an experimental investigation of a high lift device on the owl wing
TRANSCRIPT
AD-769 492
AN EXPERIMENTAL INVESTIGATION OF AHIGH LIFT DEVICE ON THE OWL WING
George William Anderson
Air Force Institute of TechnologyWrigh.t7-.Patterson Air Force Base, Ohio
March 1973
DISTRIBUTED BY:
Natonal Technical Infrmation ServiceU. S. DEPARTMENT OF COMMERCE5285 Port Royal Road, Springfield Va. 22151
DOCUM.-4T CO;TROL DATA -i & D• ,,',:y in.,,,. ,:,rl f j, ' .4 bffrear!n# , ir m, ,( f~ li be oiltet when the eve toll rpoet is1 fh,Oi IAflel
.t O1 C. .I , G C Tv I. '(Cosar. e .aW..r. 120. RLPORT SIECURTY C4.SblFICATION
Air Force In:L!tute of Tec:hnolooy , Unclassified
i vrin: ,-Pattcisor iFP, Ohio - ,
-;2 :.E:ETAL I:VESYiCG;.7I0:I OF A hIGH LIFT DEVICE ON TNE OWL WING
4. ; IcA.PTIyVIE NOTES Typt oIreponf 0nd uncluaW d41tr1)
AFIT Th-sis5. AU T"OA..5# %Fe, . ,ars.. anaddloo "sfliooi. last tl*o*)
George o . ArndcronC Cap ai, r. USAF
, b;eCI L , A I E 71. TOTAL o*. OF PAC 6b. Na. OF steFs
_- _ ^-;0.,'ch __73 _ 86 21* 4',.C" 4'.k', " OR GrIANT 40. Pa. O@IGIU WATONS MfP43RT NUMIKEIISI.
CAM/AE/73 -NS $. OTHCOR REPORT nO( Any othoW stumbe tAdo Amy be .8i0 e
10. 04 Tr'.4f'J £'.sN $1 AT'..MC1 t 1i
; .,rc for public release, dlstributic-n unlimited.
r~190e1A jF l9OJ 3?P-OMSO stod osLITANYTVIT
J/'/~'~Iit~k7 IAeronautical Systems DivisionjDe~orf I~nata SAFWrigbt-Pattrson A7B, Ohio11 D* 'ector of Information i
A study was made of the aerodynamic function of the comblike fixtures foundon-the leading edge of owl wings. Hicrophotographs of an owl's wing sho%.ed the
J ' r comb to resemble a row of spanwi-se twisted airfoils oriented to form a cascade.Smokc flow visualization tests on an owl wing showed that the comb actsas acascade which.turns the, flow close to the wlig leading edge in a spanwise direc- ztion. Flow visual ization experiments were run using flat plate and cambered air-foils with comos in a low speed three-dimensional wind tunnel. Results show;edthat the leading edge comb produced a stationary spanwise vortex that delaysfico separation at high angles of attack. The highlift device was related tothe vortex lift phenomera observed on delta wing aircraft. The comb's sr. allrelative size, simple structure, and lack of moving parts may make it attractivefor aircraft use.
2411;01xf"" byNATIONAL TECHNICALINFORMATION SERVICE . I
U S D."Wolmspt of C..umwvfSpril,,fieto VA 22151
L, .L-0 I Nov "-" Unclassified;I # &" becutny , l:amiancaati'tn
B estAvailable
Copy
14nWinga fif
SI d . ird. Fltowl..-. .
Snike Tun
v-,,
LjI
UnlasiieSecurty CossifcaiI
l-
14 -!-r, M T4
LGBE. '14WVE~V I I~S tIE4EST
REPRODU00CTIO,,N, A VAM~E OMTL
7>
RIM
A AN, EXPERIMENTAL '[WEtICtt IQN O, "
-V IGH LIFT DEVICE ON'THE OWL WJHG'
GAJI/AE/73-6 GEORGE W. ANDERSONCaptain UJSAF
Y'r,'roveU for Ptlblfc refeasq;d'os'ributaijn uIujmited.
i - -OF
-RiOAA-,#,>--K ?ID 6469 H :0W,
Vi -.
~THES IS
bi-ntdtoteFafculty ofteS 6q f: Lng~ej~n
obf ith~ ir tfor-e Inshi tu o Bf Te~V h~ hip I--'
G Aog6-W. nerst, B.SI6P~ II,
enPa~ta: FlfiMaht of th
Arodfoquict rlehe; disribuof unliijed
feac
Aji s thkEi std f hatural-egnei~sses ~nt sav
'aa age-66ee it~; 'tl i' the Ilc~rtr 60 engieeing. in fac:t-, [t,
ab r qfo, b eiaJo tnj e reij~ Wit I he nwr iray
Sytem. v" ~,6d*pe
-tenfy-grad- ted fro th-eopc ~t-1lt~at EwrsA
CH'fr Vn i,- kn6wl'edg -of the :Ohhnea f elat to f wt fgtas
thebe~nngofth s tuy Thi 1ocso td he Itd n~e]d
'Movementb fehrlchIrtel. h-!thtk- bosok- the N66ct 1o w, te:~m
is~ idntifie as i - Wo vd_'prob* in fluii-dynais N niotj~
hwvri the std udnthv eew.,practia w.thotth idns
04de by- Dr. R-.A. Kroeger -of, th~ ni Jversfity, ofD TennseeSpc nttu-
In a~wirid- tfiiel t-es't of the owl -wing§.
I made many 'friends-in- the- engineering and: bio'logical salinces dur-
Ing-my study of the owl viing. I t i4 I- be -less. -than fitting, to-omit, the,
-names of these who spent the most time in giving advice, direction, and-
encouragemfe-nt in the study. Aeronauticailengincers Jerry Martin and
Jatmes Snyder of the Preliminary Design Division, Directorate of Advanced
Systems Design, A$D, sponisored the project and acted as advisors. Pro-fessor Harold C. Larsen, Lt Colondl Frederick F. Tolle, and Captain fR.F.
- 6A414-6-
Oetgn ee heAT acy, advi-sors Cat" oe rwodAct:teotard 'WheOf the AFTgh- C-artc 1oa t~cztfue e
k-, V ff8,to , - e&
Patrsmoft~ Oudor ~ct~i Cenoter, Antohlinadai-
SMCCS%0 a(; fsbehaviout.h Tood " vhes peopl 4hoent ol~hepo
btisoan lmdrea~d ,a40 aton f lilh~binict -a"6ah, to i!
.C~ogeM.Andeson'
14-
A ~iients.
A th&4S -e4
- ~ ~ ~ I d Pr f c . . n * .* ,. i, * 9 . '2:~999i h.
