Control-system techniques for improved departure/spin resistance for fighter aircraft

c. 1 "

Paper 1689

NASA Technical ' \

KIM C O P Y : (

M - T T E C ~ ~ ~ KIRTLA~JD :A/

. I =

. o a \ 0 - p" - d B ' b= 0 - P m

-

Control-System, Techniques,.

i " 1

-. .

-

for Improved~. DepartureEpin - - r l.:::I . . . . -

: Resistance for Fighter:,Aircraft , . .I

f , .

' \ , , , . , , , . . / Luat T. Nguyen, William P. Gilbert, and Marilyn E. Ogburn I AUGUST 1980' ~ .

, . .

TECH LIBRARY KAFB, N M OD6780L

NASA Technical Paper 1689

Control-System Techniques

for Improved DepartureEpin

Resistance for Fighter Aircraft

Luat T. Nguyen, William P. Gilbert, and Marilyn E. Ogburn Langley Research Ce~zter Hamnptou, Virgirria National Aeronautics and Space Administration Scientific and Technical Information Branch SUMMARY This paper summarizes some fundamental information on control-system effects on controllability of highly maneuverable aircraft at high angles of attack and techniques for enhancing fighter aircraft departure/spin resistance using control-system design. The discussion includes (1) a brief review of pertinent high-angle-of-attack phenomena including aerodynamics, inertia coupling, and kinematic coupling, (2) effects of conven- tional stability augmentation systems at high angle of attack, (3) high-angle-of-attack control-system concepts designed to enhance departure/spin resistance, and (4) the out- look for applications of these concepts to future fighters, particularly those designs which incorporate relaxed static stability.

INTRODUCTION Designers of modern fighter aircraft are faced with the requirement of providing extreme maneuverability and a high level of departure/spin resistance. Two general design approaches are followed to attain this goal. These are (1) aerodynamic design, which involves detailed configuring of the basic airframe geometry to provide inherently good high-angle-of-attack characteristics, and (2) control-system design. In recent years, increasing emphasis has been placedon the second approach because of rapid advances in avionic technology which have made the implementation of complex control laws more practical. As a result of this design philosophy, modern fighters such as the F-15, F-16, and F-18 all incorporate control-system features that enhance maneuverability at high angles of attack and greatly reduce susceptibility to departures and spins.

Because of its unique testing techniques and experience in stall/spin technology, the NASA Langley Research Center has been active in developing the use of control-system design to enhance the departure/spin resistance of fighter aircraft. This paper is a summary of the experience gained at Langley in this area as a result of numerous piloted simulation studies, subscale dynamic model tests, and full-scale flight tests of modern fighteraircraft.Somefundamentalhigh-angle-of-attackaerodynamic,inertia-coupling, and kinematic-coupling phenomena are reviewed to establish some of the basic factors involved. Next, the effect on spin resistance of conventional stability augmentation s y s - tems not designed for high-angle-of-attack flight conditions is addressed. Simple roll and yaw dampers, for example, fall into this category and, although they are generally low-gain, limited-authority devices, their effects on departure/spin resistance can b e significant. Next, specific high-angle-of-attack control concepts, many of which are currently being used in the latest fighters, are discussed. These concepts include con- trol crossfeeding, stability-axis yaw damping, static stability augmentation, angle-of- attack limiting, and automatic spin prevention. The rationale for each system is discussed and the advantages and potential limitations are reviewed. Finally, application of the technique to future configurations incorporating advanced concepts such as relaxed static longitudinal stability is addressed. The analysis presented in this paper is con- ducted using a set of aerodynamic stability and control and inertial characteristics that are representative of many current fighter airplane designs. In this report, this set of aerodynamic and inertial characteristics is referred to as the fighter configuration, the basic configuration, or a similar description.

SYMBOLS All aerodynamic data and flight motions are referenced to the body system of axes shown in figure 1. The units for physical quantities used herein are presented in the International System of Units (SI) and U.S. Customary Units. The measurements and calculations were made in U.S. Customary Units.

an normal acceleration, positive along negative Z body axis, g units (lg = 9.8 m/sec2) b wing span, m (ft) Aerodynamic lift force lift coefficient, CL rolling-momentcoefficientabout X body axis, Aerodynamic rolling moment

w

Cm pitching-moment coefficient about Y body axis, Aerodynamic pitching moment yawing-momentcoefficientabout Z body axis, Cn Aerodynamic yawing moment

qsb

Y-axis force coefficient along positive Y body axis, Aerodynamic Y-axis force CIS wing mean aerodynamic chord, m (ft) acceleration due to gravity, m/sec2 (ft/sec2) altitude,m (ft) moments of inertiaabout X, Y , and Z body axes, kg-m2(slug-ft2) gearing ratio between yaw and roll controls static gain, deg/(deg/sec) static gain, deg/deg Mach number pitching moment due to inertia coupling,

(Iz - I x) pr, N-m (ft-lb)

yawing moment due to inertia coupling, pq, N-m (ft-lb) airplane mass, kg (slugs) criticalvalues of rollratecorrespondingto CI values Q1 and q2 where p* < pi,2,deg/sec or rad/sec s 9 1 airplane roll rate about X body axis, deg/sec or rad/sec stability-axis roll rate, deg/sec or rad/sec roll angular acceleration due to ARI, deg/sec2 or rad/sec2 airplane pitch rate about Y b.ody axis, deg/sec or rad/sec free-stream dynamic pressure, N/m2 (lb/ft2) specific values of 4 where t j l tj2 airplaneyaw rate about Z body axis, deg/sec or rad/sec stability-axis yaw rate, deg/sec or rad/sec rS yaw angular acceleration due to ARI, deg/sec2 or rad/sec2 :AR1 S wing area, m2 (ft2) t time, sec time to damp to one-half amplitude, sec components of airplanevelocityalong X, Y, and Z body axes, m/sec(ft/sec) airplane resultant velocity, m/sec (ft/sec) airplane body axes (see fig. 1) angle of attack, deg angle of sideslip, deg P aileron deflection, positive for left aileron deflected with trailing edge up, deg 6a combined yaw and roll-control deflections obtained by ARI, deg 6AFiI differential horizontal-tail deflection, positive for left tail deflected with 6d trailing edge up, deg horizontal-tail deflection, positive for tail deflected with trailing edge down, deg lateral stick deflection, positive for right roll command, cm (in.)

1s rudder pedal deflection, positive for right yaw command, cm (in.)

Ped rudder deflection, positive for rudder surface deflected with trailing edge 6r left, deg e,+,+ Euler angles, deg P air density, kg/m3 (slugs/ft3) 52 total airplane angular rate vector, deg/sec Stability derivatives: a cZ a cZ a cZ - " " CZ = -

% =q

Pb ' Z r a P a - 2v ' d a - a cZ a .

-

" c2. =- cn =-

cZ cz -a8, Pb a -

'r 6 A R I a6ARI hJ P a -

2v 2v Subscripts: aug ented crit ical DR Dutch roll Abbreviations: ARI aileron-rudder interconnect CAS command augmentation system

ccv control-configured vehicle

KIAS knots indicated airspeed LCDP lateral control divergence parameter LSRI lateral-stick-to-rudder interconnect PI0 pilot-induced oscillation RSS relaxed static stability A dot over a variable denotes a derivative with respect to time.

An arrow over a variable denotes a vector.

BACKGROUND Aerodynamic Considerations The following discussion is provided to briefly identify some of the fundamental and important high-angle-of-attack aerodynamic characteristics that must be addressed in designing a control system for these flight conditions. It is emphasized that this summary is not intended to be a comprehensive treatment of the subject. (See ref. 1 for a more detailed overview.) The discussion of aerodynamics presented herein is limited to one set of aerodynamics which exhibits general stability and control trends that are representative of many current fighter designs. Also, the particular discussion in this section is limited to configurations designed to have inherent longitudinal static stability (not requiring artificial static stability augmentation). A subsequent section of this paper, "Future Concepts and Applications," addresses airplane configurations employing control-configured-vehicle (CCV) concepts wherein the airframe is purposely designed to have low or negative static longitudinal stability. It is believed that analysis of such representative characteristics provides a very useful illustration of the effects of various control-system design features. This paper is not intended to be an analysis of the par- ticular problems of any one specific airplane design.

Shown in figure 2 are low-speed wind-tunnel lift and pitching-moment data for the subject airplane with neutral controls. The lift curve is nearly linear up to o ! = 15', where a change in slope indicates progressive stall of the outer wing. Note that maximum lift does not occur until about o ! = 30°, s o that the desire for maximum maneuverability dictates the need to be able to fly safely and effectively up to this angle of attack. The pitching-moment curve shows satisfactory levels of stability at high angles of attack.

In general, longitudinal stability at high angles of attack is not a significant problem for the current generation of fighter aircraft except those that employ CCV concepts wherein the airframe is purposely designed to exhibit low or negative aerodynamic stability. The problem may be more severe, however, for future designs, particularly those with highly swept wings designed for supersonic maneuverability. The impact of CCV concepts on high-angle-of-attack pitch control is addressed in a subsequent section of this paper.

The variation of static lateral-directional stability with o ! for the airplane is shown in figure 3 for two values of Mach number, 0.15 and 0.9. The low-speed data indi- cate a rapiddegradationindirectionalstabilityabove CY = 1 0 ' suchthatCn is zero P at CY =: 1 6 ' andexhibitshighlyunstablevaluesathigherangles of attack. The reduction of Cn and marked changes to be discussed for other stability derivatives are results P of aerodynamic interference caused by stalled and vortex flow. Lateral stability ' Cl P is maintained at a high level with increasing angle of attack such that the airplane is dynamically stable throughout the angle-of-attack range as indicated by the Cn P7dyn parameter. The Cn parameter has been used in past investigations an an indica- P,dyn tion of the existence of directional divergence ("nose slice") at high angles of attack.

Negative values of this parameter usually indicate the existence of a divergence. The data presented for M = 0.9 show that both the lateral and the directional stability can be noticeablydegraded at transonicMachnumbers. At M = 0.9, Cn is negativeabove P o ! = 13', and 9 , values are significantly less stable than those exhibited by the config- P uration at low Mach number. Near C Y = 2 0 ° , the stability characteristics have degraded tothepoint that Cn is negative,whichindicatesthepossibility of a divergence.

P7dyn Turning to lateral-directional control characteristics, rudder effectiveness is plotted versus angle of attack in figure 4. At low speeds, the data show a rapid reduction in Cn above C Y = 2 0 ° , suchthattherudder is completelyineffectiveby o ! = 40'.

