8/8/2019 19710008947_1971008947

1/30

X-15 ADAPTIVE

Staff o f theFlightResearchCenter

93523

E R O N A U T I C SN DP A C ED M I N I S T R A T I O N W A S H I N G T O N , D . C. M A R C H 1971

8/8/2019 19710008947_1971008947

2/30

TECH LIBRARY KAFB, NM

0333325. -1. Rew rt No.NASA TN D-6208"4. Title and Subtitle I 2. Government Accession No.. .EXPERIENCE WITH THE X - 1 5 ADAPTIVE FLIGHT CONTROL SYSTEM

. . .7. Author(s)By Staff of the Flight Research Center9. Performing Organization Name and Address

. - . . "NASA Flig ht Resea rch CenterP.0 . Box 2 7 3Edwards,California 9 3 5 2 3-. . ." ~ ..12 . SponsoringAgencyName and Address

National Aeronautics and Space AdministrationWashington, D. C. 2 0 5 4 615 . SupplementaryNotes. _ . ~_ .

-~" .. .. -16 . Abstract

3. Recipient'sCatalog No.

5. ReportDateMarch 19716. Performing Organization Code

8. Performing Organization Report No.~ ~~

H - 6 1 810. Work Unit No.

7 2 2 - 5 1 4 0 - 0 6 - 2 4.~ ~.11 . Contract or Grant No.. " . -13. Type of Repo rt and PeriodCovered

Technical Note.. ___ . - -14. SponsoringAgencyCode

The X-15 adaptive flight control system is brieflydescribed, and system development and flight-testexperiences in the X - 1 5 research a i rplane are discussed.Airplane handling qualities with the system and systemreliabi l i ty are also discussed.

"." ~ .- - - " . . .. ""-17. Key Words Suggested by Aut hor ($) ) .18. Distribution Statement . .Unclassified - UnlimitedAdaptive flight control system - X-15 airplane -Stability and control

". ~-" _I_ ~19. Security Classif. (of this report ) 20. Security Classif. (of this page)

Unclassified 2 73 . 0 0"- . ". . ..

For Saleby the Nat ional Technical Informat ionervice, Springfield, Virginia 22151

8/8/2019 19710008947_1971008947

3/30

EXPERIENCE WITH THE X-15 ADAPTIVE FLIGHT CONTROL SYSTEMBy Staff of the Flig ht Res ear ch Cen ter

Flight Research Center

SUMMARY

An adaptive flight control system (AFCS) was evaluated during th e X-15 researchairplane program throughout the flight envelope to an altitude of 354,200 feet (108 kil-ometers) and a veloci ty of 5660 feet per se cond (6209 kilome ters p er hou r) and duringatmos pheri c entrie s in which angles of atta ck up to 25" were maintained . The dynamicpressure var ied f rom essent ia l ly 0 to approximately 1889 pounds per square foot(90,446 newtons per square centimeter).During the preflight system development cycle, problems with hysteresis, actuatorrate l imit ing, and structural resonance were experienced.The AFCS improved the handling qualities of the X-15 airpl ane in s ome res pectsThe most significant improvement occurred during atmospheric entry in which thegreatest variation in flight conditions existed.The system exhibited high reliability during the 65 flights of the test program. In-flight electronic component failures were generally undetected by the pilot because ofthe system's redundancy, fail-safe monitors, and adaptive gain compensation,The flight-test program confirmed the following advantages of the X-15 AFCS:

(1)Nearly invariant response was provided at essentially all aerodynamic flight con -ditions; (2) accurate a priori knowledge of aircraft aerodynamic characteris t ics was notrequired to design a satisf actory system ; (3) aircraft configuration changes wereadequately compensated for; and (4) the dual redundant concept provided a reliable andfail-safe system.The flight-test program also disclosed several disadvantages assoc iated with thesystem, including the following: (1 )Commands by the pilot and other spurious inputscaused gain reduction and degraded performance at undesirable times; and (2) s u p e r -critical gain operation existed in flight, which, because of mechan ical nonli nearit iesand electrical saturation, resulted in divergent airplane motions.

INTRODUCTION

Each innovation by c ontrol system designers to advance the technology is made inpur sui t of the elusive "optimum" in handling qualities. An invariant airpla ne respo nseto pilot control commands has been a design goal that has seemed to become more

8/8/2019 19710008947_1971008947

4/30

difficult to achieve with the advent of each new manned aerospace vehicle. Some ofthe problems are increasing range of center of gravity, widely varying control effec-t iveness , and var iable aircraft geometry and configuration. The "self-adaptive" flightcontrol concept was conceived as one solution to the growing problem. This conceptwas def ined in reference 1 (page 2) as "one w hic h has the c apa bili ty of changing itsparameters through an in ternal process of measurement , evaluat ion, and adjustmentto adapt to a changing environment. " In a n a i r c r a f t , s u c h a sys tem would automaticallyadapt to flight conditions of varying aerodyn amic control effectiveness without recourset o air -data sampling.

Studies by the U . S. Ai r Fo rc e in 1956 established the theoretical feasibil i ty of de-signing a self-adaptive fl ight control system. The next step, to establish the practi-cality of the concept, was undertaken in 1957 with a num ber of con tract studies. Thesestudies resulted in f l ight-test programs to evaluate adaptive concepts in F-94 andF-101 airplanes. By 1958, the government was satisfied with the practicali ty of adap-tive systems; however, the fl ight-test results were l imited by the fl ight maneuverenvelo pes of the test a i rp lanes s o that the full potential of the system s c ould not beevaluated. To more fully prove the concepts, a true aerospace environmen t wouldhave to be covered in a f l ight- tes t program. Therefore , a program was initiated whichwould result in flight demonstrations in the X-15 airplane, because the rate of changeand range of i ts stabil i ty and control paramete rs, for example, control-su rfaceeffectiveness, short-period damping, and frequency, would provide a seve re tes t ofthe conce pt. A detail ed history of the development of the self-adaptive concept through1958 w a s included in reference 1.

In June 1959, the U . S. A i r Force awarded a contract for the development of theX-15 AFCS. Although the primary purpose of the pro gram was to test a self-adaptivetechnique in a t rue aerospace environment , it was decided to include in the systemcerta in features which had come to e recognized as desirable in aircraft automaticfl ight control systems. These features included dual redundancy provisions for re-liability, integration of reactio n and aerody namic contro ls, rate command control ,and simple outer-loop hold modes in attitude and angle of atta ck. A prototype of thesystem was evaluated on the X-15 simulator and in l imited fl ight-test program withan F-101 airplane.

