the 100-percent

value (fig. 18(a)) produced interesting results. At low altitudes, the higher values of the effective dihedral derivative

provided more dutch roll stability and gave slightly better handling qualities. At high altitudes with the baseline

25 percent value, the dutch roll was near neutral stability and oscillated primarily in the yaw axis. With the full

9

baselinevaluesof effectivedihedral,thehandling qualities were degraded as a result of additional rolling motion in

the dutch roll mode.

One unconventional lateral-directional feedback was studied and found to provide benefits for specific situations.

For high altitudes, angle of sideslip, with the wrong sign on the gain, was often feedback to the aileron (fig. 38('0)).

Better flying qualities were produced, especially when high values of the effective dihedral parameter were modeled.

The gain used was -0.1, as higher negative gains drove the dutch roll mode down in frequency and produced

instability. Good lateral-directional flying qualities requite a more complete study.

Rapid Descent

In an emergency, the pilot may need to rapidly descend to altitudes below 48,000 ft where a pressurized envi-

ronment is not required. Although an escape capsule is viable, it has the undesirable side effect of losing the vehicle.

Two ways of rapidly descending are with either a controlled deep-stall maneuver or with a large external drag device

like a parachute to limit maximum airspeeds. Although a large external drag device is feasible, it was decided to

direct the simulation study toward the deep-stall maneuver that uses stalled wings for its drag device.

The deep-stall maneuver has been used throughout aviation history and has been shown to be safe and repeatable

(Sim, 1990). The deep-stall maneuver can be divided into three stages: entry to very high angles of attack, stabilized

flight at very high angles of attack (typically 40 ° to 70,), and recovery back to low angles of attack. To control deep

stall, the stabilizer needs to deflect through large angles. Thus, the airflow over the horizontal tail can remain attached

while the wing is fully stalled. The stabilized angle of attack is then a function of stabilizer and elevator angle, with

nominal values in the 40 ° to 60 ° range. During the simulation studies, the control system was engaged using the

high altitude, low angle of attack, and gain set.

A special technique was needed for the entry into deep stall to keep the peak pitch attitudes less than 15° to 20,

and to avoid the excessive 50' to 60 ° values that were easy to attain. With power off, the vehicle was slowed to stall

using the elevator. The elevator was first used to slowly raise the angle of attack to approximately 25°. The stabilizer

was next slowly moved to further increase the angle of attack while the elevator was relaxed to its neutral position.

The pitch attitude then stabilized at - 12° and airspeed indicated at approximately 38 knots. The entire transition to

a stabilized high angle of attack required approximately 20 sec to complete.

During stabilized high angle-of-attack flight, the pitch and roll axis were relatively steady, but the heading wan-dered. Active use of the rudder tended to excite the dutch roll mode, which could be easily damped using aileron.

For a deep-stall maneuver that stabilized at an angle of attack of 70 °, it took 4 rain and 20 sec to descend from

100,000 ft to 40,000 ft. Rate of sink peaked at 430 ft/sec at approximately 90,000 ft and decreased to 130 ft/sec at

40,000 ft.

Recovery to low angles of attack also required special flight techniques to limit the required peak airspeed. First,

the angle of attack was slowly lowered to approximately 25 ° using the stabilizer. The elevator was used to rapidly

(over approximately 2 sec) push the nose down and thus lower the angle of attack. Then, the nose was slowly pulledback and the horizontal stabilizer neutralized. The indicated airspeeds required for recovery were at least 58 knots for

a recovery at 75,000 fl and at least 52 knots for a recovery at 35,000 ft. Unfortunately, the corresponding equivalent

airspeeds required for the recovery are slightly higher than needed for any other maneuver studied.

The deep-stall maneuver is viable for this class of vehicle should rapid descent be necessary. However, incorpo-

ration of deep-stall capability may require either an increased maximum airspeed limit or a small drag control device

to limit peak airspeeds.

10

Aileron Size

The ailerons were initially sized to be similar to those used on many low-performance sailplanes. However,

during the simulation, less than half the total +25 ° maximum deflection for each aileron was used. Thus, smaller

ailerons would have been adequate.