i9 4,'9
I. .' Tnr d c i n .- -F . , ** * . * ~* .~
j14. }i ,4-0
L Iih etr Sua e 1; nT~ -Iet POO~ .~* ' -. ;j44 . i n d T u n e l' e ~ s ~ f w l W n o . . . . . ~ *. * .9 .. '3 7 ~
. .. - ~ter Tunn l ~ xpeimejt: ~ 9. . ;:. ~ ' 177
lVi. ~asimet ke*su, .. . . .. 0, 0 0 *. 0 *,ke0e~i 0. .. 0 47 3
Part~gah 2 . . . , .. .. . . . . . . 4
Theo r ol . . . . . . . .. . . * .~ . . . . 614Vorked Comb . . . . . . ,. . . . . . * * * ~ 6) . 4
Tho ary Wing* ** . .. . . . 68
iv
'----S -. ,t~< ,-,-----~ - I*-,~ az. -- -- s-e.h ~ A
-~ - ~ jaFt
-~ - -- -~ - ------
Ak1,11 Rarnsv..a..A,.&t
--
a a
was. ~pnc)usJonuand *'a.inn.wu~nuw&.OflS a 0 * * * 7)3k)lography ~ a a.., .- aa.73Ft
aaa~~*., a a a a a a . a a- a a 4- * U 0 0 4 75 :1'1 Vita / - - --
-Ii- N
- K---
C'- - - -
-
'-AV - -
-
iv"
N -
- - ~ 0'S
<*4 - - -
t '114-
-,-
- /
44 - L
-
- - - - - -- I11-
- - - L'-1
BI <- - -- -~
VC --v
-
--
- k-hI V~1 I
1-'iv-I VA --'ItitI &
V
- " 7V
:.<
V,(
s
_ ~~~ 3 rii flaty Fe hr T~'i4; ::':
UF e v n- ",n ev , 0ITg n1eo~tak .... 1
ii Stealies U'pe~ ~4 ~ ~hkilea'
'Attack~~~~~~ Jn'LaigEg ~bRmv
12~- +~o ~rn rahgb h l~o~ ob(),*. .1
:1 Cm Sae Use n,~a~ Tu1' xe iins() 1
ITIiCk 'hL06 I;0
17 Wi rg Sections 1 and 2-Owl. .. .. 27
~'16 O Wing Sect onsan i .. . . . . . . . . . . . 26-
Is Section 5. Position of Tip Feathers Under Aeroelast~cDuring Gliding Flight . .. . .. . . 1 29 1
20 Comb Blades-Mlagnified 30X 3
21 View of Comb from Free Stream Direction . . .. 33
vi
22~AII/A~t73,
24 'C6,~ae o Vewv -... .e .34
Cob6 Mld, idVio . ~*~3
'FP5 con le,- Left .063 tiue n-A 4 . . . 'A - 35-
*4
26 Co6ld~igh 4, ii u
409- e tM 6dei4 J
_A27ii A ;
now'-f'129d t ** *-. . ii ,4
~~3'O~ 0,~ee Com 4a q. - ' .a . i . . .
k2 SiOk~iorBscWnoe*33~ 30 ' 5, 4,.S8 ': 4 ;I1I * i ~ ~
S 46W tieC' 4 6 0 ti i tob
36 Spa' s e : k i,4: ibgd' -itv 0 84ttzr&. 52
37; Vr.x aS --t; SiW a. a.. a....
5, - 27, VS15, f to/tex
36 Ataceflwwwt Cm 6 Inst 'Wlhg
37~~~~- V~tx "tet h5d f'rn .Wr/sedtn E~
41 Flw SeparatiEd en atTriow Eg
01 U 19,, V w15 ft/sec .aaaaaa* aa 55
42 Flow Fully Attached Above Plain Wing Stalla 23, V *15 ft/sec aaa*aa aaa*aaaaaa 5
vil
Le~~ngEdg 'SparationA Region Ejideo
i5 -St~at) '~aa "o Reio a t 'za -5o-7, m157 fTOs 's 4 .. ., 4 4 -a. -i 4
- ~ ~ ~ ~ pr-if- tp-ak6Ta:%nvdg earto
un29'V if/ c C Y* f . .j *.~
Nil -otx1 oei~mitd
;-49 'Ba~taed
59, Apr~aciiIngj, Sta 1-- with Ba4stared;-Wi hg I nstalIled fOle 34V 15, f t/sec . .. . . . . . . . . . ..... 6
52 Separated Flow,. Blowing 0 ff
I53 Attachled Flow, Blowing-Oiu4 -35 V -25 ft/sec . ,.. ....... ..** . . . 62-
514 Vortex Sheet on Delta Wing (19) . . . . . . . . . a. 63-
55 Non-Linear Lif't for Several Delta Wings (21) .. . . . . . .*
56 Non-Linear Wirngtip Lift for Several Planforms (14) . . . . 67
57 Yewo-Dimensiorial Streamlnewith Spanwise
Slurry (16) o. . . . a . . . * . . . . 69
55 Lift Curves Obtained for Various Degrees ofSlowing (16) . . .. . a.. . . a . . 69 aaa
60 Vortex Sheets Generated by the Bastard Wing .... . . . 70
Vill
-- -- " + Lis .or., + +
T..of Taibe
I iieasureunts of, Sever -di Grea t Noie Ow ' i,:.. . .. ; -25
-4-1
2 Q
-77 77I
;L- -T9sti ef f mboifi
A 6Piling aspect rit i, OrfS -
-b Wing sae
V, Je elcty pawseb-wn
V.L Fresitream lf oefcit
40 Moas dnity of airn
~iaxniiw lit ceffiien
KsI(AE/73:6
- Asrac
toeii -found',6) on te Ilead i-og egje-4f 1iw wngs fic htops, of -4h
foi l rented tp [fre, a cascde. Smoke fo ~ufz~o et ni
owl ~ ~ ~ ~ ~ ~ ~ ~ - kwn h~4 htteci at sa-~ade, which40 t 166s thfobw
-- , 1a66o experimenhts Were itun using, fla#t-paeadcmee ifiswt
omb ina low sped heew-d imens Ional wind- tunnel Results,-soqed Lnat
4the leadlhng edge. 6mb, peOuded a. stationay spanw4iso-vorteax-thaft delays Ilwd- prab at ,h-igh angs of attack. he Ighlft. devicewa5 -re!-
I iited to-lhe, Vortex Ilft phepnomena obeVed!.on del'ta wingj ail-rcraft,..
The comnb's snalI rei atd ie s ize, simple -strututre, And lack of' MovIng
parts may -make, it -Attract ive for 4 1rc r# t- use.
0urpse
-Achiievent, of 'hg4her liqft coefficietn0ts j teremely low.airspees
is. A,:Or6be -of* geat imotnei ent, ST VTOLairaft'at Aevelo
iient-. 'in spt fit~te resrh, -1h this,-are&A miy 4f t'e urret
and04 prpse iyh 'lIft -systems suffer fto, eris s.Oretcoings. 'AiMo
dis., in vieOw of these proeblemsi an Ui4al' hitgh 1i ft Aevice, c~ude~
vis ,ioned; one ha i noMoving par.ts-, nho direct use -of -ehgn 'hru
and- negli1'ible effect on hi Oh, speed, perforMance. The'higb 11f~t dev4c-foiund on the wng o4f -is pcIe of qowl~s -deserves -the rynm tsS
Attention, in, that Tt appears to, -have 41ll of these didel- qutaVitl es. A,
previous wind. tunnel study of an actual-owl wing hbas jdentlfiid the 4iI' h
lift function of a comb-like. cascade located on' the l.eadingdge Be-
cause of the il mted &iature of 'the study, no-detail ed data -Was -gathered
on the flow mechanism or performance of the hi'gh lift device. The pur-
pose of this study wavi to identify the flow mechanism present on the
owl wing and to produce a suitable analog or model of the device from
which performance data could be measured In the wind tunnel. Further,
recommendations are made as to the suitability of this device for use
on conventional airfoils.
Method
The general method of study followed guidelines appropriate for a
bionics type of investigation. The f;rst step requlrd examining and
photographing the wings of live owls, obtaining components of owl wings
for precise sectioning and measurement, and gathering data on owl flight
77 [
from biologists, a4 natural)its. Co curntly,a irtu eacws
Wd Ito gather avaiaedaonaryamcher r*eevn to br
'fIgt OPoll h 46 h4aa i
todeig adconst~i -a :pool of ~the l],d~ing edge oiiThis, cbb wtis tested' on1vrlsml ifis s~fliiulia
tion technOqes Ita thri -dOi!*eS-i~hit Winhdtne. Ii~h and
-the -flow Visua firatioh study suggested- cage, Thn -the, co ~ *6del 1nti 1
a c46ose, duplica*tonof thi, fjlow ovrteolwn6wsatie. Sveral
f4ow visUaliZato ehiuswr tri.ed hqluding, sok tJitAnrlusa AdIo$Oe-
slurr-Y, jind''tuftsi- The smoke- tests proved, -to. be tenst, 4iccesf~ and'4were, usetd almobst 4eclusively in thOelatter sttages, of. the.stu~y be
r Vati rom the sm.oke visual izatilon sthowed t6e fIo Uleld.,toi Mvemhany,
- ~ of the, characteri'stics- associated with- the l1eadligO edge, vortex -phenpomenw
observed-on sharp lead-ing. edged_ delta wings ,operAting- it h'ig anls, of
attack. The literbture search was- -then. directed toward-studies iUn this t
area and was successful in locating-material that a1ded in, expla4inihg
the flow mechalni-sm.