During the summe r of 1961, the AFCS was installed in the X-15 airplan e. Thefirst flight was on December 20, 1961. Before the X-15 airplane program was termi-nated, the AFCS had been used on 65 flights. Much of the development and many ofthe early fl ight-test results were reported in references 2 to 8. The present repor tsummarizes the experiences with the AFCS for the entire X-15 fl ight program.

SYMBOLS

axaY

az

2

longitudinal acceleration, glateral acceleration, gnormal accelera t ion, g

8/8/2019 19710008947_1971008947

5/30

C

h

K P

Kq

K rL 'a

a

P

'h

'r

'50ecp$

variation of rolling-moment coefficient with angle of sid es li p, pe rdegreeva ri at io n of pitching-moment coefficient with angle of at ta ck , perdegreeaccel eratio n due t o gravity, 32.2 feet per second2 (980.7 centi-

meters per second2)altitude, feet (ki lometers)damper gain in the rol l axis , degrees different ial horizontal- tai lsurface deflection per degree per second of roil ratedamper gain in the pitch axis, degrees of average horizontal-tailsurfa ce deflect ion per degree per second of pitch ratedamper gain in the yaw axis, degre es of rudder surface deflectionper d egree per s econd of yaw rat eroll -control effectiveness, p e r second2rolling angular velocity, degrees p e r seconddyna mic press ure, pounds per foot2 (newtons pe r me te rt ime,secondsangle of attac k, degr eesangle of sides lip, degre esdifferential horizontal -stabilizer surface angle, obtained as thedifference between the left and right horizontal-stabilizer

deflections (aileron deflection producing a right roll is positive),degreesaverage horizontal -stabilizer surface angle, obtained as one -halfthe sum of the left and right horizontal-stabilizer deflections(trail ing edge down is positive), degreesrudder surface angle, trailing edge left is positive, degreesroll-control stick position, degreesangle of p itch attit ude, degre esangle of roll attit ude, degr eesyaw o r heading angle, degrees

3

8/8/2019 19710008947_1971008947

6/30

Subscript:L left

AIRPLANE AND SYSTEMS

X-15 AirplaneThe X-15 was a single-place, rocket-powered airplane (figs. 1 and 2) designed forfl ight research at hypersonic speeds and extreme alt i tudes. It was carr ied alof t underthe rig ht win g of a B-52 airplane and launchedat an altitude of ap proxima tely

45,000 feet (13.7 ki lometers) and a spee d of approximately Mach 0.80. After launch,a powered flight mission was performed, followed by an unpowered deceleration glideand a landing at Edwards Air Force Base, Calif . With this operational technique, theairplane was capable of attaining a Mach number of 6 and could be flown to and re -covered from an altitude in excess of 350,000 feet (106. 68 kilo mete rs). The dura tionof an ave rage free -flight mis sion wa s approx imately 12 minutes.

E-7895Figure I. X-15 irplane.Flights were made with two airplane configurations: ventral rudder on, andvent ral rudd er off (fig. 2) . The stability and control derivatives and physical charac-teristics of these configura tions were presented in detail in reference 9. The static-stabil i ty derivatives of the basic airplane (ventral rudder on) indicated that the vehicle

would be stable through out most of the flight envelope; however, at a Mach number ofapproximately 3 . 0 and an a ngl e of att ack greater than l o " , the airplane was uncontrol-lable without damping augmentation when normal piloting techniques were used.Analysis revealed that this instability was caused by an unfavorable combinationof the4

8/8/2019 19710008947_1971008947

7/30

49.5 (15.1) n-Ventral rudder> 1n

Figure 2. Three-view drawingof he X-15 airplane. (Dimensions in fee t (nwtcrs ) . )yawing moment due to aileron deflection and the dihedral effect. This problem wassubsequently alleviated by removing the ventral rudder (refs. 10 and 11). Although thechange produced lower static-directional stability and rudder effectiveness, it resul tedin a more controllable airplane with stability augmentation inoperative, particularly athigh angles of attack.

DisplaysA photograph of th e pilot's instrumen t panel is shown in figure 3 . The three-axisattitude indicator (shown in the centerof the panel) was the primary instrument fordisplaying airplane pitch, roll , and yaw attitude to the pilot. A pitch-attitude vernierwas provided on the left side of the indicator to give a more accurate indication ofpitch attitude. An instrumen t showing dynamic pressure was located above the attitudeindicator, and a t imer which displayed elapsed engine thrust t ime was position ed n -mediately above the dynamic-pressure display. Errors (difference between presetvalue and actual value) in angleof a ttack an d ang le of sideslip were p resented on thecr os sb ar s of the three-axis atti tude indicator. The angle-of-attack bar was used as av e r n i e r to establ ish a des i red angle of attack. A t high altitude, where the indicationsof angle of sideslip and angle of at tack were unrel iable, the sidesl ip bar was usually

5

8/8/2019 19710008947_1971008947

8/30

switched to present heading error. Actual angle of attack was presented on a dial gageto the left of the attit ude indica tor. Above the angle -of-a ttack indic ator was a presen-tation of normal accele ration .

E-9834Figure 3. X-1 5 cockpit.Conventional pressure-derived airspeed and altitude were displayed to the eft ofthe angle-of-attack dial. These quantities were used at Mach numbers of less than 2and altitudes of less than 80 ,000 feet (24 kilometers). Velocity, alti tude, and rate ofclimb, derived from the inertial reference platform, were displayed in the upper-r ight corner of the panel. A display of roll rate (below the attitude indicator) was in-cluded as a secondary display. The AFCS control panel was located on the forwardpanel to facilitate pilot monitoring.For several special-purpose flights late in the X-15 program, the cockpit displaywas varied somewhat from the arrangement described.

Mechanical ControlsAerodynamic control was provided through the empennage control surfaces usingthe all-movable rudder surfaces for directional control and the all-movable horizontal

tail for both pitch and roll control. The aerodynamic con trol surfaces were actuatedby irreversible hydraulic systems. Art if icial feel was provi ded by force bungees. Aconventional center stick, a contro ller locat ed on the right side of the cockpit (slaved6

8/8/2019 19710008947_1971008947

9/30

to the center s t ick) , and rudder pedals provided aerodynamic control. The side-locatedcontroller, although designed for use in high-acceleration environments, w as used bythe pilots throughout the aerodynamic flight region. This con trol system was describedin detail in reference 12.

