ARR No. 3F2U NATIONAL ADVISORY COMMITTEE FOR AERONAUTICS WARTIME REPORT ORIGINALLY ISSUED June 19'<-3 as Advance Restricted Report 3F2U VQMD-TONKEL VIBRATICN TESTS OF A FOUR-BLADE SUJGLE-ROTATIKG PUSHER PROPEIiLER By Mason F. Miller Langley Memorial Aeronautical Laboratory Langley Field, Va. UNiVERSirt'OFFLOfilDA DOCUMENTS DEPARTMENT 1 20 MARSTON SCENCE UBRARV P.O. BOX 117011 GAINESV/LLE.FL 32611 -701 T USA 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 - 327 ' |^V o7l ^^\ iTtOO^^Sf to I 'ATIOI-AL AJYISCiiY CO.,..;ITTi:^ i^GR A^ROl'AiJT ICS ADyA''Ci:i !<::. S T,R I C TiJ] D RxiPO^T 'irD-TU'"i;iiL YIERAC'IOK Ti:.STS GP A FOUil-BLAD:: SIxiGL^-i.OTATIFG FUSKj^R PK0P.;.LL^R By i-'ason F . :iiller su:;maky pell perf "b e b i f or iHOl'ft aher. f our t w i c w i n f,- vilar sr)ee er or ::ie nd fl T eac b 1 a — spe d of ths e th at at 01' ds , rat per d b v.'i t i des p d th the e ir the y s r.nd tr n tests of a f ou r-t 1 :^ de s iiifle— r ot at ing pr o— ing iri e. s i!n^^.lf. t ed pixsher condition were r,\ise the ccnil in?, t i Oil of v-sk? °nd dov/nvash vas expected to provide serious excitation ess vibrations of propellers with four or The tests lere conducte(i in the L ,AL 16— foot nnel v.rith s vin,? inourited at thrust— axis level propeller; the blpde sections at three— ropeller r:;dius oper-T.ed at appr oxisaat e ly hords behind the tr.i^ilinr edge of the tspered closest popition, iieasur eiaent s of propeller ess ''ere r.'iade for various airspeeds, engine n,':ine power s . The for an e at a f r e vibrator v;.as prac pre s sure upon the the angl no det ec increase fl5:p on the vibr over t h f^ per hour wake behind d;ewise reac c^-iency of t^' y stress i n c t i c 3 1 1 y i n d e f or const gn r e p c t i n 1 f" 2 e of at tc! ci- table in ere a of wpke ej. c t b. e lower s u atory stress stress perr the w t ionle ice t h r e ■ . s e d peade n t airs s vibr of the se of it at i r f a c e 3 n d p ^ t ai ing supxjlie s s V i b r a t i c e propeller c Oiis i der ab t of en£:ine peeds . I he a t i n - w a s v win;,: f r lii do'-'nvash ex n , As iuml of th° I'/iniT rohibited t r s p e d s h i r d se n of spe ly V br,^. ef f er V 0^ t c i t a a t e d gre he r >ier nous exci tne prope ed ; t he re ith air spe ke 'uPan ef ect of dj''; s ma 11 ; cha 3.9° pr t i n end 1 fill 1— span ^, 1 1 y i n c r e unning of than 140 m ■ tat i on Her suiting ed but feet i ve nwash nging duced ittle split ased tests i les 'v f 1 r, t v; '. s e r e a c t i li 1 e s s vibration was d e 1 1-'- c t e d , probably because the airspeeds were low for most of the critical engi/iP speeds and because the harmonic coaponents of wake e^■clt.■ tion i"ere sinalle INTHODUCTION The operation of a single— rotating propellor with four or more blades "bohinl the v/ing has create i ? ohq concern "because of the expectation that the combination of wake and downwash night supply serious excitation for a reactionless typ3 of vibration. Because rcactlonloss vibrations of a single—rotating propeller raay occur at all frequencies other than 1 , kB , and kB tl tiaes the propeller speed where k is any integer and B is the nu^ibor of blades, it is observed that the propeller must have laoro than three blades to vibrato in a reactionless manner (reference 1), A propeller vibration is reaction- less if the vibratory cotions of the blades are such that the vibratory bending sjononts and the vibratory forces of the several blades cancel each other at the propeller shaft; consequently, reactionless vibrations occur only with aerodynaEic c;xcita*"ion and are not possessed of engine damping. It was believed that with no engine dauping the vibratory stresses caused by the v.'ako and the downwash behind a wing could be unsatisfactorily high, inasnuch as the vibratory strcssos would be lit-ited only by aerodynamic damping, by hysteresis damping of the propeller blades, and by