KB No. IAI07 NATIONAL ADVISORY COMMITTEE FOR AERONAUTICS WARTIME REPORT ORIGINALLY ISSUED October 19W as Restricted Bulletin IA-IOT SPINNING OF LARGE AIRPLANES By Oscar Seidman Langley Memorial Aeronautical Laboratory Langley Field, Va. NACA WASHINGTON NACA WARTIME REPORTS are reprints of papers originally issued to provide rapid distribution of advance research results to an authorized group requiring them for the war effort. They were pre- viously held under a security status but are now unclassified. Some of these reports were not tech- nically edited. All have been reproduced without change in order to expedite general distribution. L " 96 DOCUMENTS DEPARTMENT Digitized by the Internet Archive 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/spinningoflargeaOOIang 7 (z i> l < 17 NACA RB No. 1I1IO7 NATIONAL ADVISORY COMMITTEE FOR AERONAUTICS RESTRICTED BULLETIN SPINNING OF LARGE AIRPLANES By Oscar Seidman SUMMARY Because large airplanes of the transport and bomber categories have been reported to have spun inadvertently, the available information on the subject has been reviewed. Results of model tests, as well as reports of full-scale- airolane spins, were considered. It is concluded that large airplanes should not be intentionally spun because these aircraft are not designed for the loads and speeds that may be encountered in the spin and recovery. If a large airplane is stalled, either inadvertently or for familiarization purposes, the pilot should apply sufficient down elevator to relieve the stall at the very first sign of stalling. The throttles should be closed if the airplane has started to roll off into a turn and the nose has dropped appreciably. Even after the airplane has rolled off on a wing, the pilot can regain control by promptly moving the stick forward and then using all three controls to return to level flight. For recovery from fully developed inadvertent soins, the rudder and wheel should be moved against the tarn and, about 5- turn later, the control column should be moved forward. In a spin while on instrument flight, the ball bank indicator should not be relied upon to indicate the proper direction in which to move the wheel or rudder, but the rate-of-turn indicator should be used to deter- mine the direction in which to move the rudder and to indicate when the rotation has stopped. The pull-out from the recovery dive should be started promptly to avoid building up excessive speed, but the pilot must be careful not to pull out too rapidly as the airplane might stall again or the structural loads might become excessive. In a spin the pilot would probably encounter difficulty in moving the controls and might have to make use of the tabs and other booster devices; however, he should be careful to avoid overcontrolling after spin recovery. NACA RB No. lJ+107 INTRODUCTION Pilots who fly large airplares - that is, transports and bombers - have normally rial no experience in spins of such airplanes although these pilots will have been "checked out" in stalls. Their spin training has been obtained on small highly maneuverable airplanes. Large airplanes are not intentionally spun, except on rare occasions, for reasons that will be apparent in the dis- cussion to follow. Relatively little information is generally available, therefore, on spin characteristics of the large aircraft. Inasmuch as large airplanes have been inadvertently spun or have been in various stages of spin entry, pilots are naturally interested in knowing what to expect if their airplane should get into a spin. The Safety Bureau of the Civil Aeronautics Board therefore requested that the NACA make such information available and the present report was prepared as a result of this request. The information presented herein is considered of interest to both civil and military pilots. The NACA has obtained a fair amount of data on the spin characteristics from free-spinning tunnel tests of models of large airplanes. The tunnel provides a verti- cally rising air stream in which the airplane model spins entirely unsupported except by the air forces. After the model has been launched in a fully developed spin, observations are made of the effectiveness of the con- trols for recovery when they ere operated by a remote- contrcllsd mechanical pilot.' '".lost o± the discussion of spinning in the present paper is based on results o£ tunnel tests of about a dozen models. A limited amount of actual flight data has been gathered from pilots' reports and from accident investigations. Pertinent data on pilots' spin experiences have been obtained from aircraft manufacturers, airlines, and the military services . Although the present report is primarily intended to cover spin characteristics, a brief discussion of stalling is also given. The discussion of stalling is largely based on the experiences of NACiS test pilots. The entire report has benefited from suggestions made by Mr. Melvin N. Cough, Chief Test Pilot oi' the ft AC A Langley Memorial Aeronautical Laboratory. IIACA RB No. LI4.IO7 3 DESIGN FEATURES OF LARGE AIRPLANES The large airplanes referred to are the present-day convention-;! monoplane transport, bomber, and multiengine- attack type? weighing more than the arbitrarily selected limit of 18,000 pounds. These airplanes are two- or four-engine types. As a result of the installation of ines and other items in the