Jet Reaction ControlsSmall monopropellant rocket motors were used on the X-15 airplane to producecontrol torques in all th ree axes when the aerodynamic control surfaces became ineffec-tive at high altitudes. This system was discussed in deta i l in reference 13. Manualpilot control of the reaction control rockets was provided by a single, three-axiscont rol s tick loc ated on the left-hand console, as shown in figure 3 . The s t ick wasmechanically connected to proportional propellant-metering valves which regulatedthe propellant flow for the attitude contro l rockets.

X-15 Adaptive Flight Control SystemTh e X-15 AFCS wa s a model following, adaptive variable gain, rate commandaugmentation system and autopilot. The system w a s described in detail in references 2t o 8 , thus only a brief description is included here. The AFCS was installed in serieswith the basic X-15 hydromechanical control system in the same general arrangem ent

as the basic X-15 stabili ty augme ntation system which it replaced ( refs . 12 and 14) .The components unique to the AFCS weighed less than 110 pounds (50 ki lo gra ms ). Th esystem had a dual redundant mechanization throughout, excluding the servoactuators.The simplified diagram in figure 4 illustrates the basic concept of the system.Commands were introduced to the control -surface actuators through conventionalmechanical inputs and simultaneous electrical inputs which were proportional to pitch-and roll-stick displacement. The electrical input to each axis was shaped by R s imple

network (model) which had the dynamic response desired of th e airpla ne in that zxis.The X-15 models were first-order lags with tim e constants of one-half second in pitchand one-third second i n roll. The differe nce in pilot rate command (mode l output) andthe achieved rate was then used as an error signal in the high-gain loop of t he system.The s yste m ope rate d on the principle of using sufficient lead in series with a high fo r -ward loop gain , ac t ing on th is error s ignal , so that the response of the aircraft wouldbe approx imatel y that of the model.The AFCS was designed to providea bandwidth three to five times greater than

tha t of the model, thus insuring that the airplane would follow the model closely. Thegain was automatically increased unti l the system began to oscil late at the verg e ofinstability. (This gain was the "critical*' gain. ) The gain changer, shown in figure 4,operate d by monitoring the l imit-cycle amplitude and adjusting the gain to maintainthe amplitude at less than a preset level. The l imit-cycle frequency, which wasdete rmi ned by the lead compensation, was higher than the aircraft 's natural frequencybut lower than the lowest s t ructura l f requency.

A number of ancillary functions were included in the system that are not shown infigure 4. Because the X-15 servoactuator had a l imited authority ( 3 1 5 of surfaceabout a t r im posi t ion) , a longitudinal trim followup loop was included to provide thesystem with full surface authority. The autopilot functions were additional outer loops7

8/8/2019 19710008947_1971008947

10/30

r Model

IIIII ReferenceII Amplitudeodulator

1IIIIIIII

Figure 4. Simp lified block diagram of basic concep t of X-15 AFCS.( a , e, cp, and $ hold) which he pilot could select t o assist in controlling the airplane.The attitude control rockets were automatically blended with the aerodynamic controlsurfaces to provide continuous control into the high-altitude regions where commandsrequired their use. Automatic limiting of normal acceleration was available to makeent rie s fro m high altitude easier.

The greatly simplified block diagram in figure 4 is typic al of the roll and pitchaxes. The yaw axis did not nclude an electrical pilot nput. The mechanical connec-tion, linking the pilot's control stick with the surface actuators, was not change d fromthe basic zircraft configuration, so that co ntrol of the unaugmented airplane was un-chang ed. Under certa in circu mstan ces, the prese nce of this mechanical pilot inputcapability could have been detrimental to the AFCS performance.

INSTRUMENTATION

During the initial flight-test program, parameters pertinent to evaluation of theX-15 AFCS were rec orded on standard NASA oscillograph recording instruments whichwere synchronized at 0.1-second intervals by a common timer. Later in the program,the data recording system was converted to a pulse-code-modu lation tape system. Thefollowing param eters we re reco rded:Airspeed and pressure altitudeNormal, longitudinal, and transverse accelerationsPitch , roll , and yaw angular accelerations, velocities, and attitudesAngles of attack and sideslipCockpit control positions and forcesControl-surface deflectionsControl servoactuator deflections

8

8/8/2019 19710008947_1971008947

11/30

AFCS gain computersAFCS autopilot reference errorsAFCS command signalsAttitude control rocket-chamber pressuresAn inertial platform supplied velocities, heights, and attitudes. A ball-nose flow-direction sensor provided angle of attac k and angle of sideslip. A tracking radar anda conventional air-data system were also used to obtain veloci ty and al t i tude informa-tion. Internal recordings were accurate within 3 percent of full-scale values for eachsensor .

DISCUSSION

General ExperienceThe flight-test program with the X-15 AFCS began in December 1961. The firstseven flights constituted the acceptance-test portion in the evaluation program. Theseflights were discussed in detail in references 6 and 7. System deficiencies were ex-pected during the acceptance tests; however, before the fifth flight, all known o rsuspected deficiencies had been eliminated , and the remaining three flights were usedfor contractual demonstration and acceptance of the system. A high-altitude flight forperformance evaluation terminated the accepta nce-test sequence. The AFCS had metall the req uireme nts of the performance specifications to which it had been designed.

Velocity,nlsec0 250 500 ' 750 1000 1250 1500 1750 "20x103 After thecceptance -testI I I I I I program,heystemassed on360;ld ' 58lightsyineilotsolyo a320- maximum altitude of approximately354,200 feet (108 kilometers) andwas evaluated with variations indyn ami c press ure fro m 1889 poundspe r s qu ar e foot (90,446 newtonsper square cent imeter ) to near

0 pounds per squar e foot (fig. 5).

280 -

240-

h. 200- -60 f Althougho serioust temptwasft madeoptimizeystem160- ance or to define the total capabil-i ty of the syst em, valuab le impres -sions and experience were obtained.

Much of the experience was notunique to the "adaptive" concept,however.Mos t of theunusualhandling characteristics, whichstimulated pilot comments wer ere su lts of the rate-commandadaptive concept.Velocity,tlsec controleatureatherhanheFigure 5. X-15 right etwclope for the AFCS evd ua t io n .

9

8/8/2019 19710008947_1971008947

12/30

Development ExperienceEarly de velopment of the X-15 AFCS was discu ssed in refer ence s 2 t o 8; only the

significant developmental experiences subsequently found in flight re reviewed in thisrepor t .