CONCLUDING REMARKS

The stratoplane is a conceptual design of a large, manned single-engine monoplane designed to conduct atmo-

spheric sampling and to validate its engine technology. The vehicle was designed with a gross weight of 11,100 lb,

a wingspan of lg0 ft, and a propeller 30 R in diameter. To study problems associated with the flight characteristics

of such a large, lightweight vehicle, a manned reai-time simulation was developed. In support of the simulation,

mathematical models of the aerodynamics, mass properties, and propulsion system were developed.

The simulation was initially conducted with the unaugmented vehicle to determine the needs for a control aug-

mentation. The unaugmented longitudinal flying qualifies were dominated by a lightly damped phugoid at low

altitude and a neutrally stable phugoid at 100,000 ft. The vehicle could be safely flown in the longitudinal axis,

but could not be precisely controlled as a result of the difficulty in maintaining trim. In the laterai-directionai axis,

the flying qualities were dominated by lightly damped dutch ron oscillations. The vehicle required constant pilot

attention at the lower altitudes and very high pilot workloads at high altitudes.

Control augmentation was investigated using basic feedbacks. For the longitudinal axis, flightpath angle and

pitch rate feedback were sufficient to damp the phugoid mode and produce good flying qualities. In the lateral-

directional axis, bank angle, roll rate, and yaw rate feedbacks were sufficient to provide a safe vehicle with acceptable

handling qualities.

Intentionally deep smiling the stratoplane was investigated for a safe and rapid descent from either 100,000 fl or

through the turbulent middle altitudes. R was concluded that the deep stall maneuver is a viable maneuver for this

class of vehicle. However, incorporation of deep-stall capability may require either an increased maximum airspeed

limit or a small drag device to limit peak airspeeds.

11

REFERENCES

1. Althaus, Dieter, Profilpolaren fur den ModeUflug (Polars of Airfoils for Model Airplanes), Neckar-Verlag,

Villengen-Schwenningen, Germany, 1980.

2. Chambers, Alan, and R. Dale Reed, "A Very High Altitude Aircraft for Global Climate Research," Unmanned

Systems, vol. 8, no. 3, Summer 1990, pp. 14--19.

3. Chambers, Alan B., Global Stratospheric Change: Requirements for a Very-High-Altitude Aircrafl for Atmo-

spheric Research, NASA C'P-10041, 1989.

4. Etkin, Benard, Dynamics of Flight: Stability and Control, John Wiley & Sons, New York, London, and Syd-

ney, 1967.

5. Henderson, Breck W., "Boeing Condor Raises UAV Performance Levels," Aviation Week and Space Technol-

ogy, vol. 132, no. 16, April 23, 1990, pp. 36-38.

6. Jane'sAll the World's Aircraft 1983--84, Jane's Publishing Co., New York and London, 1984, p. 843.

7. Kennelly, Robert A., llan M. Kroo, James M. Strong, and Ralph L. Carmichael, Transonic Wind Tunnel Test

of a 14% Thick Oblique grmg, NASA TM-102230, 1990.

8. Lamar, John E., and Blair Gloss, Subsonic Aerodynamic Characteristics of Interacting Lifting Surfaces with

Separated Flow Around Sharp Edges Predicted by a Vortex-Lattice Method, NASA TN D-7921, 1975.

9. Lan, C. Edward, and Jan Roskam, Airplane Aerodynamics and Performance, Roskam Aviation & Engineering

Corp., Ottawa, KN, 1980, pp. 289-329.

10. MeCroskey, W.L, A Critical Assessment of Wind Tunnel Results for the NACA 0012 Airfoil, NASA TM-

100019, 1987.

11. McGhee, Robert J., William D. Beasley, and Richard T. Whitcomb, NASA Low- and Medium-Speed Airfoil

Development, NASA TM-78709, 1979.

12. Miley, S.J., A Catalog of Low Reynolds Number Airfoil Data for Wind Turbine Applications, Texas A&M

Univ., College Station, TX, Feb. 1982.

13. Norris, Jack, Voyager, The World Flight: The Offzcial Log, Flight Analysis and Narrative Explanation, Jack

Norris, Publisher, Northridge, CA, 1988.

14. Sire, Alex G., Flight Characteristics of a Modified Schweizer SGS 1-36 Sailplane at Low and Very High Angles

of Attack, NASA TP-3022, 1990.