-Bakrund
The unusual configuration of the owl wing was first brought to the
attentonr of the aeronautical engineering community in 1934 when Lieu-
tenant Commander R.R. Graham, R.N. wrote an article on the subj%.ct in
the Joural of the Roa Aeronautical Soit (1). In thls'article en-
titled "The Silent Flight of Owls" Graham describes the hooked comb:
1"Piere Is a remarkably stiff, comb-like fringe on thefrcont margin of every feather that functions as aleading edge. The teeth of this comb are extensionsof the barbs, or fibres, that form the front webb ofthe feather. They vary in length and distance apart
2
- h.
~~'n - e- -- ,- -a -
ad -00iiu- apart ().'llgh
-this cokb iilo6 Wih eote e lar tso teowwig ascr
sl~eed y G 'ITr as conteibtUing ,tq thd I'm! qieflgt 4~h
cas6,of the-- comb ,O rqooses tha aaroynmi qutn i ud by,'a
reution :in flowvlct ttelaigdgote~n. Ff~ig
7j a.~, a bref os 'eatUf the, co I posible effect on -ud'6eert ion is-made'by Aujijit Rdspet 'I ivn46 a4tcle~bihd 16.
- c~ntvt1.d Biophysics of- Bi-rd TIlight1'()1-sptW.e thcp
.1 ndrical wires gqenerttn.iae61k Ian toe ad otsthat fiaefe
, rquenqy-df the,,emitted- tolm. is, af fntion-4h of, h;_y_~e df-te. l
tsor yas thei conuihon use of' various- types !of vre gnrtrt
dela y flwsprtonon-airfoils 'hat Oro66otpO'.the ithio tha -4is
function In this Imner and are not priMrily ntended. for aderodnm'
qqiet in'g E, Bu'ck, of the Univers'ity of Oklahomfa,.states I natce
entitled "'Bird -Ac-ocynamics" that -straih 9 in tlce or h laig
edge of an airfoil in an-arrangement eembl'ng the hooked -comb, results
In an increase in lift (3). A recent NASA Technical Memorandum-aliq
Shows that leading edge serrations result in si-gnificant increases in
maximum lift coefficient by generating chordwise orien~ted vortices (4).
The best presentation of the problem in engineering terms, however,
appears in the book Struc .urerForm-Movement by He:inrich Hertel 5)
Reynldsrnumbe-r of the comb-wing system. Hiertel makes the important
observation that the hooked comb does not resemble the vortex or tur-
bolence generators used on conventional airfoils because of the orien-
tation of the comb blades relative to the airfoil surface. The -Zon-
3
-7,
_AMEz3- 0i -
ci In- that Herief kieaches is that thgj;.ihc06ho-ofJ the cbn 61s- n
. ,~i. 71 -' p,
fovdpole nfu i i-Kodt 4hn l-~ei ai, cthifi tty; nv
nese S~il nstitute (6). The- stud ,eite q o eeAroyaCS,
contributed'~Av Jo -niesprsin ~ nOtant:-part-of the -studty
tst-f ;An low,, wing§. The, esuliits o ti teft% shqod- that. - the b ob-di-
Angles _of attiiqk.-By rb the, * 6lw f~ tiMl d'W# _Oq, of
the~~~ sWealie o'b ttetpo h ig ptined-o This, le4 0o the-
discov4ey that, -t e o b~ a r d c n , a v r e - etthehe ~' ' e di!
ing ed~ge -and directing spanwise f-low tOward :th 'dtip Thd ~t on
jstreamlines, And 'the dis covery-of, the-hi'gh lift func~tioqn of tkq oifibi
proyided, the -baisis onwwhich. the preesent study -was undertaken'.
The 46fgr*iio ;~d-j -n, atr
dloi~yxisiiile th dj i- f--g6 the ah - V h
,)owe! strtingwith t -dee jhint, Orogressin p own ;p7the pe r,
anzernatg t 'the wr, joint.i Outboard ofte rst >o r
b0on ts -control lable, by thet bir tatemaein -featesoie da
r I gt angles -to the, d Ir6ctlon- of f It. These -feathers 6 re- termed -th ,6
primary fight feathers., Takenl ai a group theiy form -a mavable outeer-
pbrt ion, of the winhg tha t is, -often, -rOerr-d. to collcivl at.mmaus.-
The-feathers of the iliauv can -be-moved indiepqp4!enly- or as a' gr ou the
muscles. FI.Iing out, the remaIn~ng plIanform of the iing, are- ethe-secon-
dary feathers. These. a re held in.,place- by the fledsh -or pa~tag! um-t surun-i
ding the bo-r~es of the foearm. The -patagium in- additilon: to securlng .tho-
secondary feathers al so forms the -contou~r of the -leadInhg -edge o' the.
inner portion of the wing. Projecting forward from the top of th'- wrist7
area are a grouping of two to three short feathers that act as a control-
)able aerodynamic "thumb". This feather system Is called the a'lula, or
bastard wing and Is well developed on the owl. The physical shape of
~1the wing is completed by the presence of covert feathers which do not
contribute to the planform of *he wing but give it additional contour or
thickness. The covert feathers are attached-to the patagium boO above
and below the ma In flight feathers and concentrate,additional thickness
o'i the wing in the region directly behind the letding edge.
~1 The primary and secondary feathers form the wing planformn arnd are of
special interest since they are also the wing's load carrying structure.
F ~ -~ ~
* AFf7~4.. -~A C C
* C
UI
C
L1~0
L..0~'C
t -
(
I1 0
1~ED
'4-C
4C13
CQ 1~ VU
'A.
I
I *
~1
6
_Thisl feathes are Omposed1of aci tra haft with aieor leafaces,
-rnigaon-1to ts lenthj, Thvane '4s bu ilt -ti _f a,46es of:petl 11 ods or, 44ats 1ba t a're 'boud& tge'ther along thei TeithbkY'
6bdbue., the. 1ab i funct-,'n4 "in 'a JManeh nlgu oho and e
soecohn.This, feature a-4lk*, the barbs,'to -fort a I' Oitsufae he prt b orma l tesses.
1tI blu ta hebf i owl wi. is "~~ abl g ,'eometry ,Odeewhth-
pft grates th ,e!S' theyust, ,and i4'ft -produding -medhanitss as. Tel as
proViding Ost ofth eans -for fIght c6htrbf., -Thtts -factors- equre
act Ii t6o4. pas sfv4 6rea s, 'th~us t rtequ -ireents. d icte On Oq~gry
oqf wing' cha*ngent -under- dict, muscular -conte6l- of 'the 'bird s -the
-OesIc prprtie odfthe i 440t feathers. 'Aerqoynmfq forces act-I
Iiig-on cheiie feathers can be--resolved-into tors iona 1 -afd' bend ing momen ts
actirv§: on' the. shaft.i Iin -f 11 ght bend ing -oments acing on te prfiarj
feather-s of the manus cause the wingtip regl6n. -to bend upwar an pen
Into A series of multiple airfoilIs. the orlentation of these -inividuala irfo ilIs i s not I n-the p) ane of the wi ng and va r ies w ith the owl't flIight
condition. A close examination of the-manus region in Figure 2 shows
that the first two primary feathers have their vane planform specific-
ally designed to-accentuate the slot'effect. Sending effects on the
secondary feathers result in a change In the physical camber of the in-
board sections of the wing. Torsional moments about the feather shafts
appear to play an important part In controlling the porosity of the wing
surface. This porosity is formed by the slots that open between the
flight feathers during the flapping or powered phases of flight.. The
+-4-- + +.+ i-+,-. - • 4+* -:++ 0 " " :,-
.+ -
4
.+ + +_ ++ + .+.+ + +_ _•+
. ,+- - A I E 7 - - . +•.
-?, . ; -,+ - -•
- + - + . + _.- • +
,.+ . + ... .+ . .. ... . 1 +
\ vs . - . + ++ + o ..4..
,. -+ 4 .-4 - S -. .
C V, + +, ' + " 4 -- ---4 - - . -- . .. .. - - -
+.+ . ,,. , : - , . - - - . - .. + +-x-+
. .. 0 _
-+
"4+ ++ + + . - .. • + '-; I.+.+, + .