Although much study and experience went into the designf the X-15 AFCS in thefo rm of analog-compute r studies and F-101 flight-test experien ce witha system havingthe same concepts , severa l problems were encountered whenbre adb oar d of the AFCSwas connected to the ground-base X-15 flight simulator (fig. 6) . The simulator con-sis ted of a full-size operating mockup of the comp lete control system: cockpit, controlst icks, cables, l inkages , electronic equipment, servoactuators , power actuators, andsimulated control surfaces. Complete six-deg ree -of -freedom airplane motion wascomputed with an analo g computer, which enab led complete mission s to be "flown"from launch to landing. When the actual AFCS hardware was used with the simulator,the first difficulties with the system became apparent.

E-6737Figure 6. X-1.5 flight control sinzulator.

The hysteresis o r lost motion in the X-15 cont rol sy stem w as large comparedwith that of the F-101 and, therefore, presented a problem, particularly during entrymaneuvers from high altitudes. A t high altitude the gains were, of c ourse , drive n totheir maximum values in trying to compensate for the almost complete lackof aero-dynamic control effectivenes s. During entry, the aerodynamic control effectivenes sincreased beyond the value for loop stability with the AFCS gainat maximum, but,beca use of the lost motion, the loop was not closed until a disturbance exceeded thehysteresis. Considerable movement of the pilot's controls resulted from these10

8/8/2019 19710008947_1971008947

13/30

so rapidly that the surface actuatorsmechanical system to thealso occurred to a lesser extent at other flight conditions when,the lack of motion, th e gains drifte d to supercr itical values a nd causedaf period of shaking when the motion exceeded the hysteresis. An example of thisshown in figure 7. By putting flow r estr ic tors i n the s ervo actua tors t o make their

rate of operation mo re consistent with th atof the s urface actuators6 deg/sec), shaking of the pilot 's controls was essentially eliminated. In addition,of the gain computer were cha nged o that the gain was rapidly reducedo a subcri t ical value before the shaking could develop enough ampli tude to be noticed

t

Figure 7. AFCS gain operation and surface activity .Maintaining the gain at a near-critical value when pilot inputs tended to drive itwn mome ntar ily was also a problem. This was most apparent in ballistic fligh t,

2 and 3) . A t altitude, a sharp pilot input coul d drive the gain low enoughits way back up.a high-pass filter and limiter to the servoused by the gain computer. In this way, the effect of the low-frequency, but-amplitude, pilot inputs was greatly reduced without affecting the small- amplitudeency signal necessary for proper gain adjustment. In addition, the reactionallow for dips in the gain values, thus preventingdisengagement of the reaction controls .

Operation of the automatic followup trim system presented a problem, particularlyhigh altitude. The trim system slowly oscillate d, which caused the control stick toat which the trim actuator functioned,of that particular loop.Very late in the develo pmen t of the X-15 AFCS, but prior to flight, a problem wasa flight in which the relatively low-gain basic stability augmentationstructural reson ance of the horizontal s tabi l izers (refs . 14 and 15).

11

8/8/2019 19710008947_1971008947

14/30

Because of its much larger gain values , it was obvious that the AFCS would require anextremely deep notch filter to avoid the structural resonan ce. The high order of thenotch fi l ter required an extensive modification ( r e f . 5). The large additional phase lagof the notch filter introduced another problem: limit cycles. A compromise had to befound, because phase lag had to be reduced to reduce the e f fec tf the l imit cycles. Todo this, it was necessary to increase the ga in at the structural frequencie s. Anacceptable compromise was found only after the maximum and minimum gains werereduced; that is, the variable gain range was shifted downward.

Surface rate limiting, mentioned previously, produced another problem. Even i nthe modified system, it was possible for the servo veloci ty to exceed thatf theactuator and thereby force the actuator to move at its rate l imit . When this occu rred ,the surf ace no longer produced the response requested by the servo, and the airplanewas unable to maintain the commanded rate, part icularly for large inputs on high-gainconditions. In the small roll ste p shown in figure 8(a), vehicle response followed thecommand very well. In the l a rge ro l l step (fig. 8(b)), rate l imi t ing occurred, asevidenced by the slope of the surfac e positi on trac e, and the resp onse de viat ion fro mthe model was significant. These deviations were also reflected in the coupled axes,

6aL l+vaFurface rate limitL

(a ) Snzall roll step. ( b ) ,argeoll stepFigure 8. Sw fa ce rate limitirzg ( X - 1 5 sirnulator).

as indicated by the large sideslip excursio n. In some criti cal area s this satu rati oncould have resu lted in loss of control, particula rly i f the rate limiting had occurredbecause of large pitch-axis commands, which would also have left the aircraft with-out roll control o r damping o r both. This was peculiar to the X-15 airplane becausethe horizontal stabil izers provided pitch and roll control simultaneously.

12

8/8/2019 19710008947_1971008947

15/30

Operational ExperienceOn the first flight, in December 1961, some deficiencies were found in the X-15AFCS. These deficiencie s were associated with airplane and systems interface, so

that the system did not need to be changed functionally. The only functional change-the deletion of two funct ions -during the ent ire program was made prior to the tenthflight. The yaw-rate-to -roll (YAR) and the lateral-acc eleration-t o-rudder (LAR)im p s were ini t ial ly included in the system to compensate for an adverse dihedraleffect which existed in the unaugmented airplaneat high angles of attac k when the ven-tral rudder was instal led. These lateral-direct io nal characteris t ics were discussedin detail in references 10, 11, and 16. When the AFCS was operating near criticalgai ns, the YAR and LAR loops o ffer ed little improvem ent in handling qualities; how -ever, they were thought to be necessary i f the system gains were great ly subcri t ical .When t he deci sion was made in September 1962 to fly the X-15 airplane without the ven-t ra l ru dd er , th e YAR and LAR loops were deleted from the AFCS. In the airp laneventral-rud der -off configurat ion, simulator experie nce had indicated that the se loopstended to destabilize the Dutch roll mode. These experiences refuted the contentiontha t the X-15 AFCS would be relat ively insensi t ive to airplane configurat ion changes.Subsequently, the LAR lo op was reactiva ted, flown for several flights, and found tohave no measurable effect on airplane stability. Thus, it was removed again, sup-porting the co ntention that th e X-15 AFCS could accommodate configuration changes.