15. Smith, J.P., Lawrence J. Schilling, and Charles A. Wagner, Simulation at the Dryden Flight Research Facility

from 1957 to 1982, NASA TM-101695, 1989.

16. U.S. Air Force, USAF Stability and Control DATCOM, Flight Control Division, Air Force Flight Dynamics

Laboratory, Wright-Patterson Air Force Base, OH, April 1978.

12

]4

...[

o

o

o

6O

50

4O

Wingloadlng, 30

Ib/tt 2

2O

10 -

050

Mach0.60

....... 0.705•

I I I I_

60 70 80 90 100 110 120 130 140 x 103

Altitude, ft glo_

Figure 2. Wing loading at high altitudes with lift cocfl3cient of 1.0.

Wind

velocity,knots,true

160

140

120

100

80

6O

4O

20

05O

- -- 90 percentile winds....... 10 percentile winds

m

m

m

m

-

I I I I I I I I I

60 70 80 90 100 110 120 130 140x10 3

Altitude, ft 91o-_o

Figure 3. Genetic winds at high altitudes.

15

2.2

2.0

1.8

Maximum 1.6lift

coefficient 1.4

1.2

1.0

.8

.01

- oC.0.400.60

0.70 _ i

oOS _'S s_ S°S_°S_ o oGSS _B

...o°"° _,y'Y y/J"

.__ ! I I r"l'_-Ill I I I I I Illl I I I I I illl

.10 1.IX) 10.00

Reynolds number, millions910501

Figure 4. Generic trends in maximum lift coefficient.

Maximumlift

coefficient

Minimum ._"

velocity _4().000 ft

10 percent fuel -._ ///..,--"

90 percent fuel _//-70-000_o

Maximum / _.j,,/ ,o,,,.s _.velocity _ ..fy/90.000 ft,.-'-

/"Mach / o..'" _-r/ ..-

o "- ......"...'" ....oo°° • e w

0.40 o'* ,.... ..,,," 100,000 ft

0.70"""

2.2 --

2.0-

1.8 -

1.6 -

1.4 -

1.2 -

1.0-

.8-

o6 I I I IJ]lll I I ! I Illll I I I I Illll

.01 .10 1.00 10.00

Reynolds number, millions910502

Figure 5. Lines of minimum and maximum velocity for stratoplane superimposed on figure 4; Reynolds numberbased on mean chord.

16

Liftcoefficient

3.0 - Mach0

2.5 - - ...... 0.40..... 0.60

2.01.01.5 _/_," , ------- 0.70

-.5

-1.0 I

-10 0

I I L I I I I I I

10 20 30 40 50 60 70 80 90

Angle of attack, deg 910_

Figure 6. Lift coefficient.

Airfoil

profiledrag

coefficient

.100 m

i

m

m

m

m

.010 -n

m

i

i

i

.001

.01

I Stratoplane I

_ht envelol:_ I

I I I Illlll I I I III!11 I I I !11111

•10 1.00 10.00

Reynolds number, millions 11105G4

Figure 7. Generic trends in airfoil profile drag coefficient with lift coefficient of 1.0.

17

Dragcoefficient

1.,I- ,_c,1.4i- .......o._o..... 0.601.2 ....

1.0

.8

.6

.4

.2/

0 1 I I __I I I I I Io

-10 0 10 20 30 40 50 60 70 80 g0

Angle of attack, deg u_o_5

Figure 8. Drag coefficient.

Lift-to-dragratio

0 w

25 -

20 -

15 -

10 -

5 -

0 -

-5 -

-10

i!

Mach

! o....... 0.40..... 0.60.... 0.70

! •

I.:i1:

V,

I I I ! J l 1 I I i

0 10 20 30 40 50 60 70 80 90

Angle of attack, deg g_o_

Figure 9. Longitudinal performance efficiency parameter.

18

Normalforce

coefficient

3.0 -

2.5 -

2.0

1.5 _1.0

0

-1.0-10

! ! I I I I I

0 10 20 30 40 50 60

Angle of attack, deg

Figure 10. Normal force coefficient.

