'/^M^-/^^ \y NATIONAL ADVISORY COMMITTEE FOR AERONAUTICS WARTIME REPORT ORIGINALLY ISSUED August 19^2 as Advance Eestricted Report A MASS-DISTRIBUTION CRITEEION FOR PREDICTING THE EFFECT OF CONTROL MANIPULATION ON TEE RECOVERY FROM A SPIN By A. I. Nelhouse Langley Memorial Aeronautical Laboratory Langley Field, Va. NACA WASHINGTON NACA WARTIME REPORTS are reprints of papers originally issued to provide rapid distribution of t advance research results to an authorized group requiring them for the wsLr effort. They were pre- ♦ vlously held under a security status but are now unclassified. Some of these reports were not tech- nically edited. All have been reproduced without change m order to expedite general distribution. L - 168 DOCUMENTS DEPARTMENT Digitized by tine Internet Arclnive in 2011 with funding from University of Florida, George A. Smathers Libraries with support from LYRASIS and the Sloan Foundation http://www.archive.org/details/massdistributionOOIang 110 O^f Hi NATIONAL ABVISORY COMMITTEE POK AERCNAUTICS ADVAFCE ilESTRICTED REPOHT A MASS-DISTKIBUTION CRITEEIO!J FOE FBEDICTING TEE EFFECT OF COrTROL MAIHIULATION OH THE EEC0V2JRY FROM A SPIN Bj- A. I. Neihonse SUMMARY Res-alts of s-cin-tunnel tests of 65 models indicated that when the airplane design aimulated that of the ear- lier single-engine type, with mass d i s t ri'b'ut ed chiefly along the fusela^^e, aileron-with and elevator- np settings aided recovery, and the rudder was the predominant con- trol for recovery. When the dcsisu approached the design of multiongine airplanes (or the more recent single-engine airplanes with wint, tanks and v. ing armament) with the mass distributed chiefly along the wings, ho'vever, aileron- against and elevator-down settings were conducive to the most r'^pid recovery and the elevator was the predominant control . The primary importance of the mass distribution of an airplane in determining itr, spinning characteristics is demonstrated and a useful criterion for predicting the optimum control manipulation for recovery, based on a non- dimensional mass-distribution parameter, is presented. Charts that should be useful for such predictions to both the pilot and the designer are included. INTRODUCTION During the past 5 years, 65 models, representing air- planes covering a wide range of dimensional and mass de- sign characteristics, have been tested in the NACA free- spinning wind tunnel. As is to be expected, these models have shown varied spin and recovery characteristics, re- flectin:5 the differences in the proportions and mass dis- tribution of the models. A consistent difference, however, in r.pin and recovery characteristics was early apparent between modal liiS'ntly ioade the win^s. I and not ?. erod tor causing t undertaken fo tests have te Por these tes varied and in chief Ij" along was distribut chiefly alon? t ion reversed s he-^-vily loaded along d along the fuselage, n an effort to establi j'namic cha ract er ipt i cs his difference, a seri r many of the models, en acci'mulated for 19 ts, the mass distribut dels Those mass distri the fusela&e were r^l ed chiefly along the ^ the wings liks^-ise ha the fusel or he^^vily s h ma s s d i , a s t h e p c s of sp ec and result represent a ion of eao but ion was oaded unt i ings. Mod d their ma age and those loadec". along s t r i bu t i n , riirary fac- ial tests ~as S of ST^ch tive designs, h model was originally 1 the mass els loaded ss distribu- A qualitative analysis of the results was obtained for 65 modela tested in the spin tunnel, as well as of the results of special tests for 19 of these models. Definite rules have been formulated concerning the effects of control manipulation on the recovery from the spin, as influenced by the airplane mass di 13 1 ribut io n. A criterion based on a nondii:ensionai mass-distribution parameter has been e3tf:.bli shed for predicting these effects. AIJAHaTUS AlTD TESTS The sp in-t est iuc: technique in the waCA free- spinning wind tunnel and the construction of spin nodels are de- scribed in dctril in reference 1. The models, constructed of balsa, arc ballasted for dynamic similarity to the cor- responding airplane by installation of proper weights at 3\^itable locations. An automatic clockwork delayed-action mechaninm or a ma.