Other minor configuration changes were made &ring the x-15 AFCS programwhich SO had no effect on the airplane's handling qualities. Wing-tip pods wereadded and removed; one horizontal stabilizer (one panel) was replaced with a distortedstabi l izer which had been removed from another airplanefter causing a ro l l mis t r im;

With A F C SNo augmentation

and one horizontal s tabi l izer wascoated with about one -half inc h(1.27 cent imet ers) of experimentalablat ive material , producing anrating Unsatisfactoryi lo t airfoi l 1 inch ( 2 . 54entimeters)thicker tha n the stand ard airfoil onthe opposite side of the airplane .

" _Satisfactory I..-;."-- 1"im

Unacceptable -0.0066 pe r d q-1.0 - . 5 0 .5 1.0 1.5 A moreonclusiveonfigura-

cmaobasiction sensitivity study was conductedwith the X-15 simulator in anat tempt to determine the benefi tsthat couldbe achieved by using an"adaptive"controlsystem.Basic

Satisfactory X-15 aerodynamicerivativesere var ied to de termine the compensa-rating Unsatisfactory t ionndaptiveystemouldro-

f a )Lor.lgituditzal stab ili ty.

Pilot ---kBasc y - = - . vide. It was readily apparent hatK

Unacceptable theFCSasxtremelyolerant ofvariat ions in the levels of bas ic a i r -craft control effectiveness and sta-bilit y. Figu res 9(a] and 9(b) show heeffect on pilot ratin g of var yin g the-2 -1 0 1 2%

( b ) Dihedral effect. magnitudes of two basicerivat ives,F'igure 9. Alcgnlcntatiou effccctivetless ofAFCS (x-1sinllrlaror).

13

8/8/2019 19710008947_1971008947

16/30

the longitudi nal stability derivative Cm, and the effective dihedral derivative Clthrough sufficiently wide ranges to render the aircraft unflyable without augmentation.The derivative values are exp ress ed in mult iple s of the basic X-15 valu es. Thedashed l ines represent the unaugmented ai rcraf t which, for negat ive valuesf staticmargin and dihedral effect, was impossible to control . As shown, the AFC S couldabsorb substantial deviations of these p arameters from the design levels a nd stillprovide acceptable handling qualities. These two examples are typical of the expe rien cewith the X-15 AFCS when it was operat ing nearcritical gain.

P'

The AFCS did not operate near critical gai n mo st of the time. An indication ofhow well the gain changer maintained the system gainat its critical value is shown infigure 10 in which the actual gain wander s to either side of the critical value. Duringent ry fro m high alt i tude, the airplane aerodynamic control effectiveness increasedrapidly f rom a value near 0 , which should have caused the system gain to drop cor-respondingly from its maximum value. Because of the mechanic al control systemhysteresis , the drop in system gain was delayed, as indicate d by the large values ofairplane gain while the system gain remained at its maximum value. When ther e waspilot control activity, electrical noise, turbulence, s t ructural vibrat ions, or rapidaircraft motions, the gains were often reduced below the critical gain value.

KP*degldeglsec

.8

-.5

-.6=. 7-

. 4 -

. 3 -

. 2 -

-aE-System

\maximum gain-OD- -"ooo - -

Supercr i t ica l gains

Subcr i t ica l gains

"""" "-ystem mi nimum ga in -0""".11 2 3 4 6 8 1 0 20 30 40 60 80 100

I I I I I l l l l ~ I I I I I I I I I~ 6 , per sec2

Figure IO . AFCS gain-charlger operation.

The effects of the deviatio ns fro m cri t ica l gainon the airplane's handling charac-ter ist ics w ere not usually perceptible to the pilot. Figure 11 shows pilot ratings(ref. 17) fo r a num ber of different fligh t phases for both the X-15 AFCS and a conven-tional fixed-gain system installed in the other two X-15 airplanes. These flight phasescovered a wide range of altitude, Mach number, and dynamic pressure. Thus, the

14

8/8/2019 19710008947_1971008947

17/30

Fixed gainAFCS"-

nI""-0" 3

Pi lo tra t ing

4 ' Launchotationlimbushover( a )Pitch mode. Iaunchotationlimbushover( b )Roll mode.

Pi lotrat ing2

Launchotationlimbushover

Pi lotrat ing

Launchotationlimbushover

ons e cha rac teri stic s of the airplanes equipped with the fixed-gain system varied.the pilot rating s d id not change appreciably nor were they substantially differentlow dynamic pres-be partially explained by the fact that both sy stems provided effective(damping ratios in excess of 0 . 3 ) at reasonable dynamic pressures.no lower thanas in the fixed-gain system, or held consta nt at 0 . 5 , as in the AFCS; he considered

In the low-dynamic-pressure flight phases (refs. 18 and 19) , the AFCS providedstem. For example. during an X-15(figs. 12(a) and 12(b)), the AFCS was effective much earlier in theof its higher gain, and damped out most of the entry misalineme nte in dynamic pressure began. Theearlier phases of the entry. Aircraft oscillatio ns that persisted untilH e could not readily damp theations manual ly b ecause of the rapidly changing frequency of the oscillation asIf he had attempted to manually damp an oscillatione of the entry, he could have instead easily aggravated the motions bypoorly timed or improper control input.

The AFCS also provided the pilot with hold modes and blended aerodynamic and

15

8/8/2019 19710008947_1971008947

18/30

'I-401 I I ~~ 1 I 1 1 I 1 I 1 I I0 10 200 40 50 60 70 80 90 100 110 120t, sec

( a ) With fixed-gain system.Figure 12. X - I 5 atmospheric-entry time history.

16

8/8/2019 19710008947_1971008947

19/30

1200

q,400

0

h ,f t

(b)With AFCS.Figure 12. Concluded.

17

8/8/2019 19710008947_1971008947

20/30

For t hese r ea sons , t heX-15 airplane with the AFCS w a s cons idered to be super ior tothe X-15 airplanes with conventional fixed-gain systems for flights out of the atm os-phere. An X-15 pilot 's assessment of factors that contrib uted to the superio ri ty f theAFCS is presented in the appendix.

The inv aria nt resp ons e c har act eris tics of the AFCS were not obvious to the pilot,becau se pil ots do not usually think of des ired resp ons e i n terms of stick deflection.Before a pilot makes a control input, he has decided what airplane response he wants.If the initial control input does not provide the desired resp onse, he modifies the input.The manifestation of the invarian t rate-c omman d res ponse charact erist icof the X-15AFCS was somewhat subtle. Trim changes normally associated with extension o r re-t ract ion of landing gear, landing flaps, speed brakes, or other variable-configurationdevices were automatically compensated for by the system and were thus masked tothe pilot. Variations in longitudinal center-of-gravity position were also not app are ntto the pilot , because t r im was automatically adjusted and response characterist ics didnot va ry. Transonic t r im changes were masked, too, as were lateral o r directionalasymmetries. The yawing and roll ing moments due to the thrust of a misalined enginewere automatically compensated for within the limits of the system's con trol autho rity.Although the pilots generally appreciated automatic tr im compensation, they weresomewhat uncomfortable about the loss of familiar cues. The t r im change that usual lyaccompanies extension or re t ra ct i on of landing gear or landing flaps serves as apositive indication of respon se to the m oveme nt of a part icular control lever .