Mach0

....... 0.40

..... 0.60

.... 0.70

I I I

70 80 90

010507

Axialforce

coefficient

.3

.2

.1

0

--°1

-.2 -

ll3 m

rood m

--o5

-10

I I ! I I I I

0 10 20 30 40 50 60


Figure ll. Axial force coefficient.

Mach0

....... 0.40

..... 0.60

.... 0.70

I I !

70 80 90

010598

]9

Pitchingmoment

coefficient

F "°c".1 _- ....... 0.40

/\ ..... o_o

....°°--ol

-.2sl "___- _.:.__.

I-.5

-.6 / I I I I I I I I I I

-10 0 10 20 30 40 50 60 70 80 90

Angle of attack, deg olos_

Figure 12. Pithing moment coefficient.

.010

.005

0

Incremental -.005pitchingmoment

coefficient

per degof elevator

-.010

-.015

-.020

--.02_

-.030-10

m

Mach0

- - ...... 0.40..... 0.60

..... 0.70

/-/m

mmmJ

I I ! I I I

0 10 20 30 40 50


(a) In pitching moment coefficient.

I ! I I

60 70 80 90

9106O0

Figure 13. Elevator control effectiveness.

2O

Incrementalaxial forcecoefficientper deg ofelevator

.0010

.0005

0

--.0005

-.0010

-.0016

-.0020 -

-.0025 -

--.0030

-10

F

m

I I I I I I

0 10 20 30 40 50


(b) In axial force coefficient.

Mach0

....... 0.40

..... 0.60

.... 0.70

! I I I

60 70 80 90

910601

.OO8

.007

.006

Incremental .005normal

force .004coefficientper deg ofelevator .003

.002

.001 -

0-11

Mach0

....... 0.400.60

.... 0.70

n

B

I

0

(c)

I I ! I

10 20 30 40


In normalforcecoefficient.

50 60 70 80 90

91O6O2

Hgure 13. Concluded.

21

or j-'_ I- ....... o,4o

/ ..... 0.60

-.010 _- .... 0.70

--n_ffielent -" .....

stabilizer

-.o._ _--'-'7-0401 --r" I I I ! i i I I I

-10 0 10 20 30 40 50 60 70 80 90

Angle of attack, deg olo_

(a) In pitching moment coefficient.

.0005

0

--,000_

Incremental -.0010axial forcecoefficient -.0015per deg ofstablllzer -.0020

-.0025

--.0030

--,0035

-10

m

Mach0

B _ ....... 0,40

..... 0.60.70

, ! .I J I I I I ....j w

0 10 20 30 40 50 60 70 80 90

Angle of attack, deg g1_o4

Co) In axial force coefficient.

Figure 14. Stabilizer control effectiveness.

22

.016

.014

.012

Incremental .010normal

force .008coefficient

per deg ofstabilizer .006

.004

.002

0-10

Mach0

....... 0.40

..... 0.60

.... 0.70

m

0 10 20 30 40 50 60 70 80 90

Angle of attack, deg slo6o5

(c) In normal force coefficient.

Figure 14. Concluded.

Pitchdamping

due topRch rate

parameter/rad

-25

- 27

-29

-31

-33

-35

-37-

- 41-10

Mach0

....... 0.40

..... 0.60.... 0.70

D

m "'''''''''''''''''''''''''''''''''''''''" .......... "''''''''''"

I I I I I I I I I I0 10 20 30 40 50 60 70 80 90

Angle ofaUack, deg olo6o6

(a) In pitching moment coefficient as a result of pitch rate.

Figure 15. Damping parameter.

23

Normalforce

dampingdue to

pitch rateparameter/

red

22

21

2O

19

18

17

16

Mach0

....... 0.40

..... 0.60

.... 0.70

mRe@elwwwwwwwww...._ewmwwewgwwmwwwwwwwwwwwwwwwwwwwwww.w--w._-w

1 ! I I I I I I I10 20 30 40 50 60 70 80 90

Ang_ of ettack, deg ol0eo?

15 -

14 I-10 0

(I)) In normal force coefficient as a result of pitch rate.


.0016

.0014

.0012

Change in .0010yawing

moment .0008coefficient

par deg ofsidesl_o .0oo6

.0004

.0OO2

- Mach

-- ° ...... 0U40

A

..... 0.60

!