6 , Comparison with the M = 0.9 data indicates significantly lower rudder effectiveness at the high-speed condition. These characteristics indicate that it would be difficult to maneuver the airplane effectively by using the rudder above a! =: 3 0 ' because of the lack of yaw control.

Figure 5 shows the rolling- and yawing-moment derivatives due to differential deflection of the horizontal stabilators used for roll control on the subject configuration.

The low-speed data at M = 0.15 .indicate that the rolling-moment effectiveness of these controls is maintained to high angles of attack and that proverse yawing moment is pro- ducedupto a! = 3 2 ' . The transonicdatapresentedfor M = 0.9 shows a degradation a marked reduction in rolling-moment derivative in these characteristics with 6d above a! = 2 0 ' and a change of yawing-momentderivativeCfromproverseto n6d adverse at significantly lower angles of attack. The overall roll effectiveness of the air- plane can be appraised by using the lateral control divergence parameter (LCDP), which is defined as C

- n6d

LCDP c2 6d A negative value of this parameter is indicative of roll reversal. When a roll reversal is encountered, a right roll-control input by the pilot will cause the airplane to roll to the left. Computed values of LCDP for the subjekt airplane are shownin figure 6. The results for M = 0.15 indicate that reversed response will be encountered above a! = 2Z0, due primarily to the high level of static directional instability discussed previously. The degraded stability and control characteristics at M = 0.9 are reflected in the LCDP values, which predict that reversed response will occur above a! = 16'. These results indicate the importance of considering Mach number effects in designing a control system for these flight conditions.

The dynamic derivatives produced by rolling, yawing, and pitching are also impor- tant parameters in high-angle-of-attack flight dynamics and must certainly be accounted for in the control-system design process. Unfortunately, these derivatives are not as readily obtainable as the static stability data discussed previously. Figure 7 shows the variation with angle of attack of the roll and yaw damping derivatives for the subject con- figuration measured during low-speed wind-tunnel forced-oscillation tests. The yaw- damping data show that the level of damping is maintained with increasing angle of attack up to and beyond maximum lift. Roll damping, on the other hand, experiences a sharp

reduction above CY = loo due to progressive wing stall. The low values of roll damping

for CY 2 2 0 ' results in a poorly damped Dutch roll which degrades tracking perform- ance. In situations such as this, designing the control system to alleviate the problem is an attractive potential solution.

Kinematic and Inertia Coupling Considerations Coupling resulting from kinematic and inertia effects are important in varying degrees to the high-angle-of-attack flight dynamics of all modern fighter aircraft. The following discussion is a brief review of several of these phenomena that are particularly significant to the control-system design process.

Figure 8 illustrates the kinematic coupling between angle of attack and angle of sideslip that occurs when an airplane is rolled about its longitudinal axis, or X-axis, at high angles of attack. If the airplane is flying at angle of attack with the wings level and the pilot initiates a pure rolling motion about the airplane X-axis, all the initial angle of attack will have been converted into sideslip after 90'of roll. Because it is undesirable to generate large amounts of sideslip at high angles of attack from a roll-performance as well as a departure-susceptibility viewpoint, most current fighters are designed to roll more nearly about the velocity vector than the body axis. It is obvious that this conicalrotationalmotion(indicatedby ps) eliminatesthecouplingbetween CY and 0.

Resolving ps into the body-axis system shows that this motion involves body-axis yaw rate as well as roll rate and that these rates are related by the expression r = p tan CY.

If this equality is not satisfied during a roll, sideslip is generated as a result of kine- matic coupling with varying as = p sin CY - r cos a.

In the case of rolling with an initial sideslip, it is seen from figure 8 that body-axis rolling results in the initial sideslip being converted into angle of attack after 90'of roll

iu varying as 6 =: q - p cos CY tan p. The second term of this expression indicates

with thatrollingwithadversesideslip(pand p havingthesamesigns)tendstoreduce a, whereas rolling with proverse sideslip (p and p havingoppositesigns)tendsto

increase CY. This latter effect can be an important consideration in that substantial

increases in a can be generated as a result of kinematic coupling if the airplane is rolled with proverse p (using excessive rudder for example).

The second form of coupling important to the high-angle-of-attack dynamics of modern fighter aircraft is a result of inertial effects. Figure 9(a) illustrates the well- known inertial pitching moment that is produced when a typically fuselage-heavy fighter airplane is rolled about its velocity vector at high angles of attack. The desirability of this type of roll from theviewpoint of kinematic coupling was discussed previously; how- ever, the nose-up pitching moment caused by inertia coupling can also be an important consideration. As an aid in visualizing this effect, the fuselage-heavy mass distribution of the airplane is represented as a dumbbell with the mass concentrated at the two ends.

If the airplane rolls about its velocity vector, the dumbbell tends to pitch up to align itself perpendicular to the rotation vector ps. This pitching moment due to inertia coupling Mic can be expressed as Substituting p = p cos (Y S and r = p s i n (Y S gives

~i~ = (Iz - 1 ~ ) p ~ 2 cos a sin a = : ( I 2 z - 1 ~ ) p ~ 2 s i n 2 0

The previous expression shows that the inertia-coupling pitching moment resulting from stability-axis rolling is always positive (nose up) for positive angle of attack, increases with increasing angle of attack, and varies as the square of the roll rate. Thus, if high r a t e s of roll can be generated at high anglesof attack, significant nose-up moments are produced which cause uncommanded increases in angle of attack and, if longitudinal stability is low, can lead to loss of control.

The inertia-coupling yawing moment which results from the combinationof roll and pitch rates is illustrated in figure 9(b). The airplane mass distribution is represented by the dumbbell and the airplane is shown rolling to the right and pitching up. As can be seen, the dumbbell tends to yaw nose-left to align itself perpendicular to the rotation vector 5. Thus, the airplane would be rolling and yawing in opposite directions.

Recalling that to minimize adverse sideslip due to kinematic coupling r must be equal to p tan a, it is seen that this form of coupling can contribute to the bu'ildup of large amounts of adverse sideslip which in turn can result in loss of lateral-directional control a t high angles of attack.

The discussion in this section is an attempt to briefly highlight some aerodynamic, kinematic, and inertial phenomena important to the high-angle-of-attack flight dynamics of modern fighter aircraft. These phenomena must therefore be addressed in designing the control system if it is to be used to enhance departure/spin resistance. When these considerations are not included in the design process, the resulting system can often have detrimental effects on departure/spin resistance. Examples of this are covered in the following section, which is a discussion of the effects of conventional stability aug- mentation systems not specifically designed for high-angle-of-attack flight conditions.

EFFECTS OF CONVENTIONAL CONTROL SYSTEMS AT HIGH ANGLES O F ATTACK Pitch Damper As discussed previously, current fighters designed to have a satisfactory level of static longitudinal stability at low angles of attack do not generally experience significant degradation of these characteristics with increasing angle of attack. Furthermore, con- ventional pitch dampers are normally designed with relatively low gains and very limited authority; therefore, such systems generally do not significantly affect departure/spin characteristics.

Yaw Damper Yaw dampers designed to augment Dutch roll damping at low-angle-of-attack flight conditions are usually implemented by feeding aircraft yaw rate through a washout (high- pass) filter to drive the yaw control within some prescribed authority limit. At low bf the yaw damper to enhance Dutch roll damping tends to decrease speeds, the ability with increasing angle of attack for several reasons. As discussed previously, the com- bination of low dynamic pressure and the tendency toward loss of rudder effectiveness at high angles of attack is certainly a contributing factor. However, an equally important the previous section, is the decrease in static directional stability effect, discussed in withincreasingangle of attack.ThisreductioninCcombinedwithsustained CL "P P causes the Dutch roll mode to become a rolling oscillation about the longitudinal body axis with little yaw rate involved. As a result, increasing the effective has little impact 'nr on the Dutch roll motion at these conditions. However, in the extreme condition where and CL are degraded to the point that the airplane exhibits a yaw departure cnP P (nose slice), the damper is beneficial in that it opposes the motion and, therefore, slows the rate of divergence. This effect is generally small at low speeds because of the lack of rudder power at low values of 4. At higher speeds, however, the beneficial effect of the damper can be stronger.

Figure 10 is a summary of the effects of a yaw damper applied to the fighter config- uration discussed in the previous section. Damping of the Dutch roll mode is computed l l l l l Ill I I I 1 111111111l1l11l1l1l f o r l g t r i m (low speed)and M = 0.9 and is expressed in terms of the parameter l / t l , 2 .

The low-speed results for the basic airplane show a decrease in damping with increasing angle of attacksuchthatthemode is undamped(unstable)for CY aboveabout 17'. Addi- tion of the yaw damper results in a very significant improvement in stability at the lower angles of attack; with increasing angle of attack, however, the effectiveness drops off rapidly. At M = 0.9, the basic airplane experiences a sharp loss in stability for LY > 15' because Cn becomes highlynegative. By CY = 2 0 ° , theDutch roll mode is P quite unstable, which indicates the existence of a divergence. Again, addition of the yaw damper greatly improves Dutch roll damping at the lower angles of attack. In this case, however, the effectiveness of the device does not decrease with increasing angle of attack as rapidly as in the low-speed case. At CY = 2 0 ° , the yaw damper significantly reduces the level of instability and the resulting departure is therefore expected to be milder.

This characteristic is verified in figure 11, which shows time histories of a split-S maneuver in which the airplane was maneuvered into the instability at M = 0.9 and LY = 2 0 ' . The unaugmented airplane experienced a fairly rapid departure "over the top" of the turn before decelerating out of the instability region. As predicted previously, the airplane with the yaw damper active encountered a much milder uncommanded motion as it passed through the unstable area.

An additional high-angle-of-attack effect of yaw dampers is related to kinematic coupling. A s discussed previously, minimization of adverse sideslip generation during rolls at high angles of attack requires that the airplane roll about the velocity vector which involves body-axis yaw r a t e as well as roll rate. Because the damper opposes any it tends to make the airplane roll about the body axis rather than the velocity yaw rate, vector, and thus is detrimental to roll coordination. The severity of this effect is more apparent at high-speed than at low-speed flight conditions.