damping produced by motion of the blade shanks in their hub sockets. Tests were conducted in the LMAL 16— foot high— speoi tunnel with a wing mounted at thrust— axis level ahead of a four— blade singio— rotat ing propeller. Measurements of propeller vibratory stress were made for variotis airspeeds^ engine powers, and conditions of the wing for the complete engine— speed range. Most of the testing was done, however, within the limited range of engine speed for which a prom- inent reactionless vibration occurred. Members of the staff of Hamilton Standard Propellers, Division of United Aircraft Corporation, collaborated in conducting the tests and analyzing the records. APPARATUS A1:D MT^rHOHS The s ingle— rot at ing propel lor tested is described as follows : Type HaTiiltori Standrrd hyclr onr t ic r.'aterial ->- 1 u a i lui iii alloy I'hi.n'ber of "blades i'' our Diameter 1j feet C iuches Blf.de design 6437-13 Hut) design 24D50 The propeller was driven by a Pratt Sc ./hitney 5.— 2800 engine geared 16:9 and niounted on ruhber mounts in a full- scale stut— wing nacelle. The engine— pr opeller—na ce lie comtination is shov/n in fi gur e 1 . The nu s li •= r conditio;! v,.'ing at thrust— axis level ler . The I-; inr:, v.'hich has t i n , y as i nverted merely v,'ay of aounting having "bei vi"br a t i n t ests reported the v; i ng I;; oun t e d a lie :> d of push e r c n d i t i r . T h e d i 1 c '■:. tion of thp v'ini" with v; i n f; set at an an-:le of a ur e 3. The V'ing '-'as loca sect ions at three-f our t hs ,1 1 a ppr ori;'; -tely tv.'ice th e d;;e of the t '•! p e r e d v i n g . for one of the tests is f t h^ s i iiu 1 a t ed flat) would the U' ins, i t v- s s s r. e w h a t t i on , was Firr.ulate aaead of the an J i-CA lev/— heca.use of c en desira'cle in reference the pr ope lie tensions of t r e G T) e c t to t 1 1 a c k a of ted in such a the pr ove lie eir chords he They s i;aulat hov/n in f igur "be attached t f or war d of t d "by mounti f ur -blade drag airfoi onvenie nee , 1 or c onduct 2 . 3' i gxir e r t s iJiul a he wi.'ig and he pr ope lie 0''-' are show way that t r radius op hind t hp tr ed split fl e 2 , In or the rear he usual fl ng a pr opel— 1 se c— this in.^ the 1 shows t e t he t he r for the n in f i g— he blade e r a t e d ailing ap used der that spar of ;p posi— 0<5 c ill ogr aph records of propeller vibratory strain were obtained by a method fully describpd in reference 2. Electrical strain gages i-ere hiount^d 1 on^^i tudinally on all the blades r-i t th° shanks and nepr th" tips. The gages were mounted on t hp cambered — that is, the front — sides of the blades (fig. "). iSecause the xaximum stress of a blade surface for a given propeller radius is at raaximuiu blade thickness for a flatwise vibration, the tip gages were aounted to laeasure stresses at .L^ixiSiUhi blade thicknesses. S orae .':ae.?s viere ir.ounted. on the wine:. The strain gages on the propeller were connected to a slip— ring device, vrhich in turn vras connected to volt — a.^e amplifiers. The strain— gage resistances varied v/ith the strains to produce fluctuating voltages the alternat- ing components of which were applied to the aniplifiers. The gages were calibrated in such a v;ay that there was a known relationship hetween the alternating voltages and the strain variations. The v/ere app osclllog t ographi lo graph ing a t i v;ave con a given ohtained an osail weve rep oies of peller s c p n he d would ha alternsti lied to OS raph, and c pppe V e lement s ; ming wave sisted in cylinder f hy proper lograph el :-r e s e n t i n g vibration peed, in o eterinined.. V e h e e n j v. u?— /olt age output cillograph elenen strain variations There were 12 amp one of the ch'^.nne on the photograph periodic impulses ired- These i mpu connection from esient.