0-10

I I I ! I I ! ! I I

0 10 20 30 40 50 60 70 80 90

Angle of attack, deg G106o8

Figure 16. Directional stability derivative.

24

-.001

--°002

_o003

Change In -.004side forcecoefficient -.005per deg ofsideslip -.006

--.007

--.008

-.009-10 0 10 20 30 40 50 60


Figure 17. l_ma_ side force stability derivative.

-- Mach0

....... 0.40

..... 0.60IT ¢tkCi i ¢£-faT

I I I I I I I I I I

70 80 90

O1O6OO

Change Inrolling

momentcoefficientper deg of

sideslip

m

miO005

-.0010 -

-.0015 -

-.0020 -

-.0025 -

_!0030 m

_J0O3_ m

--.0040

-10

\

Mach0

....... 0.40..... 0.60.... 0.70

I I I I I I

0 10 20 30 40 50


(a) Full calculated value.

Figure 18. Effective dihedral derivative.

I I I I

60 70 80 90

910610

25

ChangeInrolllng

momentcoefficientperclegofsideslip

II!II!I

I I I I I I

0 10 20 30 40 50


Co) One-fourth of calculated value.


Mach0

....... 0.40..... 0.60.... 0.70

! I I I

60 70 80 90

g10611

Incremental

rollingmoment

coefficient

per deg ofaileron

.OO8

.007

.006

.005

.004

.003

.002

.001

Mach0

....... 0.40

..... 0.60

.... 0.70

Figure 19.

0 10 20 30 40 50 60


(a) In ro]ling moment coefficient.

Aileron control effectiveness.

70 80 90

910612

26

.0012

.0010

.OOO8

Incremental.0006yawingmoment .0004

coefficientper deg of

aileron .0002

0

--,0002 --

--.0004

-10

I _." /'_, I ;, ,,/!: _."°°/ tl;

I I I I I ' I0 10 20 30 40 50 60 70


Co) In yawing moment coefficienL

Mach0

....... 0.40

..... 0.60

.... 0.70

I I

80 90

910613

Incrementalside forcecoefficientper deg of

aileron

.0040 - Mach0

.0035 - - ...... 0.40..... 0.60

.0030 - 0.70

.0025 -

LI

.OolsL l.;//'X, \

"lOfy't0/I I I I-10 0 10 20

I I I30 40 S0


(c) In side force coefficient.

[ I I !

60 70 80 90

I)10614

Rgun: 19. Concluded.

27

.00006

.O00O5

.O0OO4

Incmmenta! .00003rolling

moment.00002

coefficientper deg of

rudder .00001

0

-.00001

m uun

.--.:.-._..

Mach0

....... 0.400.60

.... 0.70

I_I. tit; t

I t , |li__l

] | I I ! I0 10 20 30 4O 50

Angle of attack, (:leg

(a) In rolling moment coefficient.

J ! I I

60 70 80 90

910615

,00010

.OOOO5

0

Incremental -.00005yawing

momentcoefficientper deg of

rudder

-.OOO10 -

-.00O15 -

_eO0020 m

-.00025 -

--.00030

-10

m

Mach0

- - ...... 0.40..... 0.60

0.70

- -...--__,

I I I I I I

0 10 20 30 40 50


_) In yawing moment coefficient.

R_u-e 20. Rudder control effectiveness.

I I I I60 70 80 90

910616

28

.0016

-.0014

-.0012

Incremental-.0010sideforcecoefficient -.0008perdegof

rudder--.0006

-.0004

--.0002 --

0-10

Mach0

....... 0.40

..... 0.60

.... 0.70

i

I I I I

0 10 20 30


40 50 60 70 80 90

9t0617

(c) In side force coefficient.


Rolldamping

as a result ofroll rate

parameter/md

.4 _ Mach.2 ....... 0040

O- --0.

I ! I I I l ! I

20 30 40 50 60 70 80 90

Angle of attack, deg olo61o

I!I'-.4 - i!I!

-.e - I?/

-1.0 - _.//,.I

-1.2 -" I I

-10 0 10

(a) In rolling moment coefficient as a result of roll rate.

Figure 21. Damping parameters.