Roll Damper Conventional roll dampers are also generally designed to enhance low-angle-of-

attack flying qualities - in this case to quicken the roll mode. However, the influence

of the roll damper on high-angle-of-attack characteristics can be much stronger than that of the yaw damper. There are several reasons for this characteristic. One reason is that roll controls, particularly differential tail, can produce strong yawing and rolling moments at high angles of attack, whereas rudders generally tend to lose effectiveness.

A second important factor is the combination of high-angle-of-attack aerodynamics and kinematics. A s mentioned in the previous section, the Dutch roll mode becomes pri- marily a rolling motion at high angles of attack; therefore, a roll damper is more effective in stabilizing this primary lateral-directional mode of motion. Some further insight into this phenomenon can be gained by examining an expression for the damping of the Dutch I I I, I roll. In reference 2, an approximation for the damping index of this mode is developed and, with some manipulation, the damping can be expressed in terms of the parameter 1/tl,2 as follows: An indication of the variation of damping with roll damping is obtained by taking the : derivative of the above expression with respect to Cl P For a stableconfiguration Cn > 0, thisexpressionshowsthatthesensitivity of PydYn Dutch roll damping to CL increases withangle of attack and varies withthelevel of P dihedral effect. Higher dihedral effect results in a greater increase in Dutch roll damping due to increased roll damping. This result substantiates the intuitive explanation given earlier that high levels of dihedral effect at high angles of attack tend to make the Dutch roll more of a rolling oscillation and, therefore, increasing roll damping would be expected to be effective in damping the mode. Thus, under these conditions, a roll damper designed for low angle of'attack to augment the roll mode damping with little effect on Dutch roll damping tends to have the opposite effect at higher angles of attack.

Of course, the effect of the roll damper on high-angle-of-attack Dutch roll stability is directly related to the system gain (surface deflection per unit roll rate), which is generally low. The upper portion of figure 1 2 illustrates this effect for the example fighter configuration. As discussed previously, at low speeds the airplane exhibits an

undamped Dutch roll mode above a = 17' duetoloss of CI and Cn , while CI

P P P remains high. The oscillation is almost entirely about the roll axis and the t e r m "wing rock" is often used to describe this type of motion. As predicted, addition of a roll damper designed for the airplane has very little effect on the Dutch roll damping at low at higher angles of attack is evident; however, the angles of attack. The beneficial effect improvement in damping is small because of the low gain of the system. When the gain is quadrupled, however, the data show that the damping above a = 10' is greatly aug- mented, to the extent that the wing-rock tendency is eliminated.

Although roll dampers can have a stabilizing effect at high angles of attack if indicates good stability and dihedral effect is high, experience has shown that Cn B ,dvn , _ " their effect can become quite the opposite as the basic airframe stability becomes mar-

ginally stable or unstable low o r negative . This characteristic is also indi-

cated by the preceding expression for a ( l/tl/z),$CI , which shows that if

P < 0 while dihedral effect is maintained, increasing roll damping is destabilizing.

Cn P 7 dyn The lower portion of figure 12 illustrates this effect; values of Dutch roll damping are shown for the example airplane at M = 0.9. It is seen that at a = 15' (wherethe configuration is stable), the roll damper enhances stability. However, at a = 20°, where the basic airplane experiences a moderate instability due to a marked loss of static directional stability, the data show that the roll damper, even with its low gain, significantly aggravates the instability. Time histories of split-S excursions through are shown in figure 13. As shown previously, this region with and without the roll damper the basic airplane experiences a moderate departure. Addition of the roll damper, how- ever, results in a much more severe and prolonged loss of control.

The detrimental effects of a roll damper are further aggravated if the driven roll control exhibits adverse yaw. For example, if a departure to the right occurs, the damper applies left roll control which, due to its adverse characteristic, produces addi- tional nose-right yawing moment to drive the departure.

Because of their potential for strong adverse effects at high angles of attack, most conventional roll dampers not designed for these conditions a r e simply deactivated at high angles of attack. Systems that employ high-gain roll dampers to correct specific high-angle-of-attack deficiencies such as wing rock should also incorporate safeguards that allow the dampers to operate when their effects are beneficial and to deactivate when their effects are detrimental to stabilityand departure resistance.

EFFECTS OF PITCH AND ROLL COMMAND AUGMENTATION SYSTEMS Many current and evolving fighter designs use control laws which are more sophis- ticated than simple stability augmentation systems. These concepts, often referred to as maneuver-demand, o r command augmentation systems (CAS), are designed to provide the pilot with airplane response characteristics that remain relatively constant over a large flight envelope. The behavior of an airplane with these control laws can be quite different from that of an airplane using more conventional direct stick-to-surface control. For example, airplanes using such CAS concepts are usually equipped with a control logic s o that a given stick force (or deflection) commands a specific response, such as specified pitch or roll rates and/or normal acceleration response. In a typical pitch CAS (fig. 14), a mixture of pitch rate and normal acceleration is commanded proportional to pilot stick position (or force) to provide a uniform pitch response over a wide angle-of-attack and speed envelope. However, with such a response-command system, for any nonaccelerated flight condition, the stick must be in a neutral position. Whenever the stick is held aft of neutral, the pilot is commanding an accelerated flight condition as opposed to commanding a particular trim angle of attack (which would be the case in an airplane with conventional augmentation). For example, at a fixed power setting (thrust level), consider a pilot rolling such an airplane into a turn, applying a fixed partial aft stick deflection (or force), and holding this control input. If the pilot is commanding an accelerated condition that cannot be sustained with available thrust, the airspeed decreases and the pitch control system commands increasing angle of attack in an attempt to maintain the commanded pitch rate and normal acceleration. If such a system is not authority limited o r angle- of-attack limited, the pitch CAS eventually commands a full pitch-control deflection and increases the airplane angle of attack to the maximum trim angle of attack, usually an angle of attack at or beyond maximum lift. This characteristic is illustrated in the time histories shown in figure 3 5; in this maneuver, the pilot banked the airplane into a turn and applied a step stick input (commanding normal acceleration and pitch rate) near t = 8 sec, establishing an initial angle of attack of 14O. Thereafter, the stick position and power setting were held constant. As the airplane decelerated, the angle of attack increased to over 3 5 ' as the pitch-control system attempted to maintain the commanded pitch rate and normal acceleration. Such behavior can lead to inadvertent stall entries if the pilot is unaware of this control-system characteristic. Moreover, if the airplane exhibits degraded lateral/directional stability and control in the stall, inadvertent loss of control and spin entry may occur.

With regard to the lateral axis, the CAS concept can be used to provide a uniform roll response to lateral stick inputs over a wide range of flight conditions. With this characteristic, as the aileron rolling moment becomes lower becauseof either lower dynamic pressure or reduced control effectiveness, the control system (CAS) drives the 1l1l1l11llll I I l l 1 rnI I I IIIII control surfaces to larger deflections in seeking to provide the commanded roll rate.

Consider, for example, an airplane flying in a decelerating windup turn to track a target airplane. A s the maneuver progresses, the airspeed steadily decreases while the angle of attack is increased to maintain the desired turn rate. A s the pilot attempts to track the target airplane, the aileron deflections increase significantly as the CAS attempts to maintain the constant roll-rate response. If the airplane control system incorporates a roll/yaw interconnect, which deflects the rudders in proportion to aileron deflection for coordination, rudder deflections also increase markedly as the maneuver progresses into the low-speed, high-angle-of-attack regime. A s with the pitch CAS described pre- viously, the pilot is often unaware that such large control deflections are occurring.

When the lateral-directional stability of the airplane is low, as is often the case near maximum lift, such large-amplitude directional control (rudder) deflections can excite oscillations which may degrade fine tracking at high angles of attack.

Another potential adverse characteristic of a roll-rate CAS at high angles of attack arises when the pilot attempts to roll the airplane using only the rudder pedals. The practice of rolling with rudder-pedal inputs is a very common technique in many current fighters where roll response to lateral stick inputs is either low or reversed from the desired direction. In such airplanes, the pilot applies rudder to generate proverse side- slip which produces the desired rolling moment via the dihedral effect of the airplane

Cl is sufficientlystable . In thiscase, whenthe lateralstick is kept.cen-

assuming

( P )

tered (neutral), the roll CAS causes the ailerons to oppose the roll rate generated by the rudders. This CAS response is similar to that of a conventional roll damper system, except more dramatic in that roll CAS systems are often high-gain, full-authority as opposed to low-gain, very limited-authority roll dampers. An example of such a response is shown in figure 16 which shows a roll-reversal attempt using only rudder inputs during a pull-upmaneuver. At t = 6 sec and a ! =: 15O, thepilotappliedfullleftrudderwhile keeping the lateral stick neutral. Although about 7 ' of steady-state sideslip was gen- erated, the roll response was comparatively slow because the roll-rate CAS applied over 2 0 ' of opposite aileron deflection in countering the uncommanded roll rate. In the extreme situation, attempting to roll such an airplane at high anglesof attack with rudder pedals can result in the inadvertent application of nearly full "cross controls" (roll con- trol opposite to rudder), which is a prospin control input for most current fighter air- planes. For conventional airplane designs, it is quite common for the pilot to use the rudder alone for roll control at high angles of attack; therefore, most pilots tend to use rudders instinctively for roll control at high angles of attack. Consequently, when pilots are transitioned into airplanes using the CAS concepts, training is required to insure that the pilot fully understands the important differences in airplane response to controls for the CAS type airplane. With a properly designed roll CAS, the pilot can roll the airplane I I 1 1 1 1 1 1 I I Ill I I I over a large angle-of-attack range using conventional lateral-stick inputs. The pilot should not attempt to roll the airplane using only rudder pedals since the roll CAS opposes the roll rate and degraded roll performance results.

HIGH-ANGLE-OF-ATTACK CONTROL CONCEPTS Within the last decade, there has been increasing emphasis on including the fore- going high-angle-of-attack considerations in the control-system design process for fighter aircraft. As a result, a number of high-angle-of-attack control concepts have emerged, many of which are now being very effectively used in the latest fighters (refs. 3 to 6). These concepts include roll- and yaw-control interconnects, stability-axis yaw damping, static stability augmentation, angle-of -attack limiting, and automatic spin prevention.

