1、REPORT No. 825ANALYSIS OF WIND-TUNNEL STABILITY AND CONTROL TESTS IN TERMSOF FLYING QIMLITIES OF FULL-SCALE AIRPLANESBy GERALDG. KAYTKNSUMMARYfie analytis c$ re8ult8 of wind-tunnel 8taMliiy and control#estaof powered airplane nmde18in tern-uof thejlying qualitimof full-scale a.irpla.neais admcated.
2、In order to indicati thetop”c.supon which cinnmenti are conkdered desirable in the re-port of a wind-tunnel stability and control inw in- stead, the wind-tunnel data are usually presented in coeffi-cient form in a voluminous series of curves and tables.As judged in flight testing or service, the fly
3、ing- qualitiesof an airplane are characterized not by dimensionkxs coefE-cients but by forces, velocities, accelerations, angles, and othermeasurable quantities which actually define the stability andcontrol oh.aractwistics of an airphne in flight.Various means may be employed for determining fromwi
4、nd-tunnel test data the particular dimensional valuesdescribing the airplanes flying qualities, but this type ofaudysis has not generally been considered the provinco ofwind-tunnel peTSOnnd. It is believed, however, that thevalue of wind-tunnel tests wodd be increased considerablyif an analysis of t
5、his nature were imduded in every stabfityand control investigation. Although the +mIysis wouldaugment rather than replace the measured data as usuallypresented, the greater portion of the wind-tunnel report wouldconsist of 8 discussion of the actuzd flying qudit.ies of theairplane. The inclusion of
6、such a discussion wouId eIiminatethe confusion oftm caused by mere presentation of thetest remdts, facilitate the practical application of tunneldata, and provide assurance that no flight difficulties willpass undetected because of failure to put the accumulatedinformation to its proper use. Moreove
7、r, test programs d+signed for this purpose cou.Id be planned more efficiently withregard to the amount of required testing. Many programs inthe past have been Iaid out arbitrarily without a completeunderstanding of the manner in which the resulting datashould be appIied. This lack of understand has
8、at timesresulted in insufEcient data concerning trim conditions andconsiderable unnecessary data for untrimmed ocnditions.The purpose of this paper is to outline a suggested formof presentation of the results of a stability and control inves-tigation in terms of flying qmditiee as de.6ned .refereme
9、1and to systematize and review briefly the analytical workrequked for this type of presentation. LNOeffort is made tospecify deilnite test procedures.Reference 2 contains a review of testing technique for usewith powered wind-tunnel models and a fairly complete dis-ction of most of the standard test
10、s necessary for the ml-lection of data used in the suggested amdy%. It shouldprove useful in the preparation of any stability and controltt progrwn. Although ht measurements and observa-tions are made with respect to the airplane body ases, theuse of the “stability” axes as recommended in reference
11、2wiU probably be satisfactory for the nomml range of testconditions, and these axes are used throughout the presentpaper.It is asamed that alI necessary tumd corrections will bemade before any amdysis is attempkd and that the measureddata will be sufbiently accurate for use in predicting flightchara
12、ctqiatics with reasonable precision.The NACA requirements for satisfactory flying qualities,as explained in “detaiI in reference 1, are used as a basis forthe procedure suggested herein and nstitute the list of sub-jects on which it is believed comments should be made in thepresentation of wind-tunn
13、eI data reIating to stability andcontrol. The complete series of tests is not considered essen-tiaI for every airplane; the list of requirements is included inits entirety for the purpose of pointing out the desired formof analysis for any phase of stability and ccntrol investigated.It is realized,
14、of course, that the requirements for satis-factory flying qualiti= may undergo constant revision withtime. By methods simiIar to those indicated in the presentpaper, however, wind-tunnel tests may be used for the in-vestigation of any revisions of the present requirementsor for the investigation of
15、a completely different set ofcriterions.523Provided by IHSNot for ResaleNo reproduction or networking permitted without license from IHS-,-,-524 REPORTNO. 825NATIONAL ADVISORYWh_TTEE FGRAERONAUTICSFor purposes of clarity and convenience, each of the presentNACA flight requirements is given k“ the te
16、xt, accompaniedby recommendations regarding its relation to tunnel test.Unless otherwise specified or implied, the requirementsshoulcl be investigated for all conditions of flight, speciaIattention being given the conditions that appear to be “themost critical.COEFFICIENTSAND SYMBOLSa.irphme. weight