29

Yawdamping

as a result ofroll rate

parameter/red

.2

.1

0

--.1

--.2

--.3

-.4

-.5

--.6

-10

Co)

i

/L I1:

_i;\ ::! /

Mach0

....... 0.40

..... 0.60

.... 0.70

I I ! I I [ I I I I0 10 20 30 40 50 60 70 80 90

Angle of attack, deg oloelQ

In yawing moment coefficient as a result of roll rate.

Rolldamping

as 8 result ofyaw rate

parameter/rad

.18 F Mach0

#!.i:: ....... 0.40l_z_ _ ..... 0.60

I .q,_| III it. : _ .... 0.700:J I :/

/ !11I lil

ItlI''

I i.i, I I I I r i

-10 0 10 20 30 40 50 60 70

Angle of attack, deg ut_

.16

.14

.12

.10 ifi

(c) In rolling moment coefficient as a result of yaw rate.

Figure 21. Continued.

I !

80 90

3O

Yawdamping

asa result of

yaw rateparameter/

rad

Mach0

....... 0.40

..... 0.60

.... 0.70

I I I I I I I

0 10 20 30 40 50 60


(d) In yawing moment coefficient as a result of yaw rate.

I I I

70 80 90

910621


Rollingmoment

coefficientbias

.0030 -

.0025

.0020

.0015 -

.0010 -

.0005 --

O-

--eO005 m

-.0010 I-10 0

Figure 22.

B

I I I I I I

10 20 3O 40 50 60


Rolling moment coefficient trim bias.

Mach0

....... 0.40

..... 0.60

.... 0.70

I I I

70 80 90

910622

31

Yawingmoment

coefficientbias

|

It ;I [

#1;* "

,,,f;.I II,;, I|Ill.:. III ,I;: .11ill_ :ItIll b IIII_il tlI#1 I !1

Mach0

....... 0.40

..... 0.60.... 0.70

.00030 -

.00025 -

.00020 -

.00015 --

.00010 -

•00005 --

0 -

we00005 m

-.00010 I I I l I i I-11 0 10 20 30 40 50 60


Figure 23. Yawing moment coefficient trim bias.

I I !

70 80 90

91O623

.O03O

.0025

.0020

.0015

Side forcecoefficient .0010

bias

.O005

0

--.0005

-.0010-10

!i -n

'VII1|.iI_ II.I, i: I P.II I I.o, li Ili t! rl

Mach0

....... 0.40

..... 0.60

.... 0._

I I I I I I I I

0 10 20 30 40 50 60 70


Figure 24. Side force coefficient trim bias.

I I80 90

910624

32

Propellerefficiency

1.1

1.0

.9

.8

.7

.6

.5

.4

.3

.2

.1

- RPM_ 500

....... 250

Blade stall

0

m •

i •

m t, ,m _,

i

I I I ! I I 1 I I I I I

0 10 20 30 40 50 60 70 80 90 100 110 120x103

Altitude, ff91O625

Figure 25. Propeller efficiency.

- 4

Short-periodpole

X

Phugoidpole

I

I I I I I I I I-7 -6 -5 -4 -3 -2 -1 0 1 2

910_26

Figure 26. Unaugmented longitudinal eigenvalues for low-altitude flight condition.

2

33

-8

Rollpole

L- 14

Dutchrollpole

X

X u

I I I J I I !-12 -10 -8 -6 -4 -2 2 4

O1O627

Spiralpole

,,,, ,,,,,

0

Figure 27. Unaugmented lateral-directional eigenvalues for low-altitude fright condition.

RollX

pole

| | I I I ! •

-.70 -.60 -.50 -.40 -.30 -.20 -.10

Splmlpole

XI

.0 .10

w

.20910528

.7

.6

.5

.4

.3

.2

.1

0

Figure 28. Closeup of unaugmented dutch roll and spiral eigenvalues for low-altitude flight condition.

34

Short-periodpole

X

Phugoidpole

I I I I I I I I-2.8 -2.4 -2.0 -1.6 -1.2 -.8 -.4 0 4

- 2.4

- 2.0

- 1.6

- 1.2

-- .8

-- .4

- 0

891O629

Figure 29. Unaugmented longitudinal eigenvalues for high-altitude flight condition.