Roll- and Yaw-Control Interconnects The basic rationale for roll/yaw interconnect systems, often called aileron- rudder interconnect (ARI), arise from the high-angle-of-attack kinematic coupling phe- nomenon discussed previously. Recalling that avoidance of large slideslip generation during high-angle-of-attack rolls requires body-axis.yawing as well as rolling, it is seen that the proper response to pilot roll inputs is an appropriate mixture of yaw- and roll- control deflections. Furthermore, because the required relationship between yaw and r o l l rates is r = p tan a, the needed ratio of yaw to roll control increases with increasing angle of attack. An additional factor that must be taken into account is the yawing moment produced b y the roll control. As discussed in the section on aerodynamic considerations, the general trend with increasing angle of attack is toward adverse yaw. For situations where the adverse yaw due to roll control is significant, the interconnect gain must be sufficiently high to provide enough rudder deflection to both counter the adverse yaw and to minimize kinematic coupling. This factor is best summarized by using the lateral control divergence parameter (LCDP), which, in the presence of an interconnect system, is given by where K is the gearing ratio between the yaw and roll controls.

1 7 - Returning to the example fighter configuration, figure 1 7 shows computed values of LCDP for the airplane with and without an optimized ARI system. A s shown pre- viously, the basic configuration exhibits negative values of LCDP, which indicates suscep- tibility to control-induced departures because of improper response to pilot roll com- mands. Addition of the ARI results in large positive values of LCDP throughout the angle-of-attack range, which indicates normal roll response and high resistance to control-induced departures.

ARI systems are generally implemented in one of two methods. The first method involves driving the rudders directly proportional to roll-control deflection, which results in a true aileron-rudder interconnect. In the second implementation, the rudders are driven by pilot lateral-stick deflection, so that it would be more accurate to refer to such a system as a lateral-stick-to-rudder interconnect (LSRI). The tru,e ARI has an advantage over the .LSRI in that with the former the rudder is driven by roll-control deflection so that it can more accurately compensate for the characteristics of these con- trols. Nevertheless, the LSRI implementation is often used in situations where, in addi- tion to bringing in the rudder, it is also necessary to fade out roll-control deflection with increasing angle of attack because of excessive adverse yaw o r insufficient rudder effectiveness.

In summary, ARI systems can provide several very significant improvements at high angles of attack. The first is the proper coordination of high-angle-of-attack rolls, which thereby inhibits departures due to improper roll response as discussed previously.

A second benefit which results directly from the improved coordination is the enhance- ment of roll performance as illustrated in figure 18. Peak roll rates are plotted v e r s u s angle of attack. These rates are obtained for full lateral-stick inputs applied during turns at M = 0.6 andh = 9144 m (30 000 ft). Theairplane was equippedwithan ARI system and data were obtained with and without the system active. As can be seen, the basic airplane exhibited a sharp drop in roll performance with increasing angleof attack primarily because of low static directional stability. Above CY = 2 1 ° , roll reversal w a s encountered. Activation of the ARI resulted in much improved roll performance through- out the angle-of-attack range and the roll-reversal characteristic was eliminated.

An additional important benefit of the A R I is that it greatly simplifies the pilot work load at high angles of attack. Without the ARI, the pilot must properly coordinate the rolls himself, which is a very difficult task, particularly in the air combat environment involving rapid, vigorous maneuvering with the pilot's attention out of the cockpit. As a result of the high work load, coordination is often poor and results in the degraded roll performance and departure susceptibility problems discussed previously. With an ARI system, the coordination task is automatically performed by the control system so that the pilot can use the same technique to roll the airplane at all angles of attack.

This significantly reduces the pilot work load and enhances the ability to maneuver the airplane effectively.

Because an ARI system automatically applies coordinating rudder in the direction of the stick input, it inhibits the pilot from obtaining cross controls (yaw and roll controls in opposite directions) which are departure and spin ‘inducing. Thus, an additional benefit provided by an ARI is that it prevents the departure caused by inadvertent pilot applica- tion of cross controls.

Despite the many enhancing features of ARI systems, there are also some potential problem areas that should be kept in mind in designing these systems. An important example is a case in which the airplane exhibits very poor Dutch roll damping character- istics, such as the wing-rock phenomenon discussed previously. In this example, the addition of ARI (particularly if the system gain is too high) can lead to aggravation of the oscillation in the form of divergent lateral pilot-induced oscillation (PIO). Some insight into this phenomenon can be gained by considering the pilot as a simple roll damper trying to damp the wing-rock motion. Thus, the roll inputs of the pilot are in a direction which opposes the roll rate, which, as w a s discussed previously, is the primary angular motion seen in this type of oscillation. The rolling moments resulting from the pilot inputs would thus augment CL , which is stabilizing; however, the yawing moments P obtainedthrough ARI produce a negativeincrement in Cwhich is destabilizing. If “p’ the ARI gain is highenough that the adverse effect of Cn is stronger than the stabi- P lizing effect ofCL , the net result is an aggravation of the low damping condition.

P An indication of the maximum ARI gain above which closed-loop instability can occur is obtained by reexamining the approximation for the damping of the Dutch roll modeshown earlier. The pilot is modeled as a simple roll damper with a gain of Kp degrees of control deflection per deg/sec of roll rate. Consider the combined yaw- and roll-control deflections obtained through the ARI as a single control deflection hARI

with the resulting derivatives CL and C . As discussed previously, the effect

% R I nhAFtI of thepilotclosingtheloop as a roll damper is equivalenttomodifying CL and Cn P P with the increments given by Adding these increments into the approximation of the Dutch roll damping parameter shown previously and taking the derivative of that expression with respect (1/t1/2)DR to Kp gives FortheconditionsCnp > 0 and C < 0, theeffect of thepilotclosingthe loop , Y IP is stabilizing if with an ARI system o r C- I1 cZ 6ARI cos CY > sin (Y IZ I X By convention, Cz and C a r e negative, so the above expression can be 6AFtI n%RI written as

I cn%RIl cos (Y < lc'bARIl sin CY

I Z IX

Multiplying both sides by $ 4 1 I 6mI I yields

The two t e r m s in parentheses in the preceding expression are simply theyaw and roll angular accelerations caused by 6 ~ ~ 1 , so that the requirement for enhancement of closed-loop stability is simply Recalling that proper roll coordination at high angles of attack requires roll about the velocity vector with r = p tan a, the preceding inequality states that in order that the ARI not degrade closed-loop stability, its gain must not be so high as to overcoordinate the roll. Thus, from the viewpoint of closed-loop stability, it is desirable to design an ARI such that some adverse sideslip is generated during stick rolls at high angles of attack. More specifically, it is important to avoid excessive ARI gains which result in overcoordination which generates proverse sideslip, especially if the airplane exhibits low inherent damping.

An example of the potential adverse effect of ARI systems on closed-loop stability is contained in figure 19. The time histories, which were obtained in a piloted simulation study, are indicative of piloted attempts to track a target airplane performing a steady windup turn. The study airplane exhibited wing rock at angles of attack greater than 1 7 ' .

The time histories on the left-hand side of the figure show the performance obtained with a properly configured ARI system designed with the aforementioned considerations in mind. It is seen that as the pilot pulled to about a = 2 0 ° , the airplane exhibited a rela- tively modest amplitude wing-rock motion. However, there was no tendency toward PIO, and the pilot stated that the oscillations did not significantly degrade tracking performance.

The maneuver was repeated with the A R I gain doubled such that stick rolls were over- coordinated with significant generation of proverse sideslip. The time histories on the right-hand side of figure 19 show that a rapidly divergent PI0 occurred as the pilot pulled into the wing-rock region. After three oscillation cycles, the pilot realized that he was driving the instability and centered the stick. The oscillation amplitude immediately decreased, which verified that the divergence was in fact a PI0 caused by the excessive ARI gain.

The preceding results illustrate the importance of proper tailoring of an ARI system to the particular characteristics of the airplane and identify another consideration that should be addressed in the design process. As discussed previously in the section on aerodynamic considerations, the high-angle-of-attack stability and control characteristics of modern fighter aircraft often vary significantly with Mach number. To properly account for these effects, it may be necessary to include Mach number as a scheduling parameter in the ARI design. If this is not done, the design is compromised, which not only results in degraded system effectiveness but may also lead to the closed-loop stability problem discussed previously.

Another potential disadvantage of ARI systems results from the fact that they do not allow full cross controls to be applied. As discussed previously, this is beneficial in preventing inadvertent application of such controls which can result in departures.

However, if a departure or spin entry should occur, the optimum spin recovery pro- cedures for modern fighter aircraft involves maximum cross controls (roll control with and yaw control against the rotation rate). For this reason, the design of an ARI system should include the capability t o deactivate the system or some other provision for allowing the application of full recovery controls in the event of a spin entry.

Two other potential problem areas arising from the use of ARI systems should be briefly mentioned. The first is associated with the nose-up inertia coupling resulting from stability-axis rolling. Because A R I systems promote this type of roll, use of such systems tends to accentuate this form of coupling. The second potential problem is that some pilot adaptation is usually required in flying an ARI-equipped airplane for the first time. The coning motion resulting from stability-axis rolling and the associated lateral acceleration at the cockpit may be disconcerting to the pilot at first. However, experi- ence has shown that after a brief familiarization period, most pilots adapt quickly to flying with an A R I system and are able to take maximum advantage of the benefits pro- vided by the concept.

Stability-Axis Yaw Damper A stability-axis yaw damper is often used in conjunction with an ARI system. The combination is a logical one. The ARI applies the appropriate mix of roll and yaw con- trols in response to pilot inputs to make the airplane roll approximately about the velocity vector, and the stability-axis yaw damper reinforces this motion by attempting to damp out any residual rs. As shown in figure 20, these devices are normally implemented by driving the yaw control. with an approximation for stability-axis yaw rate rs =: r - pa.

Many of the benefits of this concept are similar to those provided by ARI systems

discussed previously and involve the advantage of stability-axis rolling - minimization

of sideslip generation and the resulting improvement in roll performance. An additional benefit of the damper is that it can enhance Dutch roll damping up to high angles of attack,

particularly if static stability is maintained Cn . At low angles of attack,the

( P,dyn ' ")

body-axis yaw-rate component dominates and effectively augments damping. As discussed previously, body-axis dampers tend to lose effectiveness with increasing angle of attack; however,the p a cpmponentcanmaintain rs dampereffectivenessto muchhigher angles of attack.Figure 2 1 illustratesthischaracteristic. Computedvalues of a r e plotted for the example fighter configuration with and without a stability-

( DR

axis yaw-damper system. A s can be seen, the damper effectively augments Dutch roll damping throughout the angle-of-attack range.