^netic rcmot e- cont rol mechanism is in- stalled in the model to actuate the controls for recovery. The model with the rtidder set --ith the spin is launched in the ypin the tunnel, rate 01 des a fixed hei generally a full with t may be arra recovery is munt of the tior.. The charact oris oy hand into the vertical upward air stream of The airspeed is adjusted to equal th? vertical cent 0"^ the model pnd uhe model is thus kept at ght until recovery is attempted. F.ecovery is tteir.pted by reversal of the rudder alone from o full against the spin, although the mechanism ■iged to riove any or all cf the controls. Tau .judged by the nuuber of turns from the move- rudder to tiiC C'3srat ion of the spinning rcta- effect of aileron setting on the spinning tics is usually evaluated by a comparison of the numlDer of tTirna necesspr/ for recovery "by rudder re- versal alone from spirs for '^hich, for example, trie ailer- ons are set (not moved) -^ith the spir* (rifeht aileron up in a right spin) and the nuicher of tvirns necessary for recovery from spins for "yhiclx the ailerons are set against the spin. Results of spins in rhich the elevator is full up are compared '^-ith results obtained for spins ™ith ele- vator neutral or full dorn. In a f exc instances, for the special tests, the effects of aileron and elevator set- tings have "been "based on a comparison of the vertical speed and the attitude of thj steady spin. The Eodels tested in the spin tunnel have covered a 77ide range of dimensional and mass characteristics and include seaplane and landplane, bip3ane and high- and Iott- wing monoplane types, and multiengine and single-engine designs. The 19 models used in the special tests repre- sent different typer-. For th«^ special tests, the mass characteristics were varied oy moving ballast Treights from either the ^ing tips or the fuselage extremities to the center of gravity or "by moving ballast "heights to either the wing t ir s or the fuselage extremities from the center of gravity, the position of the center of gravity being kept constant. RESULTS The data analysed are presented in figures 1, 2, and 3. These figures are an atten-pt to represent graphically, by a single point, the important 'nass-distribiit ion char- acteristics of each model. In table I the models are given numerical designations to permit their identifica- tion in the -figures. In the Su.ler equabicns of motion, the influence of the mass distribution depends en three factors: I^ - ly , ly - Ij. and I2 - Ix.> where I^, Iv . sni I^ are the moments of inertia about the X, Y, and S body axes, respectively. For presentation in the figures, these factors have been -zade nondimensional by dividing by rab" , where m is the r.ass and b is the span of the airplane. The parameter -= was taken as the ordinate for the mb^ figures. This parameter is a factor affecting the inertia 4 pitching moment and increases rhen msss is added along the Ty - T7 fuselage. The atscissa — ,j-— is the factor affecting mt) the inertia rollin.^ rcCiTeut and the neg'ative valf'es nntaer- ically increase as weight is a(?ded along the '7ings. Inas- much as the sura of the tnree mass parameters is equal to Iv - Iv zero, the val'ae of the third 'raramst <; r , — , may ho mo ■-0 indicated by a third scnle at 4b to the ordinate Hud. ab- scissa scales. Tnis tnird part?.meter is a factor affect- ins; the inertia ya;;ing- mon-ent , the iprge positive v-ilvtes i^idicating that the mass dist; ritut ior is chiefly -.ilong the wings and the large ne,.2;.at i ve values indicating that the mass distribution ib chiefly alon^ the fuselage. -he three parr.met ers m-iiy also be ?