'.C+ - , q -+ + o • -.... + - -,1 -.
-+ - (' -+ + + -.. - ' + - - . -+ .
-.+ + + " "-- + + ,- ++ +" .+++ '4 , 1+,+ ,. +d
!.- ,+ + . . +,2_ V,
- I - ; ,+ - -. + -+ >. +'I¢ ' .+ ;,, I
+ " + + q - ++
-- 44,,
+! ' + .+ . ,+,. + .. .. , +., . . ++"+ ., f..t+ ' I'+
I+" + + *... = + #o +•<- 4. + '+ -'2 • + - , + : ',- ',+ +, _ + -+.;+.+4+ ...+, ,,
1- 44 4+. . . ' +,, .+++ -,+ +- . ,.+ + ,.- ,++ .
4 ' 44 44 C +'+. . "',+ -. + . .
,+ ,
.4 ;1 + ' - , 4 . ' . -I
+ .+ ' + + '
+
.r5: + + - " ,." "'"-: ..;+ " o..I Y + - 4 ., + .
+ .. + . - - . + ,
7 .r- - + " S + ' + ' :+
'>+'" ''': " '.+-• " - - 4 . "
I'-. -
1 '_ .+, ::-
44 4i -
+ + , , - . . -4+
4 4 +, - -., - -,. + + - .: ". , - .
-: , ; 4-,
" +- ,
-- j4- 4-,,
4- - - -.
-. - - Ae4* ,4.4b4
+ ji~
+ , ;- + +:+ +I,+ B
-fuiitt'on a-nde deee' of -ti oo-it snt wel, unertod, ho e,
severl stdies'havebeen;e dcted ThL #1i 'fel bytewusas(
* Teo) lno 0oissesses- a.- ntqber of chaeactterlJtlc nIot geheeal-1
fo'Mo te i dwn The fi est of these- is i'thei booke&.d Vojbon.
the leaI I edgeof, t he0 +it and-second pr'imary faFi#4 Vguei.3,
-And, 4 iow :the- c6f,omb pbttioh' bn the fles't pimaey' 'featue. -the-,;b
Is ormd~ y- he nds.ofth babs hatradatef~rm-he saft of- the
feathrs. ward-theextreity f ch ~bafr~,2 thebarblesta om~~
hodtevafne tqe#e', t a~frv lo 0 the : tip of" -the "barbs -to 04st!
aln -reA axis-. I't canS be s een- lthait the barbs tape sharply. at, th1e
ends aindt~retai 4n1 hthe tba eb'u lsv the ou,4tboaird -i~e he ite of the
lnivdtaV ~ th cmbttis frmtd varies W i th- the- sie: fid
specIes of, owl- as wel-l as;:with -the, position .along the sh&Oft Of th-fe-
'thtr.i Rersetoi~ d --ensins are 2 idn. for the separtdpotnf
oi4 ig4-hf~nigta occurs -in the vanes that make up the
ti-111§ -dge-potios o th-wig.This Is shown i iue5 nta
of bndig te vne ogeherout to the apex of the barbs, the~ barbules
again release their hold and the resulting separation of the barbs forms
a discontinuous surface. The appearance of the separated barbs is quite
different from that of the leading edge comb. On the trailing edge the
barbs formi soft, flexible streamers from which minute barbule filaments
-adiate in all directions. The final characteristic is the downy upper
surface on the portions of the vanes that come In contact with adjacent
feathers during flight. The contrast between downy and smooth areas on
the upper vatte surfaces is shown in Figures 6 and 7. Because the downy
surfaces are present only In areas of physical contact, it is presumed
777 777 7-77-17
-~- ~ Jut
Z"1 x '0'Ic Fill;
U))
&i
nl4
& ,,,- -011A
~- ~ <tY~..<~9AY ~ ~ -q r5 '5'J -- - - .
- -' - -. -~ ~ ~ - - - - - --~
Ft GA$1AEp73-6
- - Ir-~~?r -~--,- -
t t
I- I
~ -s-- - IQ I
--Ii'
I------ 1- -1
-: ~ ~
'1 -~ ii'~"'-tr' It
'~1 4- 14'1 s--: ~C0 IV - ---. -~ 04'
I - $'A '4-0-a *I U- 'I
- -' - 0.. 4- - -~ .0-
-. 2"it - C)
1. S.-,- Qv~- S '4'
Sr 7- - '1;;*~*;~-t,~* U
~. I)L
'1 9-- .~~~---*
F I: ; I -~' I
I Sa II I
- ~-~--s-~-I-
1 -#
-t
7-4 frIS-nI-~,
t,.. ,~d - - I
-I
-t ii
'I
-A
ha. he prfo; wel'ther a- Avublcaiii4 or qu!ittinq. fu~tJer0 (5)
4- _A
~~12
I I
HI.
-Pro
Figure 7. Downy Fe-ather Vane Surface
13
<
e -
il.Lteaue Survey.of Te-*;t- D at a.
i he, sttudy bf bOwl 11 iht by- Kroeger (6 cohia i'ns- a gr-j6t, deal of
data- on the -aerodynamic berformance-of the 'ind iduja 1d evices-present
-on' the OW owlkhqin Thiik-djta relevant to, the hocked comb -probiem fjolsi
& ~~three Aain- categories. These; are windunltsso eee-w
wi41 water tunnel test s of-a -hooked, comb- analjog., and: free -fl'iht, ex_
petlhients, us'ing live owlsi Each-of theie-exOerjiehts is--discussed with-
respect to their contributions -to the understanding of the -function of
the hooked comb.
Win Tunnel Test of Owl EWj
A smal'l owl wing was placed in a low speed wind tunnel and the
flow over the wing studied with smoke visualization and tufts. Kroeger
found that the smoke flow over the outer half of the wing vqried with
angle of attack. As shown in Flqurcs 8 and 9, the flow over the wing
turned inboard at- low angles of atta,;k and outboard at high angles of
attack. An examination of trle flow pattern on the upper wing surface
also revealed marked variations with angle of attack. At low angles
of attack, the streamlines were primarily chordwise and no effect of
the leading edge comb on the flow field could be seen. At high angles
of attack, the comb created a spanwise flow toward the wingtip. This
spanwise flow is shown in the surface streamline sketch of Figure 10.
In addition to creating this spanwise flow component, the combs gene-
t rated a vorteg sheet at the wing leading edge. This vortex sheet exten-
ded from the bastard wing at-mid-span out to the tip of the first primary
I feather, The action of the vortex sheet appeared to delay flow separa-
tion on the outer half of the wing to an angle of attaOh in excess of
14
'--- i-. -, - -
~P~*0 ~ ~.- - - __________ -,-~- -
G~fAE/7~3~6
--- 7---
*-1I
V- Figure 8. Flow Over the 9w1 Wing ~t Low Anglesqf Attack
I* I
II
II - II~~1
I.
VFigure 9. Flow Over the Owl Wing ~t High Angles of Attack
I-
154
4.7.4
.----- - - -
r I~,.
.- . - -- I . -V
I
I-- . - - --v- I
4, S - --
I . --
I' I 4,
I -I. - . -
$'4, - -
- -I- - -. -
-I - -- ---
-~ '-
A- -
4- __ ~I - - Ca -U -,
'II -
- I S *
S.
-' -. - ---- ~.b . 0-, - -~ -z--- *1 4)
* %I.R&
- fI~
* ~'
-C't - .-. '~I -* - zI- -.
,*~'
LI'4..
0 *
I-'
C
-~ A~
~ -
IN.
u~0 LII * r... .I*J
F -''I
I .9 >4
*1
fr;~ C'
I.' *I~.
I I
i 1 4
~l4&I.4
16
3dere.When the. dob Was e oed :from thie- Tadig ed"e te wing
stalle'1d at a, much, lovier ,anqte ttcad aapigotesfc
kroeber reported- that the comb' -ef fect on the ,Flow, f ieldtvs o g
C hee dimensiohNoi A shJoW~ in -Fidure 12,j the: tusrfilng of #6o flIoW jthirough, thed comb- is .highest 4t tewgsraead dceaiseis -wit h i
creasing distanc from tihe wiinj surface uintiI the , :itw 15choidwise at-
the- comb tips. No deforniation -of the comb 4ccurred during thed -tts
in spite -of the, smalI -relative, size-of the fndividual- teeth. 'in ex-
get,-that -the vortex sbheet and -the unusual- coutretth-foft4
atmid-chord are created by the mutual action of the, -hooked cobb the
hsadwing, and the slotted wingtip.