Although the stability-axis yaw damper is more beneficial than a body-axis damper under most conditions, an exception is the case in which the airplane exhibits a rolling departure caused by loss of static directional stability. As discussed in a previous sec- tion, the conventional yaw damper can significantly reduce the severity of such a depar- ture, particularly at high-speed flight conditions. A stability-axis yaw damper would be less effective in this case because the roll-rate component p a would command rudder deflection in the direction of the departure and hence the net deflection to oppose the motion would be less than if only body-axis yaw rate was used.

Lateral-Directional Static Stability Augmentation As mentioned in the discussion of aerodynamic considerations, degradation of lateral-directional dynamic stability at high angles of attack is most often due to loss of rate feedback, static directional stability and dihedral effect. Damper systems based on such as the stability-axis yaw damper, are generally less effective in augmenting sta- bility in this situation. Obviously, the desirable solution is to directly augment Cn P and C1 b y drivingtheappropriatecontrolswith a sideslipsignal. Figure 22 shows P data for a configuration which typifies the loss of dynamic stability at high angles of

attack due to degraded characteristics of and Cn . The loss in static stability

P P

above a = 25' is reflectedin a sharpdrop in Cn , whichindicates a possible

P,dyn directional divergence at a = 3 0 ' . Alsoshown i n figure 22 are results obtained by aug- menting Cn and C 1 above a = 25'. The resulting values of Cn re main P P P,dyn large at high angles of attack, which indicates that there is no instability in this region.

In this case, the augmentation of C wasaccomplishedbyfeeding a p sig- nP

naltodrivethedifferentialtailsabove a = 25 Differentialtaildeflectionsonthe and "la

particular configuration produced large adverse yawing moments at high angles of attack, such that they could not be used for rolling in this regime. These large yawing moments, however, were used to augment stability as in the following augmentation law: + K C =C1 c1 P P b d Since CL is negative,thefeedback of p tothedifferentialtailaugmenteddihedral ‘ d effect. In addition, because C was positive (adverse), the feedback also augmented n6d directional stability by the increment K C , which was considerable since C was “6d n’d large. Note that the simultaneous augmentation of dihedral effect and directional stability usingthesinglecontrol would not bepossible if C werenotadverse.Thiscase n6d illustrates the important concept of using all available control moments for augmentation and control at high angles of attack, including those that are conventionally considered adverse, since these controls can often be quite powerful in the stall region.

This concept of augmenting static lateral-directional stability has not as yet been implemented and used on production aircraft primarily because of the difficulty in obtaining an accurate measurement of sideslip at high-angle-of-attack flight conditions.

However, the potential benefits of this concept warrant further development.

Angle-of-Attack Limiting The basic rationale for using an angle-of-attack limiter is illustrated in figure 23.

A conceptual plot is shown of the variation of stability and control characteristics for a fighter as angle of attack increases. The characteristics are shown to degrade from departure-resistant at low angles of attack to departure-prone at high angles of attack.

It should be noted that the divergence problem can be lateral-directional or longitudinal.

Whatever the case, the rationale for using an angle-of-attack limiting system is to pre- vent the airplane from entering the departure-prone region. These systems are usually implemented by using angle of attack and filtered pitch rate (for lead) to drive the pitch control to inhibit excursions beyond the set angle-of-attack limit. Mach number sched- uling is also sometimes necessary to account for Mach number effects on the aerodynamic characteristics.

The benefits provided by this concept are obvious. By preventing excursions into regions of poor departure resistance, the chances of encountering loss of control and possible spin entry are greatly reduced. Furthermore, because this is accomplished automatically by the system, the pilot’s task is simplified in that he can confidently apply up to maximum nose-up command without fear of overrotating into regions of poor sta- bility and control characteristics.

Unfortunately, the angle-of-attack limiting concept also entails some inherent draw- backs. An obvious one is that if it is necessary to set the limit significantly below the angle of attack for maximum lift, then a maneuverability penalty is incurred. Further- more, the limiter negates the ability to perform rapid, momentary rotations to extreme angles of attack, which can b e a useful tactical maneuver. Even'within the angle-of-attack envelope allowed by the limiter, nose-up pitch response, particularly near the boundary, can be sluggish becauseof the pitch-rate lead information used to minimize overshoot of the boundary.

A final limitation of angle-of-attack limiting systems is that they are not foolproof.

is of course limited by that of the pitch control through which they Their effectiveness operate. Thus, an effective technique for defeating angle-of-attack limiter systems is to decelerate to very low airspeeds in a nose-high climb; the ensuing lack of dynamic pres- sure negates the ability of the limiter to prevent angle-of-attack excursions beyond the limit value. Other maneuvers that can potentially defeat the limiter system involve sus- tained high roll rates during which sufficient nose-up inertia coupling moments are gen- erated to overpower the available nose-down control moment. This type of maneuver can be especially critical for fighters incorporating the CCV concept of relaxed static stability is discussed in more detail in a subsequent section of this paper.

in pitch. This problem Automatic Spin Prevention Experience has shown that spins can be avoided if the proper recovery controls are applied as quickly as possible following loss of control. Unfortunately, during air combat the pilot's attention is occupied with many other tasks besides control of the aircraft and as a result, there is often a significant delay between occurrence of a departure and pilot response to it. Furthermore, following recognition of the out-of-control condition, the inputs applied by the pilot may not be optimum for recovery and, in the extreme case, may even aggravate the departure and accelerate the spin entry. It would appear, there- fore, to be highly desirable to relegate the spin prevention task to an automatic system.

A system capable of this task would have several inherent advantages over the human pilot, including (1) quicker and surer recognition of an incipient spin, (2) faster reaction time for initiation of recovery, (3) application of correct spin recovery controls, and (4) elimination of tendencies toward spin reversal. Reference 3 documents some early investigations of automatic spin-prevention systems conducted at the Langley Research Center. Simulation results and flight tests of subscale dynamic models showed that such systems could be very effective in preventing spins, even for fighter configurations that are very spin prone.

The basic implementation concept for these systems is conceptually illustrated in 24. Regionsrepresentingthenormalmaneuveringenvelope(lower (Y and r) figure andthedevelopedspin(high (Y and r) are shown on a yaw-rate/angle-of-attack plot.

The area between these two regions indicates the departure and incipient spin phases, where the system activation threshold would be located. It is generally desirable to place the threshold as far from the developed spin region as practical to maximize system effectiveness and speed of recovery to controlled flight. On the other hand, placement of the threshold too close to the maneuver envelope could result in the system interfering with the pilot's normal control of the airplane. Each system must therefore be tailored for the specific stall/spin characteristics of the particular configuration. Once an impending spin has been identified, the system commands controls for spin recovery which for modern, fuselage-heavy fighter airplanes normally consist of roll controls with the spin, rudder against the spin, and nose-up pitch control.

Some recent results of application of automatic spin prevention to a current fighter configuration are shown in figure 25. These results were obtained during free-flight tests of an unpowered, subscale dynamic model (additional information on this test technique is given in ref. 7). The time histories shown on the left-hand side of figure 25 were obtained for the basic configuration not equipped with a spin prevention device. The spin- entry attempt was initiated from a split-S maneuver by applying and holding full cross controls starting at t =: 5 sec. The control system of this airplane incorporated a roll/yaw control interconnect which not only drove rudders with lateral-stick inputs but also reduced maximum roll-control deflections at high angles of attack. However, to make possible the application of full-recovery controls in the event of a spin, the inter- connect was deactivated when yaw rate exceeded a preset value. As a result of the inter- connect feature, much less than maximum prospin control deflections were obtained even though the pilot was holding full inputs. Nevertheless, yaw rate and angle of attack increased steadily. At t =: 12 sec, the yaw rate exceeded the value for deactivation of the interconnect system, and full cross controls were obtained which accelerated the spin entry. Shortly thereafter, the model stabilized in a fast, flat spin. The results obtained with an automatic spin-prevention feature incorporated in the control system are shown on the right-hand side of figure 25. Although essentially the same inputs were made by the pilot, no spin entry occurred. Each time yaw rate began to increase and exceeded the activation threshold of the system (t =: 8 and 12 sec), antispin controls were automatically applied which quickly reduced the rotation rate. The resulting motion can best be described as a steep spiral fromwhich recovery was immediately obtained when the pilot inputs were neutralized.

An inherent characteristic of automatic spin-prevention systems which is potentially a cause of concern is the high control authority that is generally required for maximum effectiveness. This means that upon activation they necessarily take some if not all con- trol away from the pilot. Under nominal conditions, this is a satisfactory situation; how- ever, the implications under a failure condition are more serious. For example, if a malfunction in the system causes it to activate when the airplane is in fact within its normal maneuver envelope, the control action caused by the system could drive the air- plane into a spin without the pilot being able to prevent it. Reliability is, therefore, a critical factor in the design of these systems.

In concluding this discussion of automatic spin prevention, a comment should be made regarding the effectiveness of this concept compared with departure prevention.

Experience has shown that although spin-prevention systems are desirable from a safety viewpoint, much greater improvements in tactical effectiveness would be expected with systems that prevent departures rather than allowing them to occur. However, departure-prevention control schemes are often incompatible with the requirements for spin prevention and recovery. An example mentioned previously is the inability to obtain optimum spin-recovery controls (maximum cross roll and yaw controls) with an active ARI system. This shortcoming can be effectively resolved by deactivating the departure- prevention devices at the appropriate time and bringing in the spin-prevention system.

Thus, ideally, use of both spin- and departure-prevention systems should be considered for fighter aircraft so that the desired combination of high maneuverability and departure/spin resistance is attained.

FUTURE CONCEPTS AND APPLICATIONS Operational experience with the latest series of fighters has verified that proper control-system design can significantly enhance the departure/spin resistance of fighter aircraft. It is expected, therefore, that this technique will be further developed for application to future configurations. Although it should be possible to use the current high-angle-of-attack control concepts on these advanced designs, it is likely that these advanced vehicles will exhibit aerodynamic characteristics that are significantly different from those of current fighters. Furthermore, future configurations will probably employ unusual types of controls, such as canards and thrust vectoring, and use them in novel ways. All these factors will introduce new high-angle-of-attack problems, the solutions for which w i l l require further development of the control-system design approach. It is beyond the scope of this paper to discuss in any detail all these potential future concepts and problems. Rather, the remainder of the paper is a discussion of one concept, the (RSS). This principle will very likely be incor- CCV principle of relaxed static stability porated in future fighters, but it has serious implications on the high-angle-of-attack control problem.