Bec aus e of the rate-comma nd featur e, the AFCS did not respond conventionally tochanges in airspeed o r dynamic pressure. The airplane's at t i tude did not change unlesscommanded by the pilot. A s a r e su l t , f o r a neutral control stick position, the nose ofthe airplane did not f a l l through in a conventional manner as a i r speed decreased , nordid it pitch up as airspee d increas ed. Thus, the airplan e had no apparent speed sta-bility. This characteris tic alone woul d not have been too disturbing to the pilot i f theother cues usually associated with changes in airspeed had been available. Theseother cues, howev er, were also missing . There was no change in st ick force, st ickposition, o r aircraft response to control inputs as airspee d varied . The lack of speedstabil i ty was part icularly noticeable during the landing. Figure 13 is a time histor y ofa typical X-15 landing approach, with stick force presented for the conventional fixed-gain system and the AFCS. Airspeed constantly decreased at a rate of approximately4 knots per second during and after flare. In the X-15 airplanes equipped with con-ventional fixed-gain control systems, the pilot was continually working with a positivestick-force slope. H e was usually pulling back on the stick and occasi onally retrim-ming as speed decreased dur ing the level decelerat ion fter f lare , which were normalprocedures . In the X-15 airplane equipped with the AFCS, however, there was noobvious requirement for the pilot to continually pull back on the stick as airspeed de-creased, because, once the flare was completed and the airplane was in a near-levelattitude, the control system would automatically maintain that attitude. In addition, atthe in stant of flare completion, the pilot had to actually push abruptly on the stick (asshown in the time history) to prevent the airplane from ballooning becau se it was at ahigher angle of attack and, thus, higher pitch attitude than required to maintain 1 gflight. If the pilot had simply released the contro l stick, the system would havemaintained that attitude and the airplane w ould have started to climb. After f lare , thepilot was actually working with a negative stick-force slope, as shown in the lowerplot, which felt very unnatural. With the rate-command system, the X-15 pilots re -sorted to tr imming-in a nose-down pitch rate before the f lare to preload the s t ick,

18

8/8/2019 19710008947_1971008947

21/30

thus achieving some semblance of speed stability . Other subtle manifes tations of theinvaria nt r espons e of the AF CS were the reducti on f the influence of ground effectand the elimination of normal dihedral effect.500 152

Radaraltitude, Radaraltitude,ft rn1I I - 0Pul l 40 . - 178 PUII

Slope trendFixed-gain- / / Fixed-gain-systemlongitudinalstick force,"b NTr im nput

-40 I 1 1- 178-178 Pu llAFCS

0 stickorce,stick force, - longi tudinallongi tudinalAFCS

Ib N

-40 10 10 20 -17830t, secFigure 13. Comparison of stick force for the AFCS aud fixed-gain system during a tJlpical X - 1 5 a/)pro[Ic'harld larzding.

A s indicated previously, noncritical gain operation was experienced on all X-15flights and was usually of no consequence. Occasionally, however, when there wasexcessive control activity, turbulence, buffet , or unusual airplane motions, the gainswere substantially noncritical and noticeably degraded system performance and v e -hicle handling qualities. An examp le of gain reductio n caused by electric al noise isshown in figure 14. In this part icu lar instance , the performa nce of the AFCS becames o poor that the pilot resorted to the use of the manual reaction control system, ratherthan continue to use the poorly performing A FCS blended reaction controls.

Early in the AFCS development, it was recognized that large commands from thepilot could not be followed by the control-surface actuators, particularly at high gain.Servo motion was reflected back to the pilot's stick as s t ick kicks, and system in-stability was experie nced becaus e of the inability of th e sy ste m to follow the commandedrate. This problem was encountered on two fl ights in which the airplane became un-controllable in roll for a short period of t ime becau se all available control was being

8/8/2019 19710008947_1971008947

22/30

M i n i m u m

Max imum

Yawservoactuator,de g

t , sec

Figure 14. Time historyof control system servoactuator mo tions and gains show ing effect of lectrical disturbance.

20

8/8/2019 19710008947_1971008947

23/30

axis. A flight record of one experience is presented in figure 15.rate exceeded the recorder l imits dur ing the maneuver , as indicated bydashed lines. The sawtooth-shaped segments of the time history indicate that therate limit was exceeded. The incident w as initiated by the pilot with a ra therlimiting of the servo in pitch p roduce da reduction in the pitch-damper effective-becaus e of lag and caused the rol l axis of the airplane to become unsta ble. One

se of these incident s was that the pitch gain wasat its maximum value, far in ex-of the critical gain value for the flight condition, which made electrical saturationsmall inputs. Other contributing factors are related to the manner ineries with the mechanical system and used the exist ing,and authorityreal meaning, because the pilot could make larger commands through thecontrol system. Such commands immediately saturated the electronic part ofor rapid, the effect on the gain changer

r low -frequency signals which were large enough to saturate the electrical limits, thusto values ex-unstableate l imi t s were reached, at which timeitself.

S e r v o 1 5position, Or- Jdeg -15 I l l 1 I I I

1201 ,Figure 15. AFCS saturation instability.

In addition, the X-15 AFCS was installed in an airplane which used the same aero-ds on the system (eitheror command control) in oneaxis compromised and even prevented effective

axis. Saturation in one axis resulte d in comple te loss of controlrred immediately following recovery from a high-last flight of the X-15 airplane with the AFCS. The resulting lossf damping and effective control allowed the post spin aircraft motions in pitch, roll,aircraft's21

8/8/2019 19710008947_1971008947

24/30

s t ruc tura l limits. A more det ailed ac coun t of this incident is included in reference 20.