LtrTunnel Experiment
Kroeger attempted to duplicate the vortex sheet flow created by
the ombsystem in a small water tunnel. The shape shown In Figure 13
proved successful. The comb was made of .0014 shim stock and each of
the blades were twisted until the tips had a zero angle of attack with
respect to the free stream. Although the analog app~eared to create a
vortex sheet downstream of the comb, no tests were conducted to measure
its function as a high lift device.
Free Fl ightTest of Live Owls
A Barred owl (Strix varia alleni) was used In Kroeger's i'light
tests, The owl wc constrained to fly a relatively fixed flight path
and measurements were taken by meains ofl photogmyhs. By computing velo-
city and glide angle during the !Oliding portions of the owl's flight,
17
M--:
I4
In
J I
.19
-.4 CAIE/3-6 -
par~-'t reorte vatTe of 1 tot for ai Sorig black d to~ '.
4 the wing r,, 1ot used 20t the ih-lda,~e u6 eh§ i hdek, ir~
I' id 1-6-_'
20
*i- -_
4
N I
'if 4 16. and j
a04 pSio Iose 'o Att
44
ingjs ede-nd i ad into the surroundingcovrt: contourfeathes.
'The ci nrt fethersA~,hfi espcill on- thel 4hne wi, ar cereylih
6seve author that'ng ft he forme a- cplt s1rac that func-
complnturface tJnod delaye trnitio4nell halenetbiseyyKa
wasinuce t fy btweneleatd ercesPna out~pdor encosureah
~i feetlo Onh -mos o te ligth owheiretlovrh.
observer~~~~~jri~ athfhso - et.Anmetf hr celt~ &r
no-ted d -ur ing he flgts:td i h-apr'int ii-oito h ed
an-dw. Ths owlharantexrstic lngs gldec phase during el
sevralawhch tht wing reris fra comel iean surformtatfu
tin odlybudrylyrtasto 5,) hsaiiyo
comlintsufae o ely rasiionha benesabisedbyKrme
.rr rr~~C A- - - 4S4~ ~
r'C-
-c- 'ot~V . - - err'--- - ~
-V - - - r - - IFl K- '-- -~ K A" -
IWI
-- 4
I --
7 1
A ~-,
A 1A - - -- V 1-A
~' I itt --- j It - - j.- 'A 'I,-- - F
H- 7;H
e~-4' I ,
-~ -
rrrl -7---- -7 '-'
t. C)£ I
Li <A - -
- - HiV
)- -.
4'1' - - -.
0~'-~
-' C)- -.
*
LI~.2'
-- 4
I I
I
I 4
C
22
4
<j
- V"'!- i -"
.d -
-! - .- - '
5 23
3 S'
ihilatw4 fioo -- h *ifii§ tr ,. :4Y-'l'- -
1i The sJ'-t~, h, 3 -iik lq
Ut- clo.ase t ftha -obsrve anwin ,w 1 t esid, o-4y~mq Ch -,oI 4-d qen
3.. an Thei dffecutirqehth Te tie feter -wa Ofal buthi-
owlW~s nt A I4-'i wa comy u visibeetg. 6a-h---I q
-4l of thtsis41. The glideht k-yng64s n~bwas apfrroimat' 6h5 tghe
chr - uItngt tke git mi&-sga phaeoc-y 5t/. Th --4 g
loading auens of.674fft-I ls o htmkue oohroe
I -ato betid suchasreinei tis ve an ow ii~ wriede- to conudntions airo--
crayn.mi cmatrisotics. 10 The n Tos ourd iTe a glrf dh c gucap-
(1).Fo us1 and t1 show e thbe win etakninse tha t take tprsetaiv poiton aln h pn l h etosrsml oiveowln arailabl iht whes cputedsibted wbyg disrbng a y4 cathickths snelpever (a) Tamer linh, ein nube was ae n the
tcdkegt takt ofthid-gan d vigliy ofbd 2Section 2 taen wngI loadieng tae stadh e6n7 ~f/the rgclose tovha eredcvr fortherspe--
'24
Measurements of Severed Gr eat H Owl '1 .
Vi. i q eht, I b* ,-1.7"
Wing areak*, in2 398.
Wing Span,, in 8.0
Wing sweep angle, deg. -01
Chord at half span, in 10.
Wing loading, lb/ft2 0.64
Wi'ng aspect ratio .5*8
Table II
Owl nj Section Measurements
Chord Camber Max. Leading
Section (c), in Camber Position Thickness Edq Radius
1 9.7 , 14c .32c .68c, ,03c
2 9.5 .09c .44c .07c .03c
3 9.4 .08c .Alc .03c sh arp
4 8.1 .09c .51c ,02c sharp
*Weight calculated from wing loading value of reference 5.**Wing area includes the body area intercepted by the wing.
25
AI/AE/7-
- -I-.-.---26
~ - - 4
GAH/AE/73-6 -
I1
(.
I
4)N -
0
L -
4.,
U
C - IQ
*IUU
U)
C,'
4* 'U. 4
Cgo
C
- 041
C - C Uo 0 Ci
- 404J 44U UV 4) C
4- U) U)
r-.
4)1..
b.
I
*'1
27
A.
GAN/AE/73-6 V
C -
IIN
-4, -
0 --
ED
4.,-' 4 - -o
C.4 I) - -
0 I.'ft
'4-
C I .~. 4-
£ V0 1'~
-vEW~ 4-iLi
0 4
4 -~ 04-i .LJLi Li0 Ii
(J~
-4 01..
II
~ ~ 128
* 1
A-
S i 5f -
rean farl thc u a eue abr.Amre hnei hp
multple ios. Thoslaivestion s of Ti e these aidrf sectio
oarsw in seFigure 19d. Cosierige strnstre ofate ing hancs
In the gaoinet.ry of the Inboard scatlons duri'ng giliding f:light should be
relatively small. Table 11 lists the section values of camber, maximum
thickness, and leading edge radius. Using these values, the following
comments-can be made:
29
. etion 1 has a -Maxi MrA ca mbert o f l4w ,6,ic s muc dh,
higheo.r than on tonfvent ional a irfoils where -val'u'es
ra'rely exceed -4%. This -high,,camber sugest high.
section C1, and large pitching moiments. (131.Imax,
2. "ectks 3 and 4 have, -thicknesses-be-low- the mini-
f ~~mum -valIue of 4% usually--given as-.a structural l1imi t
fin prel.10ina- irrft eyg (14).,
3. The, maximum thick~hss -on section 4 is 'located be-
hind mid-chord..suggesting a large percentage of
laminai flow.
Hooked Comb
In order to model the hooked comb, its exact shape and orienta-
tion was studied. Since the comb is too small to inspect without mnagni-
fication, a laboratory comparator was used to select secions for micro-
photography. A 35 mmi single lens reflex camera with bellows attachments
giving negative magnifications of 10-30X was found to be suitable.
Microphotographs of the comb are presented in Figures 20 through 26.
Figure 20 shows a top view of the comb system magnified 30X. Even at
this high magnification, the surface formed by the barbules appears to
be fairly solid except ',or the fringing at the tips. Separation of the
barbules from 'he adjoinling barb occ trs some distance prior to the be-
ginninq of the taper in the outer portion of the comb. Because of this,
the barbule surface overlaps the adjoining barb for ,short distance.