/ritten as —- — — — — - ky- - kj;' and — — -j^ , respectively, i^rhere k^ , k-^ . and kf the radii of ryr^.tion abo\-'t the X, Y, ;\nd S axes, respect ively. are figure 1 recovery chnra Aileron data v.' The type of po indicates whet at;ainst the sp 2 gives similg available for v;hcther elevat arc more favor results of spc d i R t r i bii t i n . of the Ram.e no the latt(>r "a" mal loading co of both a.ilero s ho '7 s the cterist ic ere avail ints \is3d her K c 1 1 i in reduce r informa 60 of the or- up set able for c ill t est In this del are r is er^plo ndit ion. n s a n d el effect as i V. d Lie for to d e 3 ^ i-: the r t h e t u ion for models . irj],s or ceo very of 19 j £ u r f , p resent c d to d The sym va.tor s i c a t onl g n -^ t iler rns tne Th el e . F mo ie diff ed b enot bols ett i ileroii setting on the ed by routine tests. y 53 of the m.odrls. c the different models ons with the sr)in or for recovery. Figure cle"ator, dat-?. being e points indicate vator-doTvii settings igure 3 presents the Is "'ith altered mass ere nt ma s s -•■ r r a ng em c n t s y tne same number and e the altered or abnor- indicate the effects DISCUSSICK Q£ik§.?lloL_l2.r._PXeil?.Li2.:i_9Ll_2.P.LL^2.1_?.lll5.5.1s. - An in- spection of the figures sho'vs a distinct grouping of the points representing tne different effects of control scttir.j^s. Fartir.l seraration of the effects is ottained ty independent consideration of each of the three mass parameters. The most complete separation, however, appears to he given ty consideration of the inertia ya^ing-moment parr.Liet er mb" Examination of figure 1 indicates that at a ^'alue of the inertia yaring- moment parameter ■X mh' ,_i of - 5C X 10 ,^ almost complete separat place. loT larger nega usually had a favorable istics and ailerons aga effect. As the paramet proached, instances ^er had no noticeable effec For negative \'alues of than -50 X 10"^ and f effect reversed so that had a favorable effect tings '.7ith the spin rer this reversal value, a it appears that only si tion may completely rev tion to the j^reneral rul only one instance. ion of the a t ive values , effect on t inst the sn i er value of e observed '- t on the rec the para mete or positive aileron set on recovery ,• e detriment a crit ical re^ ifht variati erse the ail e '-as obtain ileron effects takes ailerons 'rith the spin he recovery character- n had an unfavorable -50 X 10~* was ap- here aileron setting overy characteristics. r numerically smaller values, the aileron tings against the spin rhereas aileron set- 1. In the vicinity of ion existed for rhich ons in mass distribu- eron effect. An excep- ed in this region in The effect of data of f i^ure 3 , a value of the yar? peiirs to be a crit ±20 X 10 ~* in V7hi bo in cither direc eter numerically g settings vere usua In several instanc very flat or very little or no effec greater than 20 X settings vrere very satisfactory recov could be obtained spin by full rudde the elevator alone tion gave satisfac elevator settings, according to the tends to reverse in the neighborhood of ing-moment parameter of zero. There ap- ical region between the values of ch tne effect of ele'"'ator settings may tion. For negative values of the param- reater than -20 X lO" , elevator-u-p lly conducive to most rapid recovery. es, horever, for models that gave either steep spins, the elevator setting had t. For positive values of the TDarameter 10 , on the other hand, el evator-dovrn definitely instrumental in effecting ery. In an extreme case, no recovery from the elevator-up, aileron-neutral r reversal alone; Tvnereas movement of from the full-up to the full-dov7n posi- tory recovery. The data from the special tests for IS models, given in fi£:ure 3, appear to prove that the separation indicated for elevator and aileron effects in figures 1 and 2 depend predominantly on the mass di st ritut ion of the models rath- er than on aerodynamic factors. The 19 models tested are helieved sufficiently rer-resentat i ve of different airplane types to permit a generalization of the conclusion. Model 15, for exairple, represents a lightly loaded, single-en- gine reconnaissance monoplane rhereas model 5 represents a high-speed, heavily loaded, tr in- engine attack airplane. It must he a-cpreciated that aerodynamic factors may modi- fy the results for some combinations of mass arrangement and extrem.e