ReliabilityA signi fica nt featu re of the AFCS was the redundancy configuration selected to pro-

vide reliability and fail-safety (refs. 2 to 8). Extremely high reliability was requiredbec aus e of the low probability of a successful entry from high altitude without augmen-tation. Fail-safety was equally important because a large transient introduced in aregi on of high dynamic pr es su re would r esul t in destr uctio n of the airplane. Completelydual damper channels, from which either or both channels could control the axis, wereprovided. The adaptive feature of th e ci rc ui tr y pe rm it te d one channel to be lost withlittle o r no loss in sys tem performance , because the remaining gain changerwouldattempt to provide the additional gain required. The gain computers were interlockedwhen operative to prevent overcritical gain following a gain-changer circuit failure andto provide a limiting effect for hardover fai lure. If a model o r variable-gain amplifierfailed, conventional monitor circuits which disengaged both channels were required.This led to the ad ditio n of pa rallel fixed-gain channe ls with fail-safe passive circui try.Because these fixed-gain channels operated simultaneously with the adaptive channelsto avo id the reliability and tra nsient penalties of switching, they effectively limited theminimum gain for adaptive operatio n. These gains were sufficiently high for satis-factory emergency performance throughout the envelope but were ess than the cr i t icalgain in the regio ns of high dynamic pressure (q I 600 lb/ft2 (76,608 N/m2)).

Reliability mDdels ind icate an AFCS mean time between failures (MTBF) of200 hou rs. The basic limitation was the servos, which had no duality. This reliabilitycompares favorably with the fixed-gain system MTBF of 100 hour s. It is interest ingto note that the AFCS electronics had a predicted functional MTBF of 100, 000 hour s.The actu al relia bilit y of the system in flight was excellent. In 65 fligh ts the syst em ex-perienced only two persistent failures which affected system performance. Only oneof these failures was detected by the pilot, who repor ted it as a direct ional mistr im.Eight other electronic component failures occurred during the program but were de-tec ted only by maintenance technicians and did not affect system performance in flight.

CONCLUDING REMARKS

The X-15 adaptive flight control system (AFCS) was evaluate d essentially through-out the X-15 flight envelope , which included a maximum altitude of 354,200 feet (108 kil-ometers) and a maxi mum veloc ity of 5660 feet per se cond (6209 kilom eters per ho ur).The dynamic pressure varied from essentially 0 to appr oxim ately 1889 pounds pe rsq ua re foot (90 ,446 newtons per square centimeter). The AFCS improved the handlingquali t ies of the airplane in some respects . The most s ignif icant improve ment occurredduring atmospheric entry in which the greatest variation in flight conditions existed.

The flight-t est prog ram confirme d certa in adva ntag es of the XFCS: The concepteliminated air-data scheduling; nearly invariant response was provided at essentiallyall aerodynam ic flight conditions; accurate a pr io ri knowledge of aircr aft aero dyn ami cchara cteris t ics was not required to design a satisfactory system; compensatio n for con-figuration changes was provided; and the dual redundant concept provided a rel iableand fail-safe sys tem.

22

8/8/2019 19710008947_1971008947

25/30

Several disadvantages associated with the system were also disclosed: Additionalanalysis was required because f the high-gain values; commands by the pilotat un-t imes; filters were required to prevent sustained resonanceof the s t ructuralsupercr i t ical gain operat ion xisted in flight which, becaus e of mecha nicals and electrical saturation, resulted in divergent airplane motions.

Nat iona l Aeronaut ic s and Space Admini s tra t ion,Edwards, Calif . , November 3, 1970.

2 3

. . . .

8/8/2019 19710008947_1971008947

26/30

APPENDIX

PILOT ASSESSMENT OF THE X-15AFCS

F r o m a handlingquali t ies standpoint, the X-15 airplane equipped with the AFCSwas vast ly super ior to the -15 airplane with the fixed-gain system in ballistic flightand during entry. Because the superiority in these two distinct regions resulted fromdifferent factors, these areas of flight are discussed separately.

Exoatmospheric FlightThe X-15 airplane with the AFCS handled well in ballistic flight for three reasons:

(1) ttitude hold modes, (2) rate command, and (3) an optimum side-located controller.The attitude hold modes provided artificial static stability at a t ime when entry

was imminent and proper attitude was vital. When out of the atmosphe re, the X-15with the fixed-gain system was neutrally statically stable. The responsibility formaintaining vehicle attitude and angular rates within relatively narrow limits wasparamount; every X-15 pilot had explored, on the simulator, the consequences ofallowing vehicle attitudes or rates to exce ed thes e limit s and had o doubt about hisprimary responsibility during exoatmospheric flight.

The rate -command feature in conjunction with an optimum side -located controllerprovided the capability of ach ieving small attitude chan ges accurate ly at low rates ofchange and of making even ma jor attitude chan ges at higher angular rates in one axisat a t ime.

EntryThe fl ight control system characterist ics which made theX-15 with the AFCS a

more desirable entry vehicle than the X-15with the fixed-gain system were: (1)blendedcontrol systems, (2) high-gain aerodynamic damping, and (3 ) use of t he same optimumcontroller for ballistic and aerodynamic control.

The blended control system reduced the pilot 's workload during entry and reducedthe l ikelihood of overcontro ll ing either the ball ist ic or the aerod ynamic sy stem. Thehigh-gain damping maintained pitch and yaw excursions during entry at a significantlylower level than those experienced in theX-15 with the fixed-gain system.

Pilot ObservationsThe true superio rity of the X-15 AFCS was that it unburdened the pilot. The air-

plane was stable at any dynamic pressure and at any angle of attack. The AFCS in-spired confidence and allowed the pilot to spend time cross -checking flight instruments,checking subsystems , and "sightseeing. ''24

8/8/2019 19710008947_1971008947

27/30

APPENDIXAlthough theX-15 with the fixed-gain system was within human controllabilityneed fo r the pi lot to provide his wn static stability during ballistic flight

e X-15 with the fixed-gain system usually experienced pitch and yaw oscillationstotal attention to aircraft control.During boost and after ent ry , the X-15 airplanes handled about equally well; noA F C S were apparent , other than duringflight and entry.

25

8/8/2019 19710008947_1971008947

28/30

REFERENCES

1.

2.

3.

4.

5.

6.

7.

8 .

9.

10.

11.

12.

Gregory, P. C. , ed. : Pro cee din gs of the Self Adaptive Flight Control SystemsSymposium. WADC Tech.Rep.59-49 (ASTIA Doc. No. AD 209389),WrightAi r Dev. C ent er, U. S. Air For ce, Mar . 1959.