This characteristic cannot be fabricated In a model cut from a flat
sheet. Figures 21 and 22 show the comb fromt the direction of the free
stream when the wing Is at an angle of attack of 30 degrees. Irn
I 30
,A/A E13-6
$Fjigure 21 A, cscade is, Aformed by the barbiules _whihreseib e aserieS
of individual -blades0 and turn the 'flow tbord" the +wi n d Cip. If -the-
+ ndividualbarbule su faces are aissuied to- be rel aclveIy impermeable,
the abiI ity Of the cascae- 0turn the flovt niy -be qutte hgf. A pri-
:mary variable in cascade geometry is blade so i|d ty- def-ined as the,
ratio between blade chord length and blade separation. In this cascade
the solidity is quite high in the base region aid tapers ofF toward the
tip as the barbule chord -length decreases.
Figure 22 shows the shape of the barbs on the undersideof the
wing surface. The shape remains rectangular until -reaching the comb
where it thanges sharply and becomes cylindrical. A critical parametert
in specifying the orientation of the barbs is the angle at which they
radiate from the vane axis. This value is approximately 25 degrees
and varies little along the span of the feather. This contrasts with
the iarge variations in barb angle present on a flight fev.ther without
the hooked comb. Further insight Into the shape of the indlvidual
blades or barbule surfaces can be gained by sectioning a single barb
from a leading edge primary feather arid photographing it from several
angles. The photographs in Figures 23 through 26 show the barb in the
top, left side, and two oblique views. (The hole appearing in all
views was made by a pin during sectioniag.) The top and side views
show the twist of the barb along its axis. No twist could be seen in
the barbule surface about the barb a is.
Conclusions
A number of conclusions were made after considering the .wlas
wing structure and aerodynamic performance:
31*
A--- ----
GAXIA/73e rod,0' Cod'
Va; / rt$1 r-p4 iur 20 ob4ae Agiid3
- .4-32
V -, -7 .- i - - : i ;.AGA/134
iIi- 4i-
11.1
4. ,i , ' jA
* JAM_ I
Figure 23. Vew of Comb from Free Stream Direction
A9, i i " -j 3"
a
Figure 22. Barbule Structure In Base Region
33
I I I I I-IlI II IIIII
I I III III I Il
-i fio-
> Figure 23- Comb Slade, Top Vietif,
Figure 23. Comb Blade, Side View
34
CAMIA 73- - - - I'
i - 1V
: )
iI
' -1tn'I- -- Figure 25. Comb Blade, LetObiu
'J-4
Figure 26, Comb Blade, k;,.Iht Obl ique
35
I • • • •• |m m e • • m '1
I mrm w • =8 sm • m = a
VA Tb-' Ifin-f efit~ r
_ng vaue f LD) ha zprtdbyKoeg
Ileaditng e,06comb fied, ib irseita g Th tnhe
owlsI gIing- pez'for Manct..
2. 6h bastard wiJng -4 used. during -9 1 gf .h
3. Exept- for 'the tip egohcig udbec-
siOdered ,an- unbroen sur'face -du-rih§, the-§Idlng,
-phase.
4.The wing -sections In -he region of the ,comb are
-thin and-h'ave sharp leading edges.
5. The booked comb share with the exception of the
barbule oVerlojp In the base region can be fabri-
cated from a fiat sheet.
6. An Important parameter in constructin~g the comb
appears to be the 25 degree angle at whicei the
axis of the-barb radiates from the feather vane.
36
WWIWfiid'Ttifn ts
-Twa wind~ -tanneis wefre use& durt-ng -thklcourse of th It study. ?IstF
-ofthe~rkwasdone in he 3fttutjej teogi~ ot.th~eA~Ig
Dyniaiits, Laboratory. Thi's tunnel Js, an openh c~etiui -loised test sc
:tion- unilt capaOblof Velo6cities up -to- 30 ft/sec. The phys'ical layout
:-of 1hi tuine (Figure 27) cons ists; -faid tube eesin eti
-diamete -and. ten-f feet long. The ,airf-low is controlled~z by- aVaria'ble
speed DC. motor- instaJ Ied at the exit of the tunnel. The motor drives
a three-bladed propeller located behind adjustable guide vanes that re-
duce undesirable rotational velo~ities in the test section. Turbulence
reduction is accomplished by use of a removable honeycomb section placed,
- in the tunnel entrance,. The test section is viewed through a large
plexiglass window on the right wall of the tunnel and accessibility to
models is- afforded by a port on top. During this test the tunnel was
equipped with a conmmercially manufactured smoke generator called the
"Cloud M-aker" (Testing M~'chines lnc-. . This unit produces a thick white
smoke by evaporating mineral oil. A high pressure bottle of carbon
dioxide is used to force the smoke into a glass jar which acts as a
reservoir. From the reservoir the smoke is injected Into the tunnelthrough a rubber tube connected to a 3-foot section of 3/16"1 diameter
tunnel was used briefly. The tunnel belonged to the AF Aerospace Re-
search Laboratory (Figure 28). It consists of a closed section, open
circuit unit with a It x 11 square test cross section. Air enters the
tunnel through tea screens and is contracted in a nozzle before reaching
37
ZIA
iAll
N +j
383
-. GAKIAE/734
U 71
-- S I4
Li -
C -,
A. 'a..
- 4,
4 -- Ift -
-. Eu
4 IW p V
I."04J '4
- ~1 -~
4,
4I .~ C
I- I 0. I:I-
0 [3-4
I.
co(.4
t 0 V
a,..
lj~I
L 3A 4.J
0
I [iJii fLJ1iiKu
39
4,
£
RE1736 - ... - ~ ji7
,~..the. *eti-secion._ Ai f1bWt indui d , -~~te;4~a, ocii
- -stant sOOOd.roo.Veoiyi controlled -10otyt~:i adlusting- louivers ih tke 4'fan section-4. Smokei generated by evapotating, hianrokeh4susing an v
tiZ~~heatiuig Blin t Th so!e s in ted, Into the- tunnet thirough-7
scireehs anhd the.-high. contract ion ratio of' the noiflet, extemffely low-
-turbaleince-flow Js, produced.in-.the, test secti
Pictures-of the smoke flow patterns required a very high speed
film. The fil'm was Polaroid Type 57 with an ASA 3000 speed. The. camera
was a Graflex UX used in conjunction with a high-intensity-stroboscopic
light source. Sest results were obtained with camera settings of f 16
at 1/100 sec.
Models
were cuzl. from .065" aluminum sheet and h.d sharpened leading edges. The
planforrn and camber of the models are shown in Fi'gure 29. Model A was
a flat plate with an elliptically shapeO tip. Model 8 was similar in
planforni but had circular arc camber. Model C was cambered only on theI forward position of the wing and had a rectangular shaped tip. The
I models were supported in the tunnel by the sting system shown in Figure
127. To minimize interference effects of the mounting and provide an endJ.1 plate for the wing root, a 10" x 10" aluminum plate could be attached to
the sting at the wing root.
IHooked Comb. The best fabrication results were attained with .004"
brass shim stock, which Is easily cut Wth small scissors and can be shaped
40
- -
bd I *A
I Model, B
7.5,
" . R: 23"
-TM odel C J" .i.1 "
____ ____ 1__" ___- " , I' '
1
Fgr29. Wing Models
411-
using tweers fh first, desin ;(Figuee 30) was- 46~trce &by lyn
-Off 4 series of p~ral~Ie 1 lines oriented- at-a 25.,di! §ee a4ngle to the
-Ieadir.9g edge. The- blades' f1ormed. by- these-'l4nes- werte-then -given Sn'.
'el.1,1tic41 taper., The second desjgn rioretents an- aittempt, to induce-,a
hfook shape intto, the comb. To produce 'the hook, gi rtular- arcs are d,,iWn
from posit -ions speaif ied -in -the dimensiohs o~f F!g~re 31.The two cobs_
of eadh type, illustrate the diffder shapes thht dan bet btined by-
varitionof the specified diiiiensjons. Table. -I gives -the dlinersion!
of vite omb htwr used iktetss As the table shows, the mostf
in working with such small dimensions, arcs were drawn with a 4 x 0
Rapido''raph pen and the parallel lines were applied from dry transfer jsheets.
Procedure
Every effort wis made to keep the hooked comb and wing models as
simle s pss~le.in he revoussections a number of other devices
A' on the owl wing were identified which could contribute to the mechanism
that delays flow separation. Tw3 of the devices we~re considered for
-model testing. These were the bastard wing and the slott;ed wing tip.