aerodynamic design to the extent that the con- trol effects may be dictated by the aerodynamic character- istics. Sequence of control raanirulation for recovery; . - The conclusions drav-n froir. the figures are particularly sig- nificant in that they indicate that the relative impor- tance of the different airplane controls for recovery from the spin may change radically bet^^een airplanes of differ- ent types. prior to the recent extended application of wing armament for combat types, airplane structural design procedure was such that the airplane was characterized structurally by relatively light wings. practically all the disposable load was carried in the fuselage, although som.e gasoline might be carried near the center of the wings. These ch'i ract er ist ics still apply to the private- o\Yner class of airplanes. This structural arrangement of .the airplane results in a mass loading chiefly along the Ir- Iv fuselage and the value of —— — =_± will tend to be large mb^ and positive, while the value of the inertia yawing-moment I Y - I ^- parametcr — — is negative. The installation of mb^ wing engines tends to increase the weight along the wings and it can therefore generally be assamed that multiengine airplanes have high negative values of the parameter y"" Ix""1y ^ — , and positive values of the parameter — ■ — ;; • mb" mb- Prosent-day military design of single-engine airplanes is also toward heavy wings. Tnc desire for increased range has increased the amount of gasoline carried in the wings. Guns and ammunition are carried outboard of the propeller, and the metal wings with the mechanism for retracting the landing gear are inlierently heavier than in older designs. The results of the model tests sho" that, for the earlier singl e- engine military design and the present-day -rivately OT^ned airplanes, the rudder is generally the predoir.inant control for recovery fron the spin and that f-:ill rudder reversal is the nost effective control manip- ulation. Movement of the elevator to the dovn position before the reversal of the rudder tends to shield the rud- der and retard recovery; V7hereas, movement of the elevator after the rudder has been conipletely reversed and rotation has begun to sIott up may offer a fa'^oraole pitching moment, tending to aid recovery 7:ith?ut adversely ■ aff ect ing the rudder action. iiovement of the elevator alone rarely gives recovery. Because high rates of descent v-ill probably be associated ^rith recovery T^^ith full-dovn elevator, the amount the elevator is moved do^rn will defend on lno^ tvacli assist- ance is needed froa the elevator to produce a satisfactory recovery. The effect of ailerons -"ill be contrary to the effects expected in normal flight and holding the ailerons against the spii; -ill retard recovery; -hereas holding the ailerons '-ith the spin '-^ill assist recovery. For nultiengine airrlanss and for the more recent s ingl e- eng ine military designs, tho elevator tends to be- come the predominant control for recovery. The movement of the elevator dov^n is essentiel to a rapid recovery. Eudder reversal, although of less importance than eleva- tor reversal, T-ill generally irprove recovery. Aileron position is critical and aileron settings -^ith the srin may greatly retard recovery; whereas aileron-against set- tings will be favorable. All controls for airplanes of these types have the effects that -'ould be expected of tnem in normal flight. It may be said in summarizing that, for airplanes of relatively light loading along the wings, full rudder re- versal before moving the elevator doT^n is imperative; mov- ing the elevator down after the rudder reversal is desir- able. For airplanes heavily loaded along the wings, mov- ing the elevator down is imperetivc; full rudder reversal is desirable. ArT:li cation to fl i.;ht.