- 1.0

Roll pole

X

I I- 1.6 - 1.4

Figure 30.

w

! I I I I I

- 1.2 - 1.0 - .8 - .6 -.4 -.2 0 .2 .491063O

XDutchrollpole

Spiralpole

<

I

Unaugmented lateral-directional eigenvalues for high-altitude flight condition.

.8

.6

.4

.2

0

35

Short-period _ILmode

Phugoid

modeI I I I I I I ..

-2.8 -2.4 -2.0 -1.6 -1.9. -.8 -.4 0

- 2.4

- 2.0

- 1.6

- 1.2

.8

.4

- 0

I4 8

910831

Figure 31. Root loci of flightpath angle to elevator feedback for high-altitude flight condition; gain variation from

Oto 1.0.

Shorl-period 0.4 0.3 0.2 0.1mode 0.5 A a A A . . A A X

A A

I I I l I I-1.4 -1.2 -1.0 -_ -.6 -.4

Phugoldmode

0 0I

-.2I

0 .2

Figure 32. Pitch rate to elevator; gain variation from 0 to 0.5.

m

g

.4

g10632

2.2

2.0

1.8

1.6

1.4

1.2

1.0

.8

.6

.4

.2

0

36

Pilot

I AlmraftElevator "-

Pitchrate

Figure 33. Longitudinal control augmentation with nominal high-altitude gains.

y

Flight-pathangle

910633

Dutch

roll 0.2mode

0.4A

A

Spiralmode

0.4 0.2A AAAAA_ AAX

I I I I I I I I

- 1.4 - 1.2 - 1.0 - 8 - 6 --.4 -.2 0 .2

- 1.0

- .8

.6

m ,4

.2

- 0

.491O634

(a) Gain variation from 0 to 0.5; roll mode near -13; low-altitude flight condition.

Figure 34. Root loci of bank angle to aileron feedback.

37

- 1.0

Roll mode 0.1 0.2 02X_/_AAA A A _ A A A

I I ! I I I I-- 1.6 -- 1.4 -- 1.2 -- 1.0 --.8 --.6 --.4

Dutchrollmode

A

0.1

Spiralmode0.1

I I-.2 0 .2

Co) Gain variation from 0 to 0.2; high-altitude flight condition.


.4910_5

.8

.6

.4

.2

0

- 1.2

Dutch

roll _mode

Roll mode Spiral0.4 0.3 0.2 0.1 mode

A A A A A A A A A X K

I I I I I I I I- 2.8 - 2.4 - 2.0 - 1,6 - 1.2 -.8 -.4 0 .4 .8

.8

.4

0

91O636

Figure 35. Root loci of roll rate to aileron feedback for high-altitude flight condition; gain variation from 0 to 0.45.

38

Dutchrollmode

Washoutfilter

!- 1.0

Spiralmode

_k

I I I I I I I-1.4 - 1.2 -.8 -.6 -.4 -.2 0 .2 .4

910637

1.0

.8

.6

.4

.2

0

Figure 36. Root loci of yaw rate to rudder feedback for high-altitude fright condition; gain variation from 0 to 0.5;roll mode near 3.46.

Rudderpedals Rudder

Lateralstick

Aileronm,"-

Aircraft

Yawrate

l- D rate

r

i BFigure 37. Lateral-directional control augmentation with nominal high-altitude gains.

Bankangle

y

91O638

39

- 1.0

Dutch

roll 1.0 _mode 0.5

Yaw ratewashoutfilter

¢X

I- 1.0

Spiralmode

(X

I I t ! ! I- 1.2 -.8 -.6 -.4 -.2 0 .2 .4

910639

(a) Root loci of sideslip angle to rudder feedback; gain variation from 0 to 1.0.

.8

.6

.4

.2

- 1.2

Yaw ratewashout

filter -0.4Oi_ A A A

I I I I !-1.4 -1.2 -1.0 -.8 -.6

Dutchrollmode

Spiralmode

*, A

!--°4

XA

-0.1 AA

-0.2 A

-- 0.3 A

AA --0.4 _A

--0.54-0.2* A_X

I I-.2 0 .2 .4

91064O

1.0

.8

.6

.4

.2

(b) Root loci of sideslip angle to aileron; gain variation from 0 to - 1.0.