RSS involves designing the airframe so that it has low or negative inherent static longitudinal stability and using the control system to artificially provide the required level of stability; the F-16 airplane employs this design concept. The potential perform- ance benefits of this concept are well known and fighter designs which incorporate very high levels of inherent longitudinal instability a r e now being considered. The use of RSS, however, can also introduce potential stability and control problems at high angles of attack in addition to the more familiar- lateral-directional problems discussed previously.

The fundamental aerodynamic characteristics of RSS configurations which can result in these problems are illustrated in figure 26, which shows an idealized plot of aerodynamic pitching-moment coefficient Cm versus angle of attack which might be expected for a statically unstable configuration. The two main potential problem areas are indicated by the hatched regions as an indication of where they are likely to occur.

The lower angle-of-attack region below maximum lift represents an area of susceptibility to uncontrollable pitch departures due to lack of sufficient aerodynamic nose-down con- trol moment.Asindicated, if the angle of attack exceeds aCrit, additionalnose-down moment cannot be generated and the airplane pitches up into an out-of-control situation.

Thus, a critical requirement of the control system in this case is to limit the maximum angle of attack to values where control can be maintained. Note, however, that in the cycrit there is very little nose-down moment angle-of -attack region immediately below available to prevent angle-of-attack excursions above acrit' For angles of attack higher than acrit, it is seen that the Cm curves change slope because of increasing break- down of the flow over the entire configuration. This causes the aerodynamic center to move aft, which results in stable deep-stall trim points at very high angles of attack.

If the airplane enters these deep-stall trim points, recovery may be very difficult since aerodynamic controls are generally not very effective at these extreme angles of attack.

In summary, RSS configurations can be susceptible to pitch departures at high angles of attack when there is insufficient nose-down aerodynamic moment to prevent angle of attack from increasing above some critical limit beyond which the airplane can- not be controlled in pitch. Two dynamic phenomena which can generate large-angle-of- are kinematic coupling and attack excursions during high-angle-of-attack maneuvering inertia coupling. As discussed in a previous section, substantial increases in angle of be generated kinematically by rolling with proverse sideslip; in addition, large attack can angle-of-attack excursions can occur during recoveries from steep attitude climbs to very low airspeed. Inertially, significant nose-up moments are generated during stability-axis rolling at high angles of attack. If these moments are greater than the available nose-down aerodynamic moment, a pitch departure is likely to occur. This problem is further illustrated in figure 27.

The variation with roll rate of the nose-up inertia coupling moment caused by stability-axis rolling is shown in figure 27. As noted previously, the moment varies with ps2 so that very substantial moments can be produced at high roll rates. Also shown are representations of the available nose-down control moment for a specified angle of attackattwovalues of dynamic pressure, and i2

(91 < q2 - ) . The points

of intersection with the coupling-moment curve indicate the highest roll rates p*

( s,l

and p* ) at which sufficient control moment exists to counter the nose-up coupling

s, 2 moment. If the roll rate should increase and be sustained above these values, then it is very likely that a pitch departure will occur. Note that p : , < pi,2, which indicates that the susceptibility to this type of departure becomes more acute as airspeed decreases.

Once a departure beyond the critical angle-of-attack limit occurs, the airplane is likely to continue to pitch up to very high angles of attack and potentially restabilize in a deep- stall t r i m point as indicated in figure 26. Furthermore, the results of figure 27 point

out a rather unique characteristic of RSS configurations - that the maximum sustainable

r o l l rates that can be controlled at high angles of attack may be limited by the effective- ness of the nose-down pitch control.

A fighter configuration which exhibits the characteristics discussed in this section was recently studied in wind-tunnel and piloted simulation investigations a t Langley Research Center. (See ref. 8.) The airplane was designed to operate at very moderate inherent levels of pitch instability (Static margin = -0.04) and incorporated a conventional aft-mounted all-moveable horizontal stabilator for pitch control. The configuration had a moderately swept wing and a highly swept wing-body strake to enhance lift and maneu- verability at high angles of attack.

Variations with angle of attack of the pitching-moment coefficient for neutral, full nose-up, and full nose-down stabilator deflections are shown in figure 28. The data show thattheairplane would t r i m at a = 66' withfullnose-upstabilatordeflection. To inhibit inadvertent excursions to these extreme angles of attack, the pitch-control system incorporated an angle-of-attack limiter which drove the stabilator in an attempt to keep angle of attack from exceeding 25'. Figure 28 shows, however, a sharp loss in nose-down stabilatoreffectivenessfor a 25'. Itwasnotsurprising,therefore,thattheairplane exhibited susceptibility to pitch departure caused by the inertia coupling phenomenon dis- cussed previously. Figure 29 shows an example of such a loss-of-control situation encountered during the simulation study of this configuration. An attempted 360' roll is shown with full lateral-stick input applied at a = 25' in an accelerated turn. In addition to maximum roll-control deflections, maximum coordinating rudder was also obtained to make the airplane roll about the velocity vector. As a result, the body-axis roll and yaw rates began to build up rapidly in the direction of the stick input.Initially, a dropped slightlybecause of kinematiccoupling;however, as pand r increased,the inertia coupling moment caused a significant nose-up pitch rate to build up, and a began to increase. At this point, q coupledwithptocreate a yawcouplingmomentwhich opposed the yaw rate and halted its growth (t = 14 sec); on the other hand,p was still in the kinematic generation of a large amount of adverse increasing and thus resulted sideslip (t = 15 sec). By this time, a had increased to above 3 0 ' despite the angle-of- The nose-up attack limiter system applying full nose-down stabilator deflection (25').

inertia coupling moment was much greater than the nose-down aerodynamic moment pro- duced by 6h = 25'; as a result, a pitch departure occurred as the airplane completed about 2'70' of the roll, just after t = 15 sec. During the ensuing loss-of-control period, a reached a maximum of 68' while p oscillatedbetween 26' and -26'.

, Theforegoingresults showedthattheairplaneroll-ratecapabilityat high angles of attack was too high to prevent pitch departures with the available pitch control, and that the only means of alleviating the pitch departure problem (other than resizing the control surfaces or further constraining maximum angle of attack) was to properly design the control system to reflect this fact. The control-system features developed to achieve this goal are summarized conceptually in figure 30. Obviously, an essential element is a roll-rate limiter system which inhibits the roll rate from exceeding the critical values which result in inertia coupling departures. Four parameters were used to evaluate what the roll-rate limit should be at any given time - dynamic pressure, angle of attack, symmetric stabilator deflection, and roll-rate magnitude. Angle of attack was chosen (1) the nose-up inertia-coupling moment varies with sin 2a; and f o r two reasons: (2) the amount of nose-down control moment available to counter the nose-up coupling moment decreases as angle of attack increases. Similar reasoning was used in choosing 4. Symmetric stabilator deflection w a s chosen because it directly indicates the pitch control remaining to oppose the inertia-coupling moment. Finally, roll-rate magnitude was used to schedule the total level of limiting imposed. For low roll rates, where coupling is not a factor, no limiting was imposed regardless of the values of the was used only when the roll rate other scheduling parameters. The limiting schedule approached significant magnitudes. With this scheme, the initial roll-response degrada- tion was minimized and roll performance was compromised only where it was essential to prevent occurrence of the pitch departure.

A second feature was incorporated to minimize the generation of proverse sideslip during rolls at high angles of attack. As discussed previously, rolling with proverse sideslip increases angle of attack through kinematic coupling and is therefore not desirable for RSS configurations for which there is a requirement to maintain angle of attack below some critical value. Proverse sideslip minimization was accomplished by scheduling the maximum rudder deflection that the pilot could command through his pedals as a function of angle-of-attack and roll-rate magnitude such that at significant values of a andlpl, no deflection could b e commanded. In this situation, the only direct rudder response to pilot inputs came from coordinating rudder deflections commanded by the roll-control input. Hence, no overcoordination of high-angle-of-attack rolls was possible.

The final feature incorporated in the control system to inhibit the pitch departure was an inertia-coupling compensator for the pitch axis to assure proper stabilator response during high-angle-of-attack rolling maneuvers. The system used angle-of- attack and roll-rate magnitude to drive the pitch control in the nose-down direction to oppose the nose-up coupling moment.

The effectiveness of the preceding control-system design in preventing inertia- coupling departures is illustrated in figure 31, which shows a 3 6 0 ' roll using full lateral stick applied at CY = 25' in an accelerated turn. As discussed previously, this maneuver performed with the basic airplane resulted in loss of control. (See fig. 29.)However, be completed with the modified control system. Although the pilot the maneuver could applied and held full roll command, the system began to decrease the roll-control deflec- tion as the roll rate approached critical values above which there would not be sufficient nose-down control to oppose the coupling moment. Note that near the completion of the maneuver, only about 25 percent of maximum roll-control surface deflection (authority of *20°) was used. Because roll rate was properly limited, the stabilator never reached its maximum deflection, and consequently no departure occurred. Angle of attack did not exceed the 25' limit and the maximum sideslip generated was only 3 ' .

As discussed previously, the second major high-angle-of-attack control problem introduced by use of the RSS concept is the potential for deep-stall trim. As shown i n figure 28, the subject configuration exhibited a weak but stable deep-stall trim at CY = 60' even with full nose-down stabilator deflection. Two techniques were found for generating a pitch departure which would result in angle-of-attack excursions into the deep-stall region. The first technique was the coupling departure discussed in the pre- ceding section of this paper. This departure, however, could essentially be eliminated by the control-system design discussed previously. However, a second deep-stall entry technique was found which the control system was unable to prevent. The maneuver involved putting the airplane into a steep, nose-up-attitude, decelerating climb with 0 reaching a maximum of about 7 0 ' and allowing airspeed to bleed off to about 35 KIAS a t the top of the climb. The airplane was then allowed to fall through at essentially zero g.

a large angle-of-attack excursion could not be The resulting kinematic generation of effectively opposed by the pitch control because of the very low level of dynamic pressure involved. Once the airplane was stabilized in a deep stall, no recovery w a s possible because the angle-of-attack limiter system commanded full nose-down control deflection independent of pilot input. To CoTreCt this deficiency, the control system was modified such that it was possible to reconfigure the pitch-control law in the event of a deep-stall entry. The reconfiguration involved deactivating all feedback, including the angle-of- attack limiter system so that the 0 d . y signal that remained was the pilot stick command.