wings, the distribution of mats of these airplanes as a group, as measured by the airplane moments of inertia, is greater along the wings than along the fuselage. The Douglas DC-3 is fairly representative of the class although it has more mass along the fuselage than along the wings. sause of their intended use, all these airplanes are less maneuverable than and are not designed for as high structural strength as the smaller types. The airplanes for which spin- tunnel -model results were analyzed included 10 twin-engine and 2 four-engine designs ranging in weight from 18,000 pounds to 120,000 pounds. Several of tha twin- . Lne airplanes in the "roup were of the relatively more maneuverable combat types. All were conventional in appearance although twin-boom tail arrangements were included. STALL CHARACTERISTICS Stalling The subject of stall characteristics is a much broader subject than spinning md has been covered pre- viously in aeronautical literature. The stall charac- teristics of large airplanes varj widely among different specific designs as ao those of smaller airplanes. The stall precedes the entry into a spin. In the worst case, the stall ray result in a violent rolling motion of which. the pilot receives no advance warning and against which the aileron control is completely ineffective or even detrimental. In most cases the ailerons should not or used. If the control column is net promptly moved forward a sufficient amount to unstall the wing, the wing-dropping ray lead to a spiral, b spin, or a falling leaf. In most stalls the aileron effectiveness will be reduced. In better stalls the roilin? motion may be less violent and advance notice may be given the pilot in the form of mild NACA RD No. LL1.IO7 buffeting or control shake. For some airplanes no rolling motion is involved and the airplane simply pitches nose down after mild buffeting. An airplane that normally stalls gently may show a violent stall under adverse icing conditions. For most airplanes, the wing-dropping will be more violent with power on than with power off. The stall in the landing condition (gear and flaps down) is frequently milder than in the clean condition although the worst case is almost always for the partial-power, partial-flap, approach condition. If the stall charac- teristics are good, the experienced pilot can usually make the airplane recover from the stalled condition before the spin actually gets started. NAG A test pilots have, in fact,, made slowly approached stalls in all types of large aircraft and, although various types of stalls and roll-off s have been encountered, none have been uncon- trollable or have gone beyond the very first stages of spin entry. Spin Entry Inadvertent spins generally result from stalls that have been followed by a violent dropping of one wing. When the wing loses its lift and drops, the nose of the airplane also drops and the airplane slips in the direc- tion of the low wing. This slipping motion will lead to an air force on the vertical tail tending to turn the airplane off course toward the low wing. This initial turning motion, which gives a change in heading, does not constitute a true spin. Inasmuch as the stall and roll-off Is produced solely by the high angle of attack, which is controlled by the elevator, control can still be regained by first unstalling the airplane by use of the elevator end then using rudder and ailerons as available and required to counteract yawing and rolling. If the elevator is moved down more than necessary, the airplane will pick up too much speed. If, however, the pilot fails to check the incipient spin by moving the stick forward promptly, the airplane progressively winds up into a stable spin. The rudder and ailerons will tend to blow with the spin (that is, right pedal forward and wheel to the right in a right spin) and the elevator will tend to blow upward. The number of turns before the airplane gets into a fully developed spin varies with different airplanes; the consensus is that the number of turns is greater than one but less than five. The essential point is that recovery becomes increasingly r.L.107 difficult and requires more turns and altitude loss frort the time of the initial stall until the spin steadies down. Recovery should therefore be started as promptly as possible at the very first indication of the stall. SPIN CHARACTERISTICS OP LARGE AIRPLANES It has been found that present-day large airplanes have, as a group, certain common spin characteristics; (1) The spins generally tend to be steep (airplane nose down more than 1+5° from the horizontal). The air- plane may exhibit some tendency for oscillations or, in extreme cases, for a whipping motion during which the attitude varies. (2) Rates of descent will be high, reaching from 115 to 280 miles per hour (10,100 to 2^,600 feet per minute). Inasmuch as the path of descent is almost vertical, these figures also represent the true airspeed. At an altitude of 10,000 feet, a true airspeed of 280 miles per hour is equivalent to an indicated airspeed of 2J4.O rules per hour. The rate of rotation will be relatively low compared with that for small airplanes. The time for one 'barn will be about 5 seconds for four- engine airplanes and about 2 seconds for twin-engine designs. An average Isrge airplane might, for example, drop 1000 feet at each turn. (3) As a result of the rotation, the airplane will be subjected to an acceleration of l«5g to Jg at the center of gravity. Occupants near the center of gravity will be held down by a force of 1. 