The bastard wing was modeled with a small wiriglet cut from shim stock
which could be fastened at various positions and angles to the wing
surface with modeling chy. The slotted wing tip may also be signifl-
cant but was not studied aiere because of the difficulties involved In
I duplicating its aeroelastic bending and vibration characteristics.
Of the three types of flow visualization methods attemp~ted, only
one wz's successful. A slurry made with titanium dioxide and kerosene
42
GAJIAE/73 G,
~- 25*1
IA
H~r _ _ __ __ _ Taee1obMdl
~1 _ ____ A _ __43
07
* GAM{A/73-2V-
tT( * D
I- A 25
4,4
;Arda A, .,C b _... F 'k!+'173 '-d- 4p. -diD+,tsinssonn
3 7.5 0.3 08 0.05 - ,
3 7.5 0U . .05 7.5 0.3 0 8 0.05 - -
6 7.5 0.3 0.6 0.03 - Ys
.- 7** .7.5 1 0 1.1 0.03 0.38 0.31 " -
I I
*Model I has untapered blades..**Model 7 is consti'4cted using circular arcs.
45
mixed 'aiccording to the, instructions-of reference I5 did nC g~cp
*istieft jresults .because-,of the. low,-speed- range-.of' thit wind -tunelfs.
ITuft's ,appIlie&d irectly-to the uppe whg' surface gave. yp9 iformat-ion}
on flow "ehavIor above the spanwise vortex region- exsting behhnd- thje
c h. e use, of sinok was sucdetsful in, Ahoing the effect 'of ;the, comb'sf
on the, f low, f ielId when, the- experfirnen-a sequencOe was: II. Test each- wing nodel witbut -the- cqmb,4attached
arid record stal.llng angle of attack.2.- test winig with ciomb instal ied.
a. Check flow through the comb for spanwise
turning at high anqies of attack.
b. Check for stationary vortex sheet aft of
the comb.
.~Test wing with comb and bastard wing installed
~ ii and record stalling angle of 7~ttack.
46'
flowfiel6d, -at the low tunn T -spe fs the t 2igh 4
across th Ig.
-Part J
~The prel iminary low speed tests were conducted usihg. the- flat plate
wing (A):. The first Ifive comb models were glued, to the 'Ugpper surfAce of
the wing wi th- the comb 414des extend ing forward- over- the stharpqned- -leidi-
Ing edge. The. ow tujinl speed of 5 ft/sec Was used to obtain coherentA
smoke-patterns so that detailed streamline behavior cduld 'be observed.41
Smoke flow over the wing prior to fitting the combos (Figure 34) shtwed
that the streamlines proceed directly chordwise. With the combs ingtal-
led, the flow pattern changes noticeably. At low angles of attack,
(Figure 33) the flow through the comb continues chordwise but exhibits
considerable stagnation on the wing surface Indicat~ing some reduction In
flow velocity. At high angles of attack near 30 degrees (Figure 34),
the comb tur-ns the flow spanwise toward the wing tip. If the comb ex-
tonds k', the wing tip, the flow through tne comb becomes entcained in a
stationary vortex core that moves laterally across the wing and rolls up
in the trailing tip vortex (Figure 35). Placing the-wirnglet at mid-span
(Figure 36) destroyed the vortex. Instead of moving spanwise the flciv,
moved into the region just under the winglet and stagnated. This was
likely caused by the winglet Inducing attached flow in this area while
47
-# esofte fjq'.- on the. upper wing strface 4-maine fql ty, separaited
1 hcreaslhj. tunnel velod1:t- did' no use -attacdhedj fow at ahg4~ o
attack -above- the pl'ain *iig stall. 1 lst~a. lasclvrtxsreWai-periodically she4 dr~Ee ~dneg tigr 7. aItIp
'cnibito ndw gl:ienitit ion cad6used -nof changes n :h: ,observed,
OrelitiNary tests a tdmb wth- uhta~pered ,bl ades 'Was t riid. Hstifibon-
ary-vortex was createid'by this con figuration aid Tt wa s discont lnued.
Part -2
None of the comb model's wiere successful irk Increasing the stallin
angle of attack of the flat plate wing (-A) or' the ful ly cambered Wing
(8) Using tunnel speeds up to the maximum ava;ilable o 30;aft/slec. At
tthis point-in the experimenht, hew insi,§ht into the theory of the comb
operation suggested two changes which proved successful. The first was
to more closely madel the bionic geometric ratio of comb blade length
to wing chord. Wing C in combinatio. with comb 6 or 7 reduced this
ratio to 2.5% which approximated more closely the value of 1.8% measured
on the owl wing. Second, a sharper leading edge was ma4e by fastening
the comb to the bottom surface of the wing. This allowed the .004" shim
stock at the base of the comb to become the wing'leading edge. The tests
I of these wing/comb combinations proved successful in delaying flow sepa-
ration to a marked degree. No difference In performance could be de-
4 duced between the two comb models even though they differed greatly In
shape and solidity. At a tunnel speed of 15 ft/sec, the stalling angle
of attack was delayed from 22 degrees to 30 degrees. Stall could be
4 48
frvward - fro the n
Vitig. peforz~anc was ved sesitiv tf ches~,,ti i atr igse
Cbh§".s im root 46ntin angle -With the- wing qurface, seemd-to hav& much
Ai close-study of the smoke -flow pictures gave toisiddrable ibsight
Into the flow m-nechaism hrestponsib~e, for creatihg, t~e hi§gh iJft bdha~jor.,
4 Pictures of' the wing without -the c~irb installe'd establ ished -the s tall-r
ing angle of attack to be near 22 degrees (Figure 38). The. stall- -was4characteristically abrupt with no sign of transition of -oscillation.
At atql et -of attack -we)+' above the -stallI (Figure 39), the -flo'w was full1Y
separated at the'leadinig edge. Instal-ling the comb produced no chan%,?s
'in flow behavior on the wing at angles of attack below 22- degrees. As
sepratonwas evident at the wingtringee(Fues4-1.A
angl ofattack Increased above 22 dges h lwfrtbcm
stroglyreattached at the trailing edge (Figure 42) and then the vortex
region at the leading edge began to grow progressively larger (Figures
43-44). Movitng the smoke probe toward the wingtip (Figure 4j5) showed
the vortex region was now smaller in size and of greater intensity. InJ
the regions where the vortex was strong an apparent stagnation point
was formed by the smcke flow impinging on the wing surface directly
behind the vortex region. As the wing approached the stalling angle of
a ttpck (Figure 46), separation was evident at the trailing edge. Stall
occurred abruptly (Figure 47) much In the same mannaer a3 the plain wing.
The nature of the vortex core vi;-s examined by injecting smoke into the
49
______________ V ---
the Io~e-a~hevorexran o~tlins. tspo 'tioni on tbe::wi. lci the ba tard
wlhin - Cile ibibdel- tee jtbj :kh d the vortex, :fOcw' fted at anoe -ot
t~34z~si, gI, seperatloiit svid it
at the tai'ling edejust -ptli to, stal
At the- cohcl us ion- Of th-xenew alowwngastsedJ
tesoetunnel tefooac qpqh that -the mdel, ~A
voiil r-tex reg-ion a ob serVedd or the owl-wing, and' tho -stalITng angle,6
of attackc agreed within one .degree with- that -of the iboel.
50
- 6 E
77--J-713- --T7
4*4
i-F
Figure 32. Smoke Flow on Basic Wi|ng Model A
a 15% V 5 ft/sec
H
I,'
Figure 33. Wing A Fitted with Comb C50 V -5 ft/sec
S-Figure 34. Spanw Flow
Fiur 35, Spnws Vortex,
'I7!
3' V
52o
9" Il 71 - C. ,
V
< :.- - " ",
- '-,- ; __.___ __ __ _
, , 5 t/e
- C
3 Ii
I Iml l lI I! I I I w
I I I I
W$AE/73-
-
Figure 36. Low Pressure Region Under Bastard Winga 30", 5/ ft/sec
HFigure 37., Vortex Street Shed from Wing Leading Edge
a , 20Q, V - 15 Ft/sec
53
, , , ,-
:,.'Figure 38. Incipient Stall on Plain Wing Model C "
: ' : .a 22', V = 15 f~t/sec .