With this system, the pitch control deflected directly proportional to pilot inputs. The reason for doing this can be seen by reviewing the pitching-moment data for the maximum stabilator deflections shown i n figure 28. The data show that at the deep-stall trim point (CY = 6 0 ° ) , a comparatively large pitching-moment increment results in going from full nose-down to full nose-up control deflection (ACm = 0.1). Thus, a possibility exists to use this available control moment to initiate and build up a pitch oscillation by moving the stick in phase with the airplane motions. This would be done with the hope that sufficient angular momentum would be created during a downswing cycle to drive the airplane over the positive Cm"hump"and back down to within the normal angle-of- attack envelope of the airplane.

A recovery attempt using the pitch oscillation technique is shown in figure 32. The nose-high decelerating-climb technique was used to cause the airplane to enter the deep- stall t r i m point. Starting from this stabilized trim at (Y = 62O, the pilot reconfigured the control system and rapidly applied full aft stick at t = 71.3 sec. The resulting nose- up moment caused (Y to increase to 75O, a t which point the pilot reversed his controls and applied full forward stick. About 14 deg/sec nose-down pitch rate resulted and the associated angular momentum was sufficient to cause the airplane to continue to pitch downward until a recovery was obtained at t = 78 sec. Although the rapidity of the recoveries obtained was dependent on proper phasing of pilot inputs, this technique was found to be consistently effective in providing recoveries from the deep-stall trim exhibited by this configuration.

To summarize this discussion of the impact of RSS, the foregoing results have shown how proper control-system design can significantly alleviate the high-angle-of- attack problems of pitch departure and deep stall that can result from applying the RSS concept to fighter aircraft. Control-system concepts that enhance resistance to pitch departure are angle-of-attack limiting and roll-rate limiting. Because of the ability of RSS configurations to trim to extreme angles of attack, some form of angle-of-attack limiting system is normally required and selecting the limit angle of attack to be as low as possible maximizes the resistance to pitch departure. Naturally, the effectiveness of this approach is constrained by performance considerations which are a critical factor in determining the angle-of-attack limit value to be used. Once an angle-of-attack limit is selected, further enhancement of resistance to pitch departure can be attained by designing the control system such that a proper balance is achieved between roll per- formance capability and available nose-down aerodynamic control moment. This may involve some form of roll-rate limiting and other features to inhibit the generation of proverse sideslip during rolls at high angles of attack. Gith regard to the deep-stall problem, recovery potential can be enhanced by incorporating the capability to recon- figure the pitch-control system such that the nominal feedback paths, and in particular the angle-of-attack limiter, are deactivated s o that the pilot has direct control of the pitch control with which to attempt to oscillate the airplane out of the trim point. For new airplanes, of course, it would be more desirable to design the airframe so as to pre- clude the existence of a stable deep-stall trim condition. (See ref. 9.)

CONCLUDING REMARKS The present paper is a summary of the experience gained at the NASA Langley Research Center in the area of control-system techniques for enhancing the departure/spin resistance of fighter aircraft. The reader should be aware that this paper does not con- stitute a complete and exhaustive study, since high-angle-of-attack aerodynamics are known to be quite configuration-dependent. Rather, this study was focused on key aero- dynamic stability and control characteristics that are common to most current fighter configurations and are important to control-system design. Experience at Langley to date has shown that proper control-system design can be an effective approach for attaining the goal of high maneuverability combined with good resistance to departures and spins. Operational experience with the latest fighter designs has confirmed these results. In view of these successes, further development of this approach is required to maximize the benefits that can be realized in applications to future fighter aircraft.

From the analysis presented in this paper, several important conclusions are sum- marized as follows: 1. The strong dependence of key aerodynamic stability and control derivatives, including both static and damping derivatives, upon angle of attack and Mach number should be carefully considered in control-system design for high-angle-of-attack flight conditions.

2. Control systems designed.without proper consideration for the high-angle-of- attack aerodynamics and flight mechanics phenomena noted in this study can often have detrimental effects on airplane departure/spin resistance.

3. Several of the pilot control techniques that have been applied effectively to air- planes having conventional stability augmentation systems can cause unexpected stall entries and degraded high-angle-of-attack handling qualities if applied to airplanes using high-authoritycommandaugmentationsystems (CAS). Therefore, considerable care should be taken in training pilots for transition from old to new airplanes to insure that the unique control characteristics of CAS airplanes are well understood.

4. Properly designed aileron-rudder interconnect systems and stability-axis yaw dampers can provide improved departure/spin resistance, improved Dutch roll damping, and improved roll performance at high angles of attack.

5. In the design of roll dampers used athigh angles of attack, careful attention must b e paid to the airplane lateral-directional static stability and control characteristics to insure that favorable rather than adverse effects are obtained; proper system limiting must be observed to avoid degrading departure/spin resistance at high angles of attack.

6. Proper high-angle-of-attack control-system design should consider the use of all available control moments for stability augmentation and control, including those that are conventionally considered adverse since these control moments can often be quite powerful in the stall region.

7. Use of both departure- and spin-prevention systems should be considered for fighter aircraft to obtain the desired combination of high maneuverability and high departure/spin resistance.

8. For fighter configurations incorporating relaxed static stability (RSS), the high- angle-of-attack control-system design must account for both the classic lateral- directional stability and control problems as well as the additional potential problems of pitch departures due to high roll-rate maneuvers and low airspeed maneuvers.

Langley Research Center National Aeronautics and Space Administration Hampton, VA 23665 June 5, 1980 REFERENCES 1. Chambers, Joseph R.; andGrafton,Sue B.: Aerodynamic Characteristics of Airplanes at High Angles of Attack. NASA TM-74097, 1977.

2. Ross, A. Jean: Lateral Stability at High Angles of Attack, Particularly Wing Rock.

- 10-19.

Stability and Control, AGARD-CP-260, May 1979, pp. 10-1 3. Gilbert, William P.; Nguyen, Luat T.; and Van Gunst, Roger W.: Simulator Study of Applications of Automatic Departire- and Spin-Prevention Concepts to a Variable- Sweep FighterAirplane. NASA TM X-2928, 1973.

4. Nguyen, Luat T.; Gilbert, William P.; and Van Gunst, Roger W.: Simulator Study o f the Departure Resistance of a Lightweight Fighter Airplane With Twin Vertical Tails. NASA TMX-3510, 1977.

5 . Gilbert, William P.; Nguyen, Luat T.; and Van Gunst, Roger W.: Simulator Study of the Effectiveness of an Automatic Control System Designed To Improve the High- Angle-of-Attack Characteristics of a Fighter Airplane. NASA T N D-8176,1976.

6. Gilbert, William P.; and Libbey, Charles E.: Investigation of an Automatic Spin- Prevention System for Fighter Airplanes. NASA TN D-6670, 1972.

7. Chambers, Joseph R.; Bowman, James S., Jr.; andMalcolm,Gerald N.: Stall/Spin Test Techniques Used by NASA. Stall/Spin Problems of Military Aircraft, AGARD-CP-199, June 1976, pp. 13-1 - 13-12.

8. Nguyen, Luat T.; Ogburn, Marilyn E.; Gilbert, William P.; Kibler, Kemper S.; Brown, Philip W.; andDeal, P e r r y L.: Simulator Study of the Stall/Post-Stall Character- istics of a Fighter Airplane With Relaxed Longitudinal Static Stability. NASA TP-1538,1979.

9. Nguyen, Luat T.; Gilbert, William P.; andGrafton, Sue B.: Control Considerations for CCV Fighters at High Angles of Attack. Aerodynamic Characteristics of Controls, AGARD-CP-262, Sept.1979, pp. 11-1 - 11-10.

a Figure 1. - Body system of axes.

2 . 0

. a

. 4 C

a, DEG

Figure 2.- Variation of low-speed lift and pitching-moment characteristics with angle of attack for a typical modern fighter configuration.

.003

- 0 " = . 1 5 - U M = . 9

. 001 STA B LE

deg -1 UNSTABLE

-. o o l t

Y I S

-. 004L \o

I I I I I I I I I DEPARTURE RES JSTANT PARTURE PRONE a, DEG

Figure 3. - Variation of static lateral-directional stability characteristics

with angle of attack.

3%

-. 002-

%; -.-001-

deg -1

\\

M = . 9 5 \

a M=.15

I I I I I I I

0 5 10 15 20 25 3 0 35 40

a, deg

Figure 4.- Variation of rudder yawing-moment derivative with angle of attack.

-. 001

O M = .15

-. 003 €I" = .9

,001

deg-l

-. 0 0 1

-. 002

-. 0 0 3 L

I 1 1 1 1 1 1

0 5 10 15 20 25 3 0 35 40

a, deg

Figure 5.- Variation of differential tail characteristics of attack.

with angle

+M = .15

U M = . 9

LCDP, deg -1

"Oo40 5. 1 0 15 20 25 30 35 40

a. deg Figure 6.- Variation of lateral control divergence parameter with angle of attack.

4 1 C + C . sin a -.

b @

-. 4 L

-. 05 'F

-. 10-

Cn, - Cn' COS a

P

I I I I

-. 20

10 20 30 40

a, deg

Figure 7. - Variation of ,low-speed roll and yaw damping derivatives

with angle of attack.

@ = O

p = 0 a # O X

+

v

Figure 8.- Illustration of kinematic coupling of angle of attack and sideslip.

X

(a) Pitching moment created by roll and yaw rates.

P

(b) Yawing moment created by roll and pitch rates.

Figure 9. - Illustration of inertia-coupling phenomena.

l g LEVEL FLIGHT ( M < . 5 1

( k ) D R S 0

sec -1

0 BAS IC

-1 I). YAW DAMPER STAB LE sec -1 UNSTABLE -1 I I I I 1 I I

0 5 10 15 25 30

a. deg Figure 10.- Example of effects of conventional yaw damper on Dutch roll damping.

4 5 YAW DAMPER OFF YAW DAMPER ON 40 r deglsec 0 -20

P"

-80

-10

t "

t l l l l l l l I l l J 1 1 1 1 1 1 1 1 1 1 1 0 5 10 1 5 20 25

0 5 10 15 20 25

TIME, sec TIME, sec Figure 11.- Time histories of split-S entries into transonic instability region with and without a conventional yaw damper active.