17、, poundsarea of wing (unless accompanied by subscript a, e, or rdenoting aileron, eIevator, or rudder), square fretwing span, feetmean aerodynamic chord (M. A. C.), feetroot-mean-squme elevator chord, feetroot-mean-square aileron chord, feetroot-mean-square rudder chord, feettail length (distance fr
18、om center of gravity to elevatorhinge line), feetangle of stabilizer setting with respect to thrust line,degrees, positive When leading edge is upaileron deflection, degrees, positive when trailing edgeis downelevator deflection, degrees, positive when trailing edgeis downrudder deflection, degrees,
19、 positive when trailing edgeis to leftairplane angle ot attack (thrust line), degreesangle of attack at taiI, degrees -”angle of sidedip, degrees, positive when right wing isforwardangle of yaw, degrws, positive when left wing isforward (*= p)airspeed, feet per secondmass density of air, slugs per c
20、ubic foot()dynamic pressure, pounds per square foot pV2airplane mass, SIU9, (W/gJ()-mrelativedensity factormLift()lift coefficient qspitching-moment coefficient about center of gravity(Pit although fulLpowerstability below 120 percent of the minimum speed is not anrdxdute requirement., it is highly
21、desirable.A suftlcient number of eleator deflections within the trimrange should be tested to provide for a dependable curve ofMUng powor; T, -O.Provided by IHSNot for ResaleNo reproduction or networking permitted without license from IHS-,-,-ALYSIS OF WIND-TUNNELTESTSW TEBMS01 FULJ. -m; T. -0.The c
22、let ermination of the ele-rater characteristics inaccelerat ecl maneuvem involves the use of pitching-momentand hinge-moment curves for both stabilizer and elevatorvariatio, as shown in figures 4 and 1. These curves shouldbe plotted for a constant thrust coefficient, the value ofwhich is det wmined
23、by L and the tat ed power for themaneuvering condition. If the initial lift coeflhient investi-gated is 10T (as may be expected in the case of purstitahpkines if Norma values of load factor and CL_ are used),the thrust coefficient may be suftlciently close to zero topermit the use of idling or zero-
24、thrust data.As indicated by the equation, the elevator deflection muejsupply enough pit chhg moment to balance the two factorsdcmAC. and Acq . The amount of deflection necessary tobalance AC. is merely the difference between the elevatorsettings for trim in steady fLight at CL and at CLu. Theadditio
25、nal deff ect ion is required to oercome the dampingof the horizontal tail. or a conventional airplane, thedamping effects of the wing, fuselage, and other airpknecomponents are considered negligible in comparison withtlie damping of the horizontal tail.)The use of equation (2) cm best be demonstrate
26、d by asample solution for a steady-turn maneuver for a 6000-pound airplane with a tail length of 16.5 feet and a wingarea of 250 square feet- The airplane has a maximum up-eIevat.or travel of 25 for a 10-inch travel at the top of thecontrol stick, tith an essentiality linear variation of 66with z.Pk
27、h this deflection is within the limit ofavailable tra-rel.I-C-2. “The variation of elevator angle with normal.acceleration in steady turning flight at any given speedshould be a smooth cur-ie which everywhere has a stableslope.SII all previous criterions are assumed satisfied, this require-ment will
28、 be met if the pitching-moment, curves of figure 4 gives 0”0006 = he aue or% Thecomputations then proceed as follows:=0.524=-0.0013w/Js 0000 PP=250X0.17 =141.2F=o.524x(0.02860.0013jx141.2X(L4x3=51.5 poundsThe force gradient, then, is 51,5-=6.4 pounds pm g,which is reasonablj close to the spccifkd li
29、mit for p u reuiha.irplanesj and tho force is greater than 30 pounds.It is possible, through the use of the principle of axis rot it is believed that the ground effect on lift issufficiently smaII that it may be ignored in this problem.A “ -J.o-. .0 6$ .i!-20$ :$-.30 . .- .o f b .*.$,04.#j / 7 .to 0
30、 .lx y$“,c/$ Y /.- I0 !0 20 50 40 50Tofu! uileron defecf ion, 6=tir,degFIGURE 13.Atleron tab obamcterkt!m.Provided by IHSNot for ResaleNo reproduction or networking permitted without license from IHS-,-,-ANALYSIS OF WIND-TUNNEL TESTS IN TERMSlimits of travel should provide zero hinge-moment at anail
31、eron deflection required to balance this amount of rolling.On figure 13, for exarupe, a tab angle of approsirnatdy 10meets tie requirement.A sirnihir procedure can be used for the rudder. It mightbe desirable, if feasible, to study the requirement in a moredirect manner by making straight-flight mea
32、surwnmts ofrollii-moment and yawing-moment coeffients at varioustab settings with rudder and properly Iinked ailerons freeto float. The requirement would then be met if Cmand Clwere found to equal zero at points within the limits of tabtravel.Because the study of this item depends on the action ofsm
33、all tabs, the investigation of this requiremmt will probablybe adviaable only on modeLs of fairly large scale.III3. “Urdess changed manualIy, the trimming deviceshould retain a given setting indeiigitely.”This item does not appear suited for nornd wind-tunnelinvestigation.111. Stalling characteristi