5..4
-
"Figure 38. inciien Sepaalleo Fla ig oe
._ c =27, V =15 ft/sec
., I!
; Figure 40. Attached Flow with Comb 6 Installeda 15% V 15 ft/sec
I II
~1
Figure 41. Flow Separation 'vident at Trailing Edge
a- 19%, V - 15 f-t/sec
.5
S GAIAE/73-6--
bes i
•v~a iro
F 4r
EllKI
- jr,- -;-)
~Fiure 42. Flow Fully Attached Above Plain Wing Stalla - 23, V - 15 ft/sec
I5
I
' Figure 43. Leading Edge Separation Region Evident.. = 24 , V 1, 5 f te isec
56
L Figure 44., Large Separation Region at 60%j Span! := =27°% V =15 ft/sec
. S
' P
Figure 4t5. Smafl Separation Region at 75 ; Span
,a = 27, V = 15 f./sec
57
-. -- Pt -
I II
1 1!
* Figure 467. Trullyn deparae rFlo
= 30o 'V = 15 ft/sec
58.
iLi
"-:g 1 o:.*\:
10
Iq
I ii
Figure 48. Vorterd Cire Iuntalled-oweaahsS= 300, V = 15 ft/sec
i 5i
l II I I I I I I
1 I ! 11111Ill I H I !11 1 l! I III II I I4N . -I I i ! 1 II I !
f GAI/At/f73-6
'4 *
-4 - -- :.
F-- igur'e 50. Approaching Stall with Bastard Wing InstalledI
-I
4 -.
I- -
am 34% V 15 ft/sec
Figure 51. Fully Separated "a /0*, V = 15 .',.'
6o
'.Y 'A
th& :flow -f itld !R-e6n a Winj qdipPedwt pnls gwn
Pears t, to, be the same; -as. t44t prodijc~d by the conib. Anair jet a&s i'-
stled-at et Ileading edge -of -w!ing C. Thei jeat axis was, - rlinted span
wise aind phfotographs of' the 'flow were ihade with and -without bl~owing at
an~angle, of attc-well1 above normal- wing, stall. 'The testwas coqduc-
-ted -in -the ML.whid -tunniel which cdul'd',OproVide, 4, irutp-e 016kt; stre!ams-.
Figure 52 shoWs -the -wing fully- sita~led. (he :aifffiost o bscures -the-
j steamlnes n th~ uper lft corhe -of' th6:plcture) n Figure- 53'the
blowingj Is -on and, tevortex region -is cle0arly outlined by the smoke
streiams. No further tests were runas excellent p Ictures ,of spanwise
Lblowing were Av4Ialle for furthear tudy In referencles 16, 17 and 18i
r - Theory
Vortex Lf.It appears that the comb generates a spanwise leiading
edge vortex sheet which leads to the high lift characteristics observed
In the flow visualization experiments. This vortex sheet !s c')mposed of
leading-edge-separated vortex filaments that are formed when the flow~ Is
unable to negotiate the sharp leading edge of the airf.>il and separates.
Unswept wings with sharp leading edges shed this vorticity as a periodic
vortex street similar to that observed behind bluff bodies. Figure 37
is an excellent example of the vortex street generated by the test wing
before the comb was perfected, As the airfoil Is swept~ back in the flow
field, the flow changes character. The vortices do not periodically
shed from the wing surface but roll up into what is termed a vortext
sheet. Figure 54 shows a diagram from Mlaltby 019) of the vortex sheet
produced by a delta wing operating at an angle of attack. In the sketch,
the leading-edge-separated vortex distribution Is represented by dikcrete
61
~~GA1i/AE/73-6H>El
7'ii ill~:.1 I '-iii I -AJ~~14 ., ~1I4
4 .-
-- -.--. !' 2
Figure 52 Separated Flow, BIow~ng Off~-35,Vm25 ft/sec
U *
I I
44 I
-.
'4
-- ~ ~-
-- , i4 4
-~
40
'4 5A -~
C Cl
Figure 53. Attached flow, Blowing Onaw350 ,V.w25 ft/sec
I62
'13
LEDNGEGSEPAATIO SECNDAR FLO
Fiue5. otxSeto et ig(9
i 6-
fi-laments -bginning at the edge.1 an r ng p toforma sur-9 theading d
face, This surface +isthe so6-called vortex, het aird is en to en-
close a core of rotating fluid that grows in size add- strength as: it
flows,-streamwse along the upper wing Surface. The presence of this
saionary vortex sheet on the upper wing Surface-creates .a -low- prest-
sre region of .separated flow 'under the outer Flow field. Experiments
have.shown that the outer flow field flows.-smothly over the -sepaated-
k c+ia i n.M reaches to ne wi-g surface just behind. the vortex sheet
(20). The lift forces produced by a delta wing are characterized by a
non-linear variation with angle of attack. Figure 55 shows typical lift
; , .,r wings with aspect ratios of .5, 1.0, and 2.0. si " Zree
cases the slope of 1he lift curve increases with angle of attack until
CLma is reached, Following Polhamus (21), the lift forces are separa-
ted Into two components. The first is the conventional linear lift ob-
tained by a solution of the ideal flow about the airfoil. The second
component is termed non-linear or vortex lift and Is identified by Pol-
hamus as the force resulting from the low pressure field created on the
wing surface in the area of the separated flow containing the vortex
sheet. Comparison of the increments of lift produced by each component
shows two trends. The first Is the Increase in vortex lift as angle of
attack Is increased. This can be attributed to the increased vortex
shedding which produces lower pressures in the separated region. The
second trend is the increase in the ratio of vortex lift to linear lift
as the aspect ratio decreases. This is explained by noting that lower-
ing the aspect ratio of the delta wing implies greater sweep which in-
creases the fraction of wing area that is in contact with the vortex
sheet. Stlt does not usually occur on the delta or swept wing in the
64
ifl
I ifl
IN ODGct 10
-I0 G \.65
GAO/AE73-6
S normal Maner. Thei decrease in -CL after reSWhhig q,' -s typkial~lyJgradual_ and _s accopanied by a large increase 'in -the .drag- f9re. The
mechin'isrn involved -is thought to -be a bre akdowni f te ordered f-Iow in
the- vortex refion that occurs first at the wing trail ing egand tr-avels
forward as, the anqileof attack is increased above C, (20). The wdl.
-established, flow-pattern observed or. the delta wing. appears -to resemble
that fouhd on the unswept wng equipped with the'eadkng ,,edge comp.
The three elements that al'low the vortex sheet to .fdrunon the delta wingj
are seen as: (a) the existence of leading-edge-seParated vortices, (b)
a spanwise velocity component, and (c) a favorable pressure gradient on,
the wing surface to prevent vortex breakdown or bursting. In compari-
son, the experiments with the unswept wing/comb combinatioh show that: j(a) leading-edge-separated vortices exist at the sharp leading edge of
the wing, (b) spanwise velocities are produced at the wing leading edge
by the mechanical turning action of the cascade-like comb, and (c) a i
favorable spanwise pressure gradient exists at high angles of attack
because of the vortex sheet created by the flow around the sharp wing 4
tip region. It should be noted that the existence of the tip vortex
alone can contrioute sizable non-linear lift Increments--especially on
rectangular wings of very low aspect ratio. Reference 14 shows this
increment to be expressed as,
ACL U C1 5
where C1 represents a coefficient that is based on aspect ratio and
planform. A graph of this effect is shown in Figure 56.
Spanwise Blowinq. An excellent analogy exists between the flow
field produced by the hooked comb and the effect of spanwise blowing
66
-icros an unswept rectahgular -4,
wing. Dixon (16) and Others- -It . -:7 .
-I- conducted spanidse bAoig, . -. .1tests on a wing similar to 4
those tested in this study CIO - -V
and. reports flow patterns INS
closely 'reseM~blng -those.
found In the-comb t' ts. Abu
'Dixon's model was a flat I , -
1 2 3 4 5 6plate with a sharpened A
leading edge and had an Figure 56. Non-linear Wing-tip