. 4

l g LEVEL FLIGHT (M<. 5 1

"BASIC AIRPLANE

. 3

U ROLL DAMPER, N O M I N A L G A I N . 2 4L ROLL DAMPER, H I G HG A I N .1 sec -1

-. 1

I I I 1 I I I 10 15 20 25

0 5 30

a, deg

-

sec -1

- 2 . 4 -lw6I

-

0 5 10 1 5 20

a , deg

Figure 12.- Example of effect of conventional roll damper on Dutch roll damping.

lllll11l111lll11lll I 1 I I l l I I1 I ll11ll111l1l111l1111l1l111

-40 I

P, degl sec

-100 I

15r

1 1 1 1 1 1 -

0 5 10 15 20 25

0 5 10 15 .20 25

t, sec t, sec Figure 13.- Time histories o f split-S entries into transonic instability region with and without conventional roll damper active.

I NTEGRATO COMMAND RESPONSE P I LOT

f

SHAPING COMMAND STICK

-

PITCH AND INPUT CONTROL FILTERING

I

r .

NORMAL ACCELERATION

+

1.0

M .5

- -

I I I I I I I ! I 1

0-

an 9

g u n i t s

-

"-

30 -

20 -

10 -

0 - 1 I I I 1

" "" 8, I I I 1 deg -100 I Pitch aft - I f l I I I I I I 1 L 0 ,

" -

-

-

-

bhp -10 -

I L

deg -20 - I I I I I I I I

0 4 8 12 16 20 24 28 32 40

Time, sec

Figure 15. - Example of inadvertent stall entry caused by pitch command

augmentation system.

20 r

-40 I 1

0 -

left r

30 r

2 o L 4 ! L L 10 0

40 r

-10 I I I I I 1 1 I L

2 : k - *

-20

- 100 1

4 s 7-

right 0 2 4 6 8 10 12 14 16

-50 -

2 4 6 8 10 12 14 16 Time, sec Time, sec Figure 16.- High-angle-of-attack stick-centered rudder roll on airplane with roll-rate command augmentation system.

0 " / .005- / - NORMAL ROLL RESPONSE - NO INDUCED DEPARTURE - LCDP 0

deg -l' -

- REVERSED ROLL RESPONSE

- POSSIBLE INDUCED DEPARTURE

-

-.,005- I I I 1 I I I I

0 5 10 15 20 25 30 35

a. deg

Figure 17.- Example of effect of an ARI system on roll-control effectiveness at high angles of attack.

---BAS IC

120 r

-AR I PEAK ROLL RATE, degl sec

0 10 20 30

ANGLE OF ATTACK, deg

Figure 18. - Example of effect of an ARI system on roll performance.

N O M I N A L G A I N A R I

* 10- /?

-

-10- w

I

- 1oL

1 1 1 1 1 1 1 1 1 ~ ~ ~ ~ 1 IIIIIIIIIIIIIII

0 5 10 15 20 25 30 35

0 5 10 15 20 25 30 35

TIME, sec

T IME. sec

Figure 19.- Tracking time histories illustrating effect of excessive ARI gain on closed-loop stability.

YAW CONTROL Figure 20.- Typical implementation of stability-axis yaw damper.

0 B A S I C O STAB I LITY-AX I S YAW DAMPER

s8 r

sec- - 2 " ! I 1 I I 1

0 5 10 15 20 25 30

a, deg

Figure 21.- Example of effect of stability-axis yaw damper on Dutch roll damping.

. 0 0 5 r ,005-

.004 -

.003 -

.002- -BAS IC OR Cn -BAS IC ---AUGMENTED b d y n ’ 0 ---AUGMENTED

deg -1 deg -1

-. 001 -

-. 002 -

-. 003 -

-. 0 0 4 L -. 004 -

I I I I I 1 I I I I I 1 I I

0 5 10 15 20 25 30 35 0 5 10 15 20 25 30 35

a, deg a, deg

Figure 22. - Example of effectiveness of high-angle-of-attack Figure 22.- Example of effectiveness of high-angle-of-attack static stability augmentation. static stability augmentation.

r

DEPARTURE RES I STANT I a L I M I T

S T A B I L I T Y AND CONTROL I ST I CS CHARACTER -, I \

\

DEPARTURE PRONE I: I ANGLE OF ATTACK

ANGLE OF ATTACK 4 I

P I T C H RATE LE4TG P I T C H

CONTROL M A C H Figure 23.- Conceptual rationale for use of angle-of-attack P limiting systems.

YAW RATE \ , , T ~ ~ ~ ~ ~ ~ ~ THRESHOLD ANGLE OF ATTACK +

- ROLL CONTROL

THRESHOLD YAW RATE

SENSOR AND - t YAW CONTROL

ANGLE OF ATTACK *

CONTROL LOG1 C

- P I T C H CONTROL

Figure 24. - Conceptual illustration of implementation of automatic

spin-prevention systems.

BAS IC AUTOMATIC SPIN PREVENTION

0' 0

2 5 r

d i g -3c; 0 1 ' " ' .

FULL R I GkfL

NEUTRAL[^-

Ped

FULL LEFT

FULL R CHT

r

NEUtd A L L

\

FULL LEFT'

l 1 1 l 1 1 l 1 1 0 L l l l l l l l i l

6 1 2 18

0 24

0 6 1 2 18

TIME, sec

TIME, sec

Figure 25.- Time histories of spin-entry attempts made on a free-flying dynamic model with and without an automatic spin-prevention system.

M A X I M U M LIFT NOSE U P I bn LUIY I KUL -NtUTRAL ---FULL NOSE UP /// ,I --"FULL NOSE DOWN

cm

NOSE DOWN 0 Ucrit a, deg 90

Figure 26. - Pitching-momentvariationwith (Y for

idealized RSS configuration.

1 1 1 1111111111 I Ill 1111 I l l 1 I11111111 I I NOSE-U P COU PLI NG

MOMENT

P I T C H I N G -

MOMENT MAGNITUDE

0 p ; , 1 p ; , 2

STABILITY-AXIS ROLL RATE Figure 27.- Comparison of inertia-coupling moment for increasing roll rate with available aerodynamic pitch-control moment at two values of dynamic pressure.

6 1 .4 .2 I C

rn

-.2 I

- .4 0 NEUTRAL CONTROL

0 FULL NOSE-DOWN CONTROL

0 FULL NOSE-UP CONTROL

~~ J 20 40 60 80 100 0, deg Figure 28.- Variation of pitching-moment coefficient with angle of attack for a fighter configuration incorporating RSS.

6 2

P, deg 0

-10

- 20

- 30

-

-

- deg 0

P, 40

deglsec o - 20

-40 - 0 2 4 6 8 10121416 18 202224262830

TIME, sec

-

-80 deg/sec 0 " 0 2 4 6 8 1012141618202224262830 TI ME, sec Figure 29.- Example of inertia-coupling pitch departure encountered during 3 6 0 ' roll attempt on fighter incorporating RSS.

I

ROLL -

M A X I M U M C O M M A N D E D R O L L RATE

-

ROLL RATE CONTROL L I M I T E R M A X I M U M C O M M A N D E D RUDDERDEFLECTION

I P I M I N I M I Z E R

l p q ~~~~~~$~ 1 ~ P I T C H

COMPENSATOR CONTROL

a

Figure 30. - Conceptual schematic of control-system features designed

to inhibit high-angle-of-attack pitch departures.

I U N U l l l I I I 40 r

- 20 -1

right -

40 1 - 6lS neutral I \

r, deg/sec 0

- 40 I 1 1 1 1 1 1 1 1 ~

I eft

-

0, 'deg 0 aft

Pitch -

- 100

~ .. .

"fkL!LL

- 200 stick neutral

0 2 4 6 8 10121416182022 0; d L 0 1*1416 li 2b /*

TIME, sec

T I ME, sec

Figure 31. - Roll of 3 6 0 ' using full lateral-stick input on airplane equipped with pitch departure inhibitor system.

a, deg 40 -

-

O K 1 I .9 r

qy deg 0

- 20

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

TIME, sec Figure 32.- Time histories of deep-stall recovery using pitch oscillation technique.

I I l l I 1 I1 I I I I Ill I 1 1 1 1 1 l 1 1 l l 1 l 1 l l l 1 l 1 1 1 1 l l 1. Report No. 2. Government Accession No. 3. Recipient's Catalog No.

L A ; TP-1689

I

4. Title and Subtitle 5. Report Date August 1980 CONTROL-SYSTEM TECHNIQUES FOR IMPROVED 6. Performing Organization Code DEPARTURE/SPIN RESISTANCE FOR FIGHTER AIRCRAFT 7. Author(s) 8. Performing Organization Report No.

L-13453 Luat T. Nguyen, William P. Gilbert, and Marilyn E. Ogburn 10. Work Unit No.

9. Performing Organization NameandAddress 505-43-13-01 NASA Langley Research Center 11. Contract or Grant No.

Hampton, VA 23665 13. Type of Report and Period Covered 12. Sponsoring AgencyNameandAddress Technical Paper National Aeronautics and Space Administration 14. Sponsoring AgencyCode Washington, DC 20546 I 15. Supplementary Notes This paper summarizes some fundamental information on control-system effects on control- lability of highly maneuverable aircraft at high angles of attack and techniques for enhancing 1 fighter aircraft departure/spin resistance using control-system design. The discussion

i includes (1) a briefreview of pertinenthigh-angle-of-attackphenomenaincludingaerody-

namics, inertia coupling, and kinematic coupling, (2) effects of conventional stability aug- mentation systems at high angles of attack, (3) high-angle-of-attack control-system concepts designed to enhance departure/spin resistance, and (4) the outlook for applications of these concepts to future fighters, particularly those designs which incorporate relaxed static stability.

17. Key Words(Suggested by Author(s)) 18. Distribution Statement High angle of attack Unclassified - Unlimited

I

I Control-system design I

Departure prevention Aircraft stability and control Subject Category 01 I -I 19. Security Classif. (of this report) 22.Price'21. No. of Pages 20. Security Classif. (of this page) Unclassified A05 66 Unclassified F o r sale by theNationalTechnicalInformationService,Springfield,Virginia 22161 NASA-Lang1 ey , 1980 - - m t:j Postage and Fees Paid THIRD-CLASSBULK-RATE National Aeronautics and National Aeronautics and Space Administration Space Administration NASA451 Washington, D.C.

1 .

20546' Official Business ' Penalty for Private Use, $300 i 3 1 .1U,A, 072580 S00903DS DEPT OF THE B I B FOBCE BP IEAPOHS LAEOBATORI ATTN: TECBIICAL LIEBABY ( S U L ) KIfTLAlyD APB Ill!! 07117 - . , If Undeliverable (Section 1 5 8 -.> .Postal Manual) Do Not Return I -~ \ ' ,