Figure 38. Roll rate, yaw rate, and bank angle loops closed with nominal high-altitude gains for high-altitude

flight condition.

40

Form ApprovedREPORT DOCUMENTATION PAGE OUBno.o_-olss

Pubic I'_Bofl_ burdenior thlmoo4teoNonol informationis astimmed Io IlVerl_e 1 hourper teen,nee, includingIhe time br reviewln0_r_trucllons, Matching ezlellng datatouP...es..and ma/ettaln_tgINs data needed, aria completingand revtewln0 the collectionof Infomufllon. hncl comment= regardingINs burdeneltlmee Orany other mpec_ of this

¢mm¢x,mnof Inlo_m.allon_k_i_.inll m,N)ges#ocefoeredudn_ lidsburden, to Washingtonkieadquartam Secvk:as. Dk'ectorale Ior Info_mallonOperationsand Reports. 121S Jeffersonurns _qlnway, _,,ulte1204, Adinglon,VA 22202-4.102. imlto the Office of Manllgemenlw_lBudgeL PapensowkReductionProject(0704-01|B), Washington. DC 20503.

1. AGENCY USE ONLY (Lmave blank) 2. REPORT DATE ....... 3:'_'REI=ORT TYPE AND DATES COVERED

September 1991 Technical Memorandum4. TITLE AND SUBTITLE &. FUNDING NUMBERS

Modeling, Simulation, and Flight Characteristics of an Aircraft Designed toFly at 100,000 Feet

........ WU-464-99-01-00S. AUTHOR(S)

Alex G. Sim

?. PERFORMINGORGANIZATIONNAME(S)ANDADDRESS(ES)

NASA Dryden Flight Research FacilityP.O. Box 273

Edwards, California 93523-0273

0. SPONSORING/MONITORING AGENCY NAME(S) AND ADDRESS(ES)

National Aeronautics and Space Administration

Washington, IX: 20546-0001

i8. PERFORMING ORGANIZATIONREPORT NUMBER

H-1731

10. SPONSORING/MONITORINGAGENCY REPORT NUMBER

NASA TM- 104236

11. SUPPLEMENTARY NOTES

12s. DISTRIBUTION/AVAILABILITY STATEMENT

Unclassified -- Unlimited

Subject Category 08

15. ABSTRACT (Maximum 200 words)

12b. DISTRIBUTION CODE

A manned real-time simulation ofa conc,qxualvehicle, the stratoplane, was developed to study problems associated

wiah the flight chamacri_cs of a large, lightweight vehicle. Matlzmadcal models of the aerodynamics, massproperties, and propulsion system were developedin support of the simulation and ate pcsented. The simulation was

at first conductedwithout control augmentation to determine the needs fora conaol system. The unaugmentedflyingqualities wcr¢ domina_ by lightly damped dutch mR oscil_ons. Constantpilot workloads were needed at highaltitudes. Control augmentationwas investigatedusingbasic feedbacks. For the longitudinal axis, flightpath angleand pitch rate feedback were sufficient to damp the phugoid mode and to provide good flying qualifies. In the lateral-

directional axis, bank angle, mll rate, and yaw rate feedbacks were sufficient to provide a safe vehicle with acceptable

handling qualifies. Imentionally stalling file strampiane to very high angles of attack (deep stall) was investigated as

a means to enable safe and rapid descent. It was concluded that the deep-stall maneuver is viable for this classof vehicle.

14. SUBJECT TERMS

Deep s__.gh-aldn,de flight

Hi_,h an_lc of attack_7. SECURITYCLAS_IFICATI0b'

OF REPORT

Unclassified

ISN 7540-O 1-_-_00

_s. SECURITYCLASSIFICATION_s. SECUmTYCLASSIFICATIONOFTHISPAGE OFABSTRACT

Unclassified

15. NUMBER OF PAGES

4416. PRICE CODE

A0320. LIMITATION OF ABSTRAC'T

Standard Form 298 (Rev. 2-8S)II_maorlbe_by ANSI IRd. Z30-t IIO4-t 02

modeling, simulation, and flight characteristics of … · modeling, simulation, and flight...

Documents