Boskovich, Boris; Cole, George H . ; and Mellen, David L. : Advanced Flight Ve-hicle Self-Adaptive Flight Control System, Part I - Study. WADD Tec h. Rep.60-651, Wri ght Air Dev. Div. , U . S. Air Force , Sept. 30, 1960.Mellen, David L. ;Cole, George H. ; and Lindahl, John H. : Advanced Flight Ve -hicle Adaptive Flight Control System. Part I1 - Design of MH-96 System. Tech.Rep.60-651,Wright A i r Dev. Div., U . S. Air Force , Feb. 28 ,1961.Exby, C. B. ; Gray, J . R. ; and Harr ison, J. W. : Advanced Flight Vehicle Adap-tive Flight Control System. Part I11 - SupportEquipmentStudy.Tech.Rep.

60-651, Wri ght Air Dev. Di v. , U . S. A i r Force , Feb. 24, 1961.Reed, M. W. ;Wolfe, J . A. ; and Melle n, D. L. : Advanced Flight Vehicle Self-AdaptiveFlightControlSystem. Part IV - Notch Filter Development. Tech.

Rep. 60-651, Wright A i r Dev. Div., U . S. Air Force , June 1962.Lindahl, J. H . ;McGuire, W. M. ; and Reed, M.W. : Advanced Flight VehicleSelf-AdaptiveFlightControlSystem. Part V - Acce ptanc e Fligh t Tests . WADDTech. Doc.Rep.60-651,WrightAirDev.Div. , U . S . A i r Force, May1963.Lindahl, J . ;McGuire, W. ; and Reed, M. : Advanced Flight Vehicle Self-Adaptive

FlightControlSystem. Part VI - FinalFlightTestReport.Tech.Rep.60-651,Wrigh t Air Dev.Div. , U . S. A i r Force, Oct. 1963.Lindahl, J . ;McGuire, W. ; and Reed, M. : Advanced Flight Vehicle AdaptiveFligh t Contr ol Syste m. Part VI1 - Final Report on Study, Development, and

Te st of the MH-96 Sy st em for th e X-15. WADD Tech. Rep. 60-651, WrightAi r Dev.Div. , U. S. A i r Force, Dec. 1963.Yancey, Roxanah B. : Fligh t Measure ments of Stability and Control Derivativesof the X-15 Research Airplane to a Mach Number of 6.02 and an Angle of Attack

of 25". NASA TND-2532,1964.Taylor, Lawrence W. , J r . : Analysis of a Pilot-Airplane Lateral InstabilityExp eri enc ed Wit h the X-15 Air plane. NASA TN D-1059, 1961.Pe te r sen , Fo r r e s t S. ; Rediess , Herman A. ; and Weil , Joseph: Lateral-Dir ect ion al Co ntro l Ch ara cte ris tics of the X-15 Airpl ane. NASA TM X-726,

1962.Tremant , Robert A. : Oper ationa l Expe rience s and Char acteris tics of the X-15FlightControlSystem. NASA TND-1402,1962.

26

8/8/2019 19710008947_1971008947

29/30

3. Jarvis , Calvin R. and Lock, Wilton P. : Operational Experience With the X-15Reaction Control and Reac tion Augmentation Systems. NASA TN D-2864, 1965.4. Taylor, Lawrence W. , J r . ; and Merr ick, George B. : X-15 Airplane Stabil ityAugmentationSyste m. NASA TND-1157,1962.5. Taylor, Lawrence W. , J r . ; and Smith, John W. : An Analysis of the Limit-C ycle

and Structural-Resonance Characterist ics of the X-15 Stability AugmentationSyst em. NASA TND-4287 , 1967.6. White, Robert M. ; Robinson, Glenn H . ; and Matranga, Gene J . : Re/sumeofHandling Qualities of the X -15 Ai rp lan e. NASA TM X-7 15, 1962.7. Cooper, George E. ; and Harper , Robert P. , J r . : The Use of Pilot Rating in theEvaluation of Aircraft Handlin g Qualit ies. NASA TN D-5153, 1969.8. Holleman, Euclid C. : Control Experiences of the X-15 Pertinent to Lifting Entry.NASA TND-3262,1966.9. Holleman, Euclid C. : Summary of High-Altitude and Entry Flight Control Ex-perience With the X-15 Airplane. NASA TND-3386,1966.

0. ; and Welsh, James R. : Flight Test Experie nce WithAdaptive Control Systems. Advanced Control System Concepts, AGARD Conf.Pr oc . No. 58, an.1970, pp.141-147.

27

8/8/2019 19710008947_1971008947

30/30

ERONAUTICSN D SPACE ADMINISTRATO NWASHINGTON,. C. 20546

OFFICIALBUSINESS FIRST CLASS MAILNATIONAL AERONAUTICSPOSTAGE AND FEES PiSPACE ADMWISTRATIC

07U 001 27 51 305 71043 00903A I R F O R C EW E A P O N SA B O R A T O R Y /WLOL/KIRTL4ND A F B , NEW M E X I C O 87117

POSTMASTER: If Undeliverable (SectionPostal Manual) DoNot R

N A S A

T he aeronautical and space activ ities of the Un ited States shall beconducted so as t o contribute . . . to theexpa nsion of hu ma nknowl-edge of phenonlena n he atvtosphere and space. Th e Adnlinistrationshall provide for he widest practicable and appropriate disseminationof inform ation concerning its activities and theresults hereof.-NATIONALAERONAUTICSN D SPACE ACT OF 1958

SCIENTIFIC AND TECHN ICAL PUBLICATIONSTECHNICAL REPORTS: Scientific andtechnical information considered important,complete, and a lasting contribution to existingknowledge.TECHNICAL NOTES: Information less broadin scope but nevertheless of importance as acontribution to existing knowledge.TECHNICAL MEMORANDUMS:Information receiving limited distributionbecause of preliminary data, security classifica-tion, or other reasons.CONTRACTOR REPORTS: Scientific andtechnical information generated under a NASAcontract or grant and considered an importantcontribution to existing knowledge.

TECHNICAL TRANSLATIONS: Informationpublished in a foreign language consideredto merit NASA distribution in English.SPECIAL PUBLICATIONS: Informationderived from or of value to NASA activities.Publications include conference proceedings,monographs, data compilations, handbooks,sourcebooks, and special bibliographies.TECHNOLOGY UTILIZATIONPUBLICATIONS: Information on technologyused by NASA that may be of particularinterest in commercial and other non-aerospaceapplications. Publications include Tech Briefs,Technology Utilization Reports andTechnology Surveys.

Details on the availability of these publications may be obtained from:SCIENTIFICA N DTECHNICAL NFORMATIONOFFICE