34、cs111-1. “The approach of the complete stall should makeitself unmistakably evident through any or W of the follow-ing conditions:FL.The instability due to staling shouId develop in agraduaI but unmistakabb manner.”Tuft studies appear to bo mandatory in connection withthis phase of the investigation
35、 and -ivith the stalling problemin generaL Information concerning the flow phenomena andhence, to some extent, the behavior of the airplane at thestall can be acquired from observations of the action oftufts on the model as the angle of attack is increased untdthe complete stalI is reached. The deve
36、lopment of insta-bility is usually graduaI when stalling appears that is, the direction and sharpn- of pitching-momentbreaks at and beyond the std, and the apparent degree ofcontrol effectiveness beyond the stdlShotid bear a directrelationship to the possibihty of recovery from stalled ilight.Furthe
37、r investigation may be required as to the nature andconsistency of this relationship.III-3- “The three-point hmding attitude of the airplaneshould be such that rolling or yating moments due tostalling, not easily checked by controls, should not occur inIanding, either three-point or with tail-first
38、attitude 2greater than that for three-point contact.”The angle of attack corresponding to the three-pointground Me of the airplane can be determined from drawingsof the complete airplane. The angle of attack at which tuftstudies indicate bad stalling in the flapdo=, idling-powercondition should be m
39、ore than 2 greater than the groundangle.“CONCLUDING”REMARAR attempt has been made in the prwnt paper to presentmethods of analysis of wind-tunnel tests in terms of flyingquaIities of airplanes. The suggested methods have hem . . _presented in an efFort to demonstrate th practicability ofthis type of
40、 analysis and aka to stinmkte. interest anddiscussion among design and test persormeL It is hopedthat with the cooperation of interested groups, the presentmethods dl be extended, improved, refined, andmostimportant-proved by application and corrdat.ion.In the present rather general treatment of the
41、 subject, ithas naturaIIy been impossible to cover mmsual cases thatmay require special treatment. If, for example, an airplaneis known to be provided with some mechatical device thatirdluences the control forces in certain maneuvem althoughthe measured items in the wind tunnel show no effect, thisd
42、evice should be considered in the study of the relevantrequirements. In short, every eflort should be made toregard the subject of the investigation as an actual flyingairplane and not as a scaled-up reproduction of a model.In conck.ion, it is believed that wind-tunnel tests ofpowered models can, if
43、 properly analyzed, be used to examinethe flying qualities of airplanes and to determine the extentto which any particular airplane will satisfy requirements forsatisfactory stability, cantrol, and handling characteristicsin fight. It is recommended that this type of “testing,analysis, and presentat
44、ion of data be generaIIy employed inwind tunnels engaged in testing airphme models for stabilityand control. -L.MiGLEY MiXORLAL AERONAWHCAL LABORATORY,hTm?roNALADVISOBY COMMITTEE FOR AERONAUTICS,LANGLEY FIELD, VA., April 1, 194S.Provided by IHSNot for ResaleNo reproduction or networking permitted wi
45、thout license from IHS-,-,-APPENDIXSIDESLIPIN AILERONMANEUVERSSYMBOLS AND DEFINITIONSThe folowing terms, in addition to those previouslydefined, are us.e in the computation of the sideslip angle:yawing-moment coefficient due to yawingyawing velocity, radians per secondradius of gyration about 7axis,
46、 feet(nzkz=Yawing moment of inertia)anglb of bank, radianslateral-stability derivatives in terms of unitmoment of inertia of airplaneTho value of (?np for the appropriate angle of attack isfound by -the US? of reference 12. Reference 14. may beused for the estimation of C.,.SIDE9LIP FORMULAAs stated
47、 in item (II-C), the angIe of sidealip developedby a conventional airplane in a rudder-fied aileron roll maybe exprwa-d as=57.3 k,+k, cos pt+k, sin -e” (k, cos Bt+k6 sin Bt)(5)wherepbc,. FjV-1-C.ak,=57.3 k4=k1+k2k,=pv - -After the constants have been evaluated, the sideslip anglemay be determined fo
48、r any value of time by substitution inequation (5). This substitution should be made for half-second or smaller intervals covering a range of time sufficientS40to allow the airplane to reach an angle of bank of 90 (or leesin the case of an extremely large airplane).DISCUSSIONThe angles of bank rwwhc
49、d in a full-aileron roll arc far toolarge to permit the use of the usual assumption that d=siu #.The custmmary solution must thereforo bo further complicatedby the introduction of the sim of the rmglc in thu equationsof motion.The expression presented herein oflcrs a somewhatsimplified solution but suflws a corrcspcmding loss in nc-curacy. The greatest possi