5 to 5 times their weight. The acceleration at the tail might be as much as 6g. (Ij.) The flattest spins will be obtained when all three controls are deflected fully with the spin. The most rapid recovery will be obtained by reversing all three controls. Moving the control column forward after the rudder has been reversed (that is, moved against the turn) will be very effective for recovery. Moving the wheel against the spin (that is, to the sarrie side that the rudder is moved) will also speed up recovery. In most cases, the turning will have stopped by the time all three controls have been moved as recommended.. MAC A RB No. lij.107 (5) Spin characteristics for the landing condition are generally similar to those for the clean condition. If a large airplane spins while coining in for a landing, the chance of completing recovery in the height available is slight. Little consistent information is available con- cerning the effects of power (applied symmetrically or asymmetrically) on spins, although it is believed that application of power in a spin may lead to vibration cf the structure. Use of power is therefore not recommended in attempting recovery from spins, except as a last resort, For a number cf reasons, spins of large airplanes are dangerous and should not be intentionally entered: (1) The air load en the airplane in a spin may exceed three times the airplane weight, corresponding to an acceleration of 3g, which is the usual safe structural limit for large airplanes. Oscillations during the spin Lght so increase the load that danger of local failures or deformations in the structure is encountered. (Fighter airplanes, on the other hand, c?n safely take an accel- eration of 8g. ) (2) The effectiveness of the instruments will be impaired. In a spin the artificial horizon may be inoperative, and the ball bank indicator may not indicate the proper direction in which to move the wheel or rudder. The rate-of-turn indicator should still function properly. (5) After the airplane stops spinning, it is in a dive and gains speed rapidly. The pilot must pull the airplane out of the dive before the maximum permissible diving speed is reached. Very skillful piloting is required to avoid either pulling up too rapidly, which would impose severe structural loads or even stall the airplane again, or pulling up too slowly and exceedin the safe diving speed. In any case, a considerable loss in altitude would be experienced before the airplane resumed level flight. (if) All three controls will tend to blew with the spin. Because of the large surfaces and high airspeeds, the controls will be hard to move. The pilot may there- fore have to make use cf trailing-edge tabs or other booster devices to help in obtaining the desired control movements . NACA KB No. Ll+107 7 (5) High centrifugal force would affect the crew physiologically and might make it difficult to move the controls or to reach an escape hatch. This effect would be most pronounced near the tail portion of the airplane. A tail gunner probably would not be able to move about. The small airplane may spin steep or flat. A small airplane rotates faster than a large airplane and has greater rudder effectiveness for recovery. Recovery for small airplanes heavily loaded along the fuselage may be expedited by moving the wheel with the spin. Small airplanes that are heavily loaded along the wings, however, as by multiple wing guns or wing fuel tanks, will have the same elevator and aileron effectiveness as mentioned for large airplanes. Spins of small twin-engine airplanes will resemble those of large airplanes except for the higher rate of rotation of the small airolanes. -e* Considerable information is available on the spin characteristics of the Douglas DC-3 model and airplane. In appendix A, a detailed description of the model spin characteristics is presented and the effects of different loadings are described. It is shown that if a large air- plane happens to be relatively heavily loaded along the fuselage, the favorable effect of moving the wheel against the spin may be lost. The currently available information on pilots' experiences in spins of large airplanes is summarized in. appendix E. These flight experiences are, on the whole, consistent with what would have been expected f r om mode 1 test results. RECOMMENDED PILOTIIJG- PROCEDURE Reference 1 gives in detail general recommendations for piloting procedure for spinning of pursuit airplanes. With a few exceptions, the general principles specified therein also apply to large airplanes. For inadvertent spins of large airplanes, the following recommendations are made ; (1) The pilot should apply sufficient down elevator to relieve the stall (and increase the speed) at the very first indication of stalling. He must be careful not to apply so much down elevator as to increase the airspeed excessively. 8 NACA R3 No. hl+107 (2) If the stall has occurred with power on, the throttles should be closed, when marked rolling has devel- oped and the nose has dropped appreciably. Closing the throttles while the nose is unusually high may result in a whip stall. (3) If the airplane has rolled off but not yet wound up into a stebie spin, the turning motion should be checked by moving the stick forward to unstall the wing and then using all three controls to regain level flight. (i|) After the spin has become fully developed and the controls are with the spin, the most effective con- trol manipulation is to move the rudder against the turn and move the wheel to the same si de as the rudder and, about =- turn Inter, to move the control column forward as far as appears necessary. These positions of the controls should be held until recovery is effected. Once the airplane begins to respond, the forward movement of the control column should be stopped, inasmuch as this movement noses the airplane down and makes the recovery dive steeper so that the subsequent pull-out takes longer. (5) In a spin while on instrument flight, the ball bank indicator should not be relied upon to indicate the proper direction in which to move the wheel or rudder, but the rate-of-turn indicator should be used to deter- mine the direction to move the rudder and to indicate when the rotation has stopped. (6) The dive pull-out should be started as soon as the spin rotation has stopped in ord.jr to a\cid building up too much speed during the dive. The pilot should not pull out too rapidly as the airplane might stall again or the structural loads might become excessive. (7) The tabs or other booster devices should be used as much as necessary to obtain the desired movements of the control surfaces. The pilot should be prepared to readjust the trbs upon recovery to avoid overccntrolling in the ensuing dive . Although spinning of large airplanes lias been suc- cessfully accomplished in several inatjnee?, the evidence points strongly against this practice. Even though the spins may resemble those of seme smaller airplanes, the NAG A R3 No. LL1O7 permissible overloads and diving speeds are lower and the controls are much harder to move. Large airplanes are not designed for aorobetics end should not be inten- tionally spun. L. angle y Memorial Aeronautical Laboratory National Advisory Committee for Aeronautics Langley Field, Va. 10 NAG A RB No. LI4.IO7 APPENDIX A SPIN CHARACTERISTICS 0? MODEL 0? THE DOUGLAS D0-3 Model spin characteristics of the Douglas DC-3 air- piano were obtained from tests of a 1^-f oot-span model in the NACA 20-foot free-spinning tunnel. The specific results in terras of equivalent full-scale data are described in some detail for illustrative purposes. For the fully developed spin with the elevator up, rudder with the spin, and ailerons neutral, the nose would be -'.own 55° from the horizontal; the rate of descent at an altitude of 10,000 feet would be 117 miles per hour (10,300 feet r<-3v minute) and the rate of rota- tion would be y .h. seconds for one turn. The acceleration at the center of gravity would be l«7g« Complete reversal of the rudder alone would give a recovery in 1 turn, after which the airplane would descend in a steep glide. Figure 1 shows the airplane motion during the last turn of the spin and during the recovery. After recovering from such a spin, the airplane would be in a dive at 173 miles per hour true airspeed (152 miles per hour indicated airspeed at 85 00 feet). The pilot" then world have the alternative of pulling out sharply with resultant high accelerations or pulling out gradually with consid- erable increase in speed and loss of altitude. If he increased the acceleration to 2g In 2 seconds and held this value during the rest of the pull-cut, the airplane would drop 2000 feet during the pull-out to level flight and the speed would have increased to 285 miles per hour true airspeed or 263 miles per hour indicated airspeed. This value 01 the speed is close to the maximum per- missible diving speed for the DG-3 airplane. If the pilot had wanted to use the elevator for recovery, it is estimated that he would have had to push l60 pounds on the control column to start moving it forward. If the '"*lob managed to get the control column to neutral before reversing the rudder, the spin would be a little flatter and the recovery dive would be steeper than if the control column remained back. For the model tests in the normal loading condition, the wheel position did not seriously affect recovery. For this loading condition, the model did not show ; NACA RB No. lij.107 11 usual favorable effect of moving the wheel against the spin because, as mentioned earlier, the DC-3 has a rela- tively heavy load along the fuselage. Tests of the model in the lightly loaded condition, which the load distribution was more nearly like that 12 NACA RB No. LI4.TO7 APPENDIX B FLIGHT TEST RESULTS Little information is available on intentional spins of large airplanes. Information available on inadvertent spins is of questionable accuracy because of the confu- sion of pilot and crew, the lack of prepared instrumen- tation, and the feet that the pilot is concentrating on trying to recover from the spin. This uncertainty in the information should be borne in mind in evaluating the following specific information on full-scale spin experiences . Doug l a s LC - e r i rp lane . - The i'o 1 1 dw i n g ins t sne e 3 h a ve been reported concerning spin experiences in the DC-J airplane (twin engine, 25*55^ lb): (1) A chief pilot for an airline company performed intentional spins with the DC-3 airplane several years age. The following results were obtained: Three spina were made with wheels up and one with wheels down. All spins were entered at an altitude of 80GO feet. For these tests the airplane weight was only 22,000 pounds. One spin of 2 turns was made with each engine operating at I4.5O horsepower. There was no effect of power or of lancing gear. The longest spin lasted 5 turns. All the time the airplane was spinning, considerable force was necessary to hold the ailerons in the neutral position there was a very marked buffeting of the tail sur- faces. The nose was well down, not being more than I5 from the vertical. No trouble was experienced in cringing the airplane out of the spin; it was necessary only to neutralise the controls after vhich the spin stopped in lees than ~ tern. The maximum indicated airspeed noticed during the spin was ISO miles per hour. a recovery the sirplane attained an indicated airspeed of approximately 2J0 miles oc,r hour. In making three turns, the airplane lost approximately JiOOO feet of alti- tude from the time that the spin was entered until recovery was completed and the airplane was in level flight. (2) Other instances have been reported where diffi- culty was encountered. In one Instance the spin was NACA R3 No. LI4.IO7 13 entered accidentally with wheels and flaps down and with partial power. The flaps were retracted and the power reduced. An attempt to stop the spin with the rudder brought no results. Full power was applied to the inboard engine with no effect. The rudder was then neutralized and the control column pushed forward with considerable force at which time the spin r.toppea. (5) Several pilots have reported going into 2-turn spins in bad weather or during training maneuvers. The pilots indicate that the ailerons whip toward the direc- tion of spin as the airplane enters the spin. Recovery was generally successfully accomplished by neutralizing or reversing all the controls. A 3-turn spin has been reported during which the nose was I)-5 down. The loss in altitude during a 1-turn spin and pull-out from the ensuing dive has been reported as 3000 feet. YFivi - 1 airp 1 a ne . - The YFM-1 airplane (twin engine, 18,150 lb) entered a spin inadvertently from an asymmetric-power flight condition. The rudder blew with the spin and the pilot could not push hard enough to move the pedal. The spin was steep. When the co-pilot jumped, he struck and bent the leading edge of the fin and also struck the rudder. At about this time, the pilot found that he was able to move the rudder. The pilot then applied opposite rudder and followed by moving the stick forward and giving opposite aileron, which brought the airplane out of the spin. This spin lasted 19 turns. i -2 6 ai r p 1 an e ♦ - A service pilot practicing evasive action stalled a 3-26 airplane (twin engine, 26,6^0 lb) and spun very steeply. He applied controls with the spin for one turn, then gave full opposite rudder, and after one more turn moved the stick forward. When this manipulation had no effect for two turns, he repeated the entire series of control movements; then after two more turns the airplane recovered in a vertical dive. This spin lasted about 7 turns. The co-pilot had closed the throttles after the first turn. The controls were very difficult to move. P-7J airpla ne.- Several P-70 airplanes (twin engine, 21,2ip5 lb) have been lost in spins. Details are lacking but it is suspected that high stick forces may have been a contributing factor. ill NACA R3 No. II1IO7 B-17 £J^d 3-2b_ airplanes .- Two four-engine designs, the 3-17 (S2,330 lb) and the 3-Zl. (50,000 lb) have been reported in spins several times. In some instances serious structural damage and loss of the airplanes resulted. In one esse, control forces were reported to be high but the combined efforts of the pilot end co- pilot finally moved the elevator and rudder controls and effected recovery. The spin was steep. A crew member near the middle of the fuselage was able to move about but the tell gunner was unable to move because of cen- trifuge! force. Boeing 307 airplane .- It is thought that the breaking up of an experimental Eoeing 307 airplane (four engine, 142,500 lb) in flight might have occurred during recovery from a dive subsequent bo 8 2- or 3-turn inadvertent spin. P-3 : 3 airplan e.- The P-38 airplane (11,300 lb), which is a small twin-engine design ard is similar to some of the lerge types, has been spun several times. The test pilot reported that on one occasion he was unable to move any of the three controls from their with-the-spin posi- tion after a spin of 3 turns. Ke regained control after eight turns by applying power to both engines. It would be appreciated if pilots having additional information on actual spin experiences in large airplanes would transmit pertinent data to the National Advisory Committee for Aeronautics at Washington, D. C. ERENCE Soule', h. A., and Seidman, Oscar: Influence of Loading Condition on Pilotin ! nique for Spin Recovery for Pursuit Airplanes. NACA RB, June l°/lj2. NACA RB No. L4I07 Fig urr spin \ Altitude 10.000 ft j k 1 $ % * ft p. ^$ $, § !*« c?>ft t ^ U5 ski pi i^S ^ ft RL'DDER REVERSAL IfULL WITH TO FULL AGAINST THE SPIN) \ ALTITUDE 9409 FT- I \ m ? o ft s: o - **: si' Vfi 1 Rotation Ceases Altitude am ft -j 5 rs. Uj ? 1 ft FIGURE I- STEADY SPIN AND RECOVERY OF DC- 3. UNIVERSITY OF FLORIDA 3 1262 08104 985 9 UNIVERSE Q GAINESVILLE, FL 326 "'7011USA