- The values of the criterion at which the aileron and elevator effects in the spin reverse, as sho-n by the figures, apply strictly to mod- els only. The general conclusions, however, should be applicable to flight, although, because of possible scale effects, the reversals nay occiir in flight at somewhat different values of the ci'iterion than are indicated ty the tunnel data. The meager conparative flight data availaole indicate that the values for the reversal of aileron and elevator effects will prohahly be changed somewhat hut there are not enough full-scale data avp liable to fix the flight val- ues. It is desirable that aore flight data "be obtained in an effort to establish definitely the values in flight at which the aileron and elevator effects reverse. Ex/clanation of mas_s_ _ef f ec t_s . - A possible explanation of the dependence of the effectiveness of the elevator and ailerons on the mass distribution is presented briefly The application of Euler's dynamical, equations to the case of an r.irplane in a stecidy spin gives, for the iner- tia yawing moment about the body axis, the expression (I^ - ly) sin ngle of wing tilt to the horizontal, .. _ _.' _j _-.-i. _;.-_ J- down is the a positive v.'hen right wing is a angle of attack Q angular velocity .>->bout spin axis For a spin in any the inertia yawing s i gn s of I ^ ■Y the tunnel results the spin leads to the ailerons again of 4= . For models setting the ailero able effect in tha ative and will act rection of rotatio designs where ly spin will produce of the spin. The chart 1, the rever when Ijj — ly is given direction, the algebraic sign of moment depends only on the algebraic ai:id the angle . In a right spin, indicate that setting the ailerons with a positive value for <+>; whereas setting st the spin leads to a ne^iat ive value loaded so that ly - ly is negative, ns T-tith the spin will produce a favor- t the inertia yawing moment will be neg- to turn the airplane away from the d i- n (against the spin). Coi.versely, for - ly is positive, ailerons set with the an inertia yawing moment in the direction fact that, for the results presented in sal of aileron effect does not take place zero can be attributed to secondary aerodynamic factors. A similar explanation may "be applied to the elevator effect, as the model results indicate that setting the elevator up usually leads to a positive value of and elevator down to a ncgati\'-e value of

TaTION OIT FI&UEE3 [Unless other"'ise iiidicTted, cocpar*' t i\'e recovery data available for toth r.ilerons and. elevators] Model Airplane 1 Uodel Airp lane designat ion represent cd j decignat ion r e-cresent ed (a) 1 1 aF-50 34 r3-2A 2 XE-2 35 XF-46 3 X5-A3-3 36 Xi:51I- 1 4 XF dF- 1 37 S3'-l 5 XF-38 38 tXOSlT-1 6 YFi'/:- 1 39 ::f-i 7 .^-20 40 XFoF-2 8 Yi2-1 41 F-3 6A 9 BT-14 42 XF4F-3 10 XF4-J-1 43 ITaCa C-ilTAED 11 E-25 44 F-39 12 01J31T-3 45 XF4F-2 13 ^Nzr-i C43 P-2 6A 14 BT-9 C47 eX0Si:-l 15 C-52 48 ^XSB2U-1 16 XSS-1 49 ''^XOSS-l Cl7 V-143 ^50 Model 159 18 FB-2 C5I X3FB-1 19 F-35 Co2 ^X0S2U-1 20 A- 17 53 XF-40 21 XlT-13 54 yp-37 <^22 XP 2.-^-2 55 XS32A-1 23 XSE2 3-1 ^56 ^X0S2U-1 ^3-± F2A-1 57 ^XOSS-l 25 XBI-11 53 1 ^■X303C-1 23 XK3^--2 59 1 ^XSOSC-l 27 ^:T3r-3 ^60 XN3Y-1 28 XFL-1 eei F43-2 29 XF2A-1 C62 F2F-1 50 XBT-12 63 S3D-1 C51 -V-143 ! °^ YF-43 32 F-44 55 XF-47B C33 XF3F-1 j ^Letter "a" ucrraal a ^Seaplane. ^ Conparat iv ^Lengthened after model number indicates loading varied from s i-'-dicated. ^Comparative recovery data for elevators only. ly. ^Landplane. recovery data. e recovery data for oilerons empennage. Sjjo comparative NACA Fiptl J,. 01 ' OOP V) C o o T3 •o ■o o w ID c aDDissnJ 5uo|D psppo Ss d lAI ,0 o r f il >- m (1 CPO) TJ .t E a> '= 0, ^0 1 U- ^^ , w I D 1 h u> NACA Fig. 2 m (0 to -CP ' c <: o o 0' c o n o ID cu ■o ■o T5 O a O CM a o « N o 7-Q o 1 >- k .01 xQOt' OOf OSZ 002 OSI 001 sbcissrij- 6uo|D peppo seD(^ ^:^ .1^ 01 c o =1 .L o +- O "n O Q. gj ■^i" Q. ou 3 riACA Tip. 3 (,.01 X OOf O^ C i- 3 x> ^ m o o X o c o w o en o 1 ~ o o ■0 0) - ro X) o Ul §-£ O VI o a C i- r\i ^ .2 "-^ ZI X. J3 »- o - tt) o i- i. (M 4-^ O 10 O M ■a " ^" n ^ o in •>- ' D £ o l^ o 1 o tl ( jr "D u ii- 1 M- -vt (^ u L +- 1.1) (J) I > ^ (11 o M ( 1 o i^ d) ii. 0) L UNIVERSITY OF FLORIDA 3 1262 08103 280 6 UNIVERSITY OF FLORIDA DOCUMENTS DEPARTMENT 120 MARSTON SCIENCE LIBRARY RO. BOX 117011 GAINESVILLE, FL 32611-7011 USA