- 'J.'he purpo engine speed is t in terms of eithe rder that the cau A direct record s t as sui I able . s of the siTiplifiers ts of a recording were recorded on pho— lifiers and 13 oscil- Is was ured for record- ic paper. The -liming occurring each time Ise records v/ere a spark-plug lead to se of having a timiug express the frequen— r engine speed or pro- se of the vibration of propeller speed (See reference 2„) All the amplifiers v/ere calicrated s iim_.ltane ously at intervals during the test oy applying a known alter- nating voltage to their input terminals. The amplitudes of the resulting oscillograph traces v/ere measured after the tests, and a definite relationship between oscillo- graph amplitudes and amplifier input voltag-es was thereby obtained 3ach inch of amplitude on the photographic paper therefore represented a known amplitude of strain on a propeller blade. Stress values v/ere determined by multiplying strain values by the modulus of elasticity for aluminum. The static natural frequencies of reactionless pro- peller vibrations with Hamilton Standard 6487-12 blades v/ere predetermined by iiieas ur eir.ent and are shown in the following table; I ' ode -" r t cuency ! (eve) 1 r 16 .7 ! 70, .2 i' 1 a t w i 6 e ! < 13 7, . 3 ) 187, ,5 .230. ,5 id£;ewise 45, .8 , j uar an old re -I vi"b the ^•r a t i rain el'- str m i li due tun rnod th- vit qu'- hor mc a ."" i V f our -t lad'-' prop'11-r, r^stini-: on its hu"b which v/ns ■-strainf'd, was •-■xcit'-^d r- 1 oi\r of th--^ l)lade tios with "Ipctrical eycitr^r of varir'bl.'' fr^qu'^ncy. It wr.s con- '-red n.ot n''- ces siivy to restrain thP hut "b ■■' c an s ■" , for a ctionlpss vitratioi), the vibratory "D ■ ' n. d i n o--^ :uom"ntc and ratory forc'S of th four 1 1 ,■-: d <■■ c? canc-^l finch oth'T at huh. Froni a low frf^qu-acy, th*-^ pxciter frnqu'-ncy w^'.g dually incrppsed; wh'-^n th-"^ vnrious r.-- •■ ct ionlps s vihr.r,— ns appp^r'-'d, th'-' f r pqur nc i-'^ s w:r<- accv.ratf^ly det^r — d. Osc illogr -jph r^^cords of strain wrre trikf-^n, usinf: ctrical strain gagf'S for pickups; th- f roqu'^nc ics of a in variation app'^aring on thrsr' r--cords v;pro d'-'t^'r — C'd. (iccurat"ly hy comparing thf>-a v:ith the traces pr o— -d hy an accurately ccxlihr at^d , electrically excited ing fork. ifrequenciOG for the first five flatv/ise es and the first edgewise luodo v/ere determined, Because present r':^port derils larincipally with reactionlcss rations, th'-^ m^^thod of do t r-rniining static na turpi fr*^ — ncies of nonr eact i onle s s vilrations is not discussed f^in. A oior.-^ co.'iplcte discussion of t)'!'"^ iviethods of suring static natural frequenci-B of a propnllor is en in reference 2 , A prop^ll.^r vihration ic terrfi.-^o flatwise if the vibratory motions of thr tlade s^^ctions r.rc primarily perpi-^ndicular to th'-^ blade chords; v/hereas an edgewise propeller vibration is one with the vibratory li-iotions of the blade sections prinarily along the blade chords. Some flatwise motion generally exists near the blade tips dur- ing edgewise resonance because of the coupling supplied by the blade twist. Cpntrifugpl correction factors were applied to the static natural frequencies of reactionless vibration in accordaxice \','ith th-^ accepted formula, which ie -xplained in r ef ere ric" -''■ , f' --^ f + f-n' (1) where f natural frequency at s given prop<-^ller s-'^epd fo static natural frequency n pro-p=ll-r speed K a constant for a given laode of a given propeller Th^- r.ethod of predicting engine speeds f or . react ion- less vitrations of frequencies 2n and 6n is shown in fifi'ure 4. Tiio values of I^ used for figure 4 are as f ol— 1 ows ; First fletwise inode Second flatwise oodp i'irst edgewise mode 1.7 5.6 1.12 The critical engjin-^ speeds are those at v/hich the straight lines intersect the lines representing natural frequencies, Only the first and th? second flatv;ise modes and the first edgewise raode are considered for figure 4 hecause, v;ithin the engine operating speeds, the strais^^ht lines represent— irij.; frequencies of 2n and Gn do not intersect the nat ur al— frequency lines for the higher laodes. The reac— tionless vibration having a fr.^-quency jf 2n is of aost importance and the r eac t i or-l:-6 s vibration of frequency 6n is also of interest. 'Excitations having frequencies higher than 6n v.'ere expected t o he negligible. The tpst conditions are given in the follov.-ing table: -. ngme oiSf^p (ib/sq in. ) 100 15 2C0 Airspeed (mph) ingine speed ( r p m ) ngle of attach, Irt to 2c0 ■- I'.'G to 2^0 90 to 2G5 i ' 1250 to 28;:->0 I \ 1350 to 2860 i 2400 to 2 85 jAngle of attack, C*^; simulated split flap on wing; 1 00 150 ; 100 to 1-5 2400 to 2 35 aSl^-^RAL j:i I so US SI OK OF PP.OF^LL^R VIBHATIOIvS CAUSJiD BY A fima AIIxiAD Oi^ TES paOPiiLLiiH Sizable excitation forces for reacticnless propel- ler vibrationp are expected if the propeller operates in the wake and dov;nvr?.6h region "behind a ving. Although nonreact i onless vibrations having other excitations are also important, they are outside the scope of the present report . In accordance with the result of an snalysis show- ing that re act ion less vibrs'cions can occur for all fre- quencies other than 1, kB , and kB • + 1 tiraes the pro- peller speed, a propeller aust have more than three blades to vibrate in a reactionless manner, and a reactionless vibration of a four— blade propeller can occur for frei^uen- cies of 2n, 6n, 10 n, 14n . . . (reference l). It may be not<=d tha.t a reactionless vibration can occur at a fre- quency of 3n for propellers with four or raore blades. The wak'-' behind a tionless vibration of a blades if the freqtiency a natural frequency for blade of the propeller regions per revolution, ■".'elocity with respect t change of angle of atta therefore act upon the vibration at a. freo^uenc of forvtird velocity act shov;n in figiire 5. The lov/er forward velocity attack. The decrease i V, , hov;ever , slightly of the angle of attack, st at ic— pressure variati in f i g ur e 6 . v/ing Lipy result in a serious reac- propeller with four or more of excitation 2n is equal to a reactionless vibration. Sach passes through two low— velocity The periodic change of forward each blade causes a periodic ck of each blade. Periodic forces blades to produce a propeller y of Sn. Tiie effect of r- change ing upon the propeller blades is greater force occurs for the because of the greater angle of n magnitude of resultant v&locity offsets the effect of the change Typical t ot al— pr os r.ure and ons in thp wake region are shown The excitation provided by the wake is not sinusoidal and "xcitaticn'^ ?t frequencies thrt are haraonics of 2n therpfore erist. These ha.ruicnic components, hovrcver , are sinall'^r thpn the f undac'iental coaponeut. The excitation having a frecuency o^" 4n vill not f!-:cite a roao 1 1 c.iies s vibration of a four— blade propeller. Although an excita- tion having a frequency of 6n can product s reactionless vinrr.tion, the third h'-.rnonic component of wake excitation i? evpectrd to bP uiiit-:' e ;:u 1 1 and to give little troutlo. Higher harmonics of vakP excitation ere expected to "be negligible . A reactionless vibration h^^ving a frequency of 5n can be n first, a second, or a higher inoae , depending upon th-"^ propeller s.pe'Jd. The hi'-^-hest mode that can be obtained v/ith this frequenc" of 6n depends upon the upper limit of propeller speed, the natural frequencies of the modes, and the increase of natural frequencies of the codes with propeller speed. (See refer ices 2 and 3.) For modes of vibration higher than tho first mode, th^ vibratory V'loc— ity cf some parts of t h« blade is 18G° out of phase v;ith that at other parts of the blade (see fig. 7); and, v.'ith thu excitation p.cting in the same sence over the entire blnue leni?jth, some parts of t h^^ blade absorb energy from the excitation, v;hile the remaining parts dissipate energy. If a renctionless vibration of frequency 6n appeared at a relatively high propeller speed, it would be one of the higher modes and therefore subject to thr- cancelation effect. The cancelation effect is somewhat decreased, however, because an excitation acting near a blade tip is moro effective than the same excitation acting near the blade shank, I'he pr^s^r.c of dovnwash b'^hind a ving is .^xo^ct^d to supply excitation for ? prop'-llor vi:ration at a fr'— qu'-'ncy of In, ^s shov^n in figur-" 8. Th'~ downv;ard compo- nent of velocity in th.-; plan of the propell-'^r disk in- creases the anglT of attack of a proprll:-r blade during on'''— half revolution of the propeller and r"'^creasc-s this angle during the remaining on.— half revolution. This p-riodic change of angl.-' of attack of the blades caus^'S p-rioaic forces to act on th- blades and th-rffore results in a propeller vibration at a frequency of In. Also, the r-sultant v.^locity V- of th*^ air with respect to the blades is variabl-^ with the sam--* frequency as th". angle of attack and aids the loiriodic change of angle of attack to produce th^ vibration. Th. vibration at a freqnevcy of In i-xcitcd by the down'-'- oh is not reactionless but, if it occurs s imult ane oxisly with the r .^ac ti onl^s s vibra- tion, is nxp-ct"d to increase th'- e ' r i ovisn'- s ^ of thi^ reac— tionl^ss vibration --xcitrd by th:^ wak-. Incr'-asing th' angle of attack of a wing caus' s an incr'-ase of downwash angl'^ and, as a r-'sult, a vibration -xcited by downvrssh would b ■•^ expected to b-comr- ■ior'- pro- noimced. Although thj= wake Dohind a v/ing follows the dov.'nwash, the chans^e of magnitude and shspe of p woke profile is siQ-.ll for a change of angle of attack less than ahout 5^. (See. reference 4,) A rplatively suifll ^ chanf:e of win,:- angle of attack is therefore expected to CQ produce little chcn^e of a propeller vihrstion excited 'l ty the wake . 1^ The us° of a split flap on a xving is espected to hroaden and strengthen the wake (reference 4) and thereliy c ons i derahly increase the propeller vibratory stress occurring at a frequency of 2n. A split flap on the l0'.i;er surface of a wing also directs the air dov/nward "be- hind the wing and is expected to increase the excitation at a frequency of In, For constant airspeed and propell^'r speed, the pro- peller vibrations excited hy the vmke aid the downvrash "behind a v.'ing would he P::pected to he less affected hy the engine hrake ia"an effective pressure than those pro- peller vitrations excited "by the engine. The trailing edge of a wing may possitly vi"brate hecause of aerodynamic excitation supplied oy a propeller operating close "behind it, if its natural frequency is equal to the frequenc?/ of excitation (reference 2). The freo^uency of importance is Bn; each blade passes the closer pa.rt jf the trailing '^dge once per propeller revo— lut ion. Discussion or I13SULTS The results of the present t^st are presented in fig- ures 9 to 12. Thp stress peaks are labeled with vibration frequencies in terms of propeller speed n and engine speed 17 — for evanplf", 2n and 'iJ-U. The stress peaks of the curves representing total vi'bratory stress have more than one freo.uency component, and the frequency com- ponents are given in order of impor taic« . Some of the stress curves are giv^n only for g frp-quency of 2n; these stresses v;ere meaSTired v/ith a v;ave analyzer. A prominent propeller vibration having a frequency of 2n appeared at an engine speed between 2780 rpm and 2340 rpm. ' (See fig. 9.) This vibration was evidently edgewise, inasmuch as the first mode of edgewise vibration at a freo.uency of 2n was predicted for an engine speed of 2380 rpm (fig. 4). The engine— speed prediction was 10 somewhat high 'but is considered ^ood, any prediction with- in 50 rpm "being satisfactory for test purposes. The curves of figure 9 are plotted for th-^: leading— ed?;e posi- tion of the shank for tv/o reasons: (l) i' or an edgewise vibration, the stresses at the shank are maxiinum at the leading edge and 180° around the shank from the leading edge, as discussed in reference 2 (the leading— edge posi- tion is in line with the leading edge at approvimn te ly the 42— in. station of the blade); and (2) jTor a first mode of vibration, the maximum stress along a blade is near the propeller hub because stress depends upon c/p , where c is the perpendicular distance from the neutral axis to the extreme fiber and p is the radius of curvature of the neutral axis. The effect of airspeed upon the vibratory stress for the edgewise vibration appearing at an engine speed of 2890 rpm is shown in figure 9. These curves demonstrate that the vibratory stress increased v;ith airspeed. Part of the total vibratory stress was produced by engine exci- tation, as evidenced by the frequencies 4— il. III, and l-'LiT ^ ■^ IS-' ' The effect of engine brake mean effective pressure upon the vibratory stress is shown in figure 9. The vi- bratory strpssps having frequencies of 2n change only slightly v/ith brake mean effective pressure for a given airspeed; the very slight variation can be due to experi- mental error. The vibratory stress of frequency 2n woiild be expected to be practically independent of brake raean effective pressure, as previously diecussed; however, if the total vibratory stress is composed of some engine- excited components, some engine damping exists that may vary with brake mean effective pressure to produce such slight variations of stress as found in figure 9. At first glance, the bottom curves seem to vsry considerably with engine brake mean effective pressure, but it must be noticed thet the curves are plotted for slightly differ- ent airspeeds. The do'.vnwnsh behind the ving should provide excita- tion at a frequency of la. .'ith a v/in.^ angle of attack of , traces of vibrations aaving a frequency of In were found for ©n airspeed of 3PC miles per hour, but no indication of the frequency In appeared at the lower airspeeds. (See fig. 9.) The effect of dcwn'-zash upon the total vibratory stress at the engine speed of 28L0 rpm is small, probably because a frequencj' of In at this J 11 "'ngin-' ep^"?d is not s. natural frequency of proppllf^r vi"bration and becansp thf dovnwash. is limit'^d in a con— strict'd air strean of 16— foot diamr-tor. Chanfin.j th--= angl'^ of attack of th" vin'i; from 0° to 3.9° produced littl(' increase of prop- Her vibratory stress. (Se- fig. 10.) This rpsult shows that thp walc^ and th'' dov;nwash bf^hind tho v;ing wore affected little by the angle change. Because the pres'^nc" of the tunnc-l wall is b°li"ved to have limited the dov^nwash, the loca- tion of th'= v;ake would be pxpf-cted to change only slightly v;hen t h-^ wing angle of attack is incr^as^-d 3.9°. In accordance v;ith reference 4, a change of 3.9° in th> wing angle of attack should produce very little change of the magnitude and the shapp of a wake profile. The s i inula tr' ■fi split flap on the lo'^f^r surfac'^ of the wing grestly increased the propoll^-r vibratory stress at a frequency of Sn. (See fig. 11.) This increase of stress is attributed to the strengthening and th'^ broad- ening of th-> v.-ake. Although the presence of the simulated split flap v;as also eYp,-->cted to cause a vibration of fre- quency In, no siich vibration was detected. ''.'ith the us'= of the sirnulat'--d split flap, hov/ev^r , coiaplete re- sponse curv^-e vjerr-: not obtain^^^d for airspeeds higher than 140 miles per hour, b'''cat\se of the dangerously high stresses anticipated, Figure IS is a, stress curve covering the entire engine— speed range. Although a flatwise vibration hav- ing a frequency of 2n was predicted for an engine speed of 1150 rpai (fig. 4), no such vibration was detected, probably because the airspeed at the critical engine speed was only 21 miles per hour. In practice, the velocity of the air with resp"'Ct to either the wing or pusher propel- ler vrould be lov/ for an engine speed of 1150 rpm. During the test th""- r.-sultb of which are presented in figure 12, the c''2Pnt bonding the gages to the blades softened some — v;hat . The actual magnitudes of the str':'sses are there- fore approximate, but the frequencies are correct and provide a reliable indication that no vibration having a frequency of 2n was present. Heither a flatwise nor an edgewise vibration having a frequency of 6n was detected (fig. 12). Figure 4 shows that the critical engine speeds for vibrations hav- ing a frequency of 6n are SOO rpm, 830 rpm, and 1350 rpm. Of these cases, only the second flatwise vibration 12 occurring at f.n engine sp'^ed of 1750 rpn v^ould be exp^ctod, because the first critical speod if b'^lov; the op''^rating range and bocauso t h'"-' airspeed is lov/ for thf s-^cond crit- ical speed. The fact that no vibration having- a frequency of 6n was detected near th(^ blado tips for an airspeed of about 110 milns per hour and an Pnginf' spprd of 13b0 rpm indicat'':s that tho third harmonic componi~-nt of wake ■^ X c i t n t i n was s !.■ nil. Th>"; stress curve e of fig-ures 9, 10, and 11 shov; that the reactionless vibration of frequency 3n excited by the wake is serious. The vibratory stressoa considerably exceeded +3500 poiinds per square inch for the shanks. Ina.sir.uch as propellers ivith more than four blades are a.lso subject to reactionless vibrations at a frequency of 2n , the wake is expected to provide seriou? excitation for edftewiae react 1 onlee E vibrations of propf^llers vith four or 'zov^ b lades , ■To vibration of the wing v/as det'ct^d that could be attributed to asr od5''namic excitation provided by the pr o— ■poller. COrCLUo lOrlS The results of vibration tests v/ith a v/inf- mounted at thrust— axis level ahead of Vj four— blade s ingle— r otat ing propeller to simulate a pusher condition in a constricted air stream of 16— foot die meter indicate the follov/ing c one lus i ons ; 1. The wake behind the v/ing supplied serious exci- tation at a frequency of tv;ice the propeller speed for an edgov/ise reactionless vibration of the four— blade pro— P'-'ller. 2. The vibrat;ry stress for the reactionless vibra- tion increased considerably with airspeed, but was prac- tically independent of engine brake m^^an effective pres- sure for constant airspeeds. 3. The effect of downwash upon the serious reaction- less vibration was very small. Changing the angle of attack of the wing from 0° to 3.9^ produced no detectable increase of downwash excitation and little increase of wake excitation. 13 4. A simulated full— span split flap attached to the v/ing greatly incrnaKed the excitation and prhobit"d the running of tests ovor the stress peak at airspeeds higher than 140 miles per hour. 5. Agreernerit "between the predicted and the jneasured value of engine s^ieed for the reactionless vi'bration was sat isf ac t ory. 6, 1^0 flatv;ise reactionless -'■ihration v/as detected, probably because the airspeeds were low for most of the critical engine speeds and because the harmonic compo- nents of wake excitation v:ere small. LaxiiPiley Memorial Aeronautical Laboratory, National Advisory Comn'.ittee for Aeronautics, Langley ilTield, Va . . ili;?I:ifiii:T:JCiiS 1. Kearns, Charles M . : Engine— Air sere v/ Vibrations. Aircraft ^-jngineer in-- , vol. XIII, no. 150, Aug. 1941, pp. 211-215. 2. Miller, liason F.: v/ind— Tunnel Vibration Tests of Dual-Hot at in.; Proaellers. i'ACA AH-i ?Cg, 31 ' 1 , 19 4."^. - * 3. Theodorsen, T.: Propeller Vibrations and the Effect of the Centrifugal Force. l'-t\'^.i^. 'i S N' .. . olG, 1 9 3b. 4. Silverstein, Abe, Katzoff , 3., anr" Bullivant, 'i . Kenneth: Dov/nwa,sh and ''Jake behind Plain and Flapped Airfoils. !I.a&>.Rcp, No. 6bl, 1939. NACA Ffg.l NACA /o'-»4 Simulated splii flap (used for one iesi on/yw Figs. 2,3 1 J Section A-A Side view riaure 2.- Simulated pusher installation with stmulaUd split flap attache^ to vving. I. J) "55 a. c 1^ (0 u -J Ir-^ , 1 I. »^ )f^ ^V- c (L r ^^ Ws^ ■f. lUOA ns. 4 s^l^-lo/:o '/t '? ■■<. Fi; ,V., .-^''^. .>-•' / / -j Ari{;'le of. atteciv Rotational speed Forv/ard velocity ?or .vard vel oci ty Respective re.istiltarit vej ple!ierit relative to air .ree s^rea::: vva'-e of air m iree of i:ir i'l wa/;e stream of a bla^ jiifiiire o.- Efiecc of ■/■•alee 'jpor. trie a,n-:le of attacJ/: ana res-altant vcioci'O; for •' bla,de element. lUCA -v-1. i -t- ! M ! "TS — 1 ( "/r 1 ^1 i ;■> 4J cS fl 1 T'^ ■<'' T ' ; 'Vl ! i I .. LLu. .Li_.L. "T" / \ i / \ V.x ho •H .I'ACA 'ir. 7 I -C^.:it,^l<; J! li 3 u !uO.. e Sec c id node a'..iir-i znoie jTig-are 7.- Lio.ie oliarjea of v ifcratin^: blade. IIAOA 2TTn '^'O Ar.i;-le 0.:" attack for no do'vnv/ash ^rain Minimiarn angle of attac'-. psr revolution v;itli du7/ri'./ash '^T.-iax ivlaxiiniLT. a-i^le of attac-- per revolv.tion 'v/ith do7;nv/ash r Station radia.:; n Rotational speed '•''oi''^mintV.Tia:.: flespective forv/ard Telocities of air '^Ro'%r.in''^?ina;c H^Fipective resultant valocities of olaag -.leuent relative tj air i'i^^j.re 6.- Effect of dov.TiVvaah upon the an^fle of attack a.ni rsjiultant .'alocit;,'- f )r s, blada element. iTAGA ?i^. 9 il4,000 .IS, 000 ±10,000-— -a TxS.OOO b6,G00 :4,000 Total vilDratnry stress (oscili- ograpn data) ?n stret-s con- ponent (v/ave- — — '', ?e0 mph _ V, 185 npli "* j'v, 105 mph; "bi?.ep, 100 lli/sq in. «;V, 142 mph; tmep, 150 I'o/sa in. Iv,. 157 n-.ph; brr.ep, ?00 Ib/sc. in. V , ?S0 r.ph V, 185 r::ph ''■■■-5- ] [V, 105 r;iph; bir.ep, 100 lb/ so in. analyzer lata) | Ar, 142 mph; bn,ep, 150 lb/ si in. [v, 157 mph; bir.ep, 200 lb/ so in. Fig^jire 9.- Effect of airspeed and braep iipon shani'; stress at l^aiing- adge position.-^ = 0'- , no si.'-.ulatei split flap on v/ing. MCA 'ir". 10 tl-^r.OOO I i j i I I )tal vi.l'ratory stress (osciliOcrraph lata) i ! a,3.;.^- ; V,134 r^pn a ,3. 9^^; 7,183 raph j c.O"; 'M^2 ;.:0i: 0,0"; 7,185 >nph 1 j ?n t?tr.-.6-i co'..'^onent (v,'ave-a:-.alyztr :lata) *lc,000 j-— ^ ,3.-'^: 7,13.i :npu : ,3.^-^; V,138 £.:pli -! ! ^.,0-'; ^.M-i^ mph '",0^; V,1F6 ;.-pl- j ' ' ■ j ! ! ! i ! J J 2-=g:' 2SC;0 .gine speed, rp;;. :'-?'00 figure. 10.- effect o:f v.ing an^^le of attack upon shank stress at leadinf-edt,-^ position. Brake -/.ean ei'f ectiv- pressure, 150 pour.ds per r.cuPre inch. ITACA yifi. 11 * 14, 000 *lc,000 ±10,000 ." ±8,000 &■ ±6,000 ±4,000 ±2,000 Total vtbi'atory otress (oscillograph data) (1) V, 138 rcpA ^ (1) V, 186 niph ' — ' (2) V, 142 iTiph '^ (?) V, 185 raph 2n stress component (wave-analyzer data) (1) V, 138 mph (1) V, 165 mph - (2) V, 142 nph (2) V, 185 nph. (1) ?lap on win^r (2) No flap 2400 2800 3200 2400 Eiigine speed, rpm 2800 ;?oo Jigraro 11.- Effect of sinulatei split flat) upon shanl: stress at luading-edce position, n, = 0°, bmep = 150 It/sq in. lUCA 12 ±14,000 * IS, 000 -10,000 -■ 6,0-0 ^ =^,000 I Str6s,3 ua.erii-tmcs ap-'jrjj-i:r.a.t.:- beca-ise of i^oniitioii of cer^ent. -h-- = ,000 =2,000 4. _...4.._..^|_|J.„,. };;2igine tx^ita Lion _..- ! 1 __f /:^_-.,.iLi i L....-., ..1 .- ... ix p= T /--' 1 ■'■■"' ^< / , I Kx T ,- ,1-..' .. 1 -, A\ l'^ 7 1 /^ \.fV\ i \^"aK-' I '•■L-f ■■' lit" r-¥- -..4.. 1000 l-:.00 1800 ??■/: En cine &pO'.:d, rpm figure 1?,.~ 3tuTi.-r.ary curvos of :itr?'j;3 .-it IPg inches from tip. r^- no siniulatcd split ila.p on Vi/ing; V, belo?; InO niph. •3000 = C^; UNIVERSITY OF FLORIDA 3 1262 08106 505 3 UNIVERSITY OF FLORIDA DOCUMENTS DEPARTMENT 120 MARSTON SCIENCE LIBRARY RO. BOX 11 7011 GAINESVILLE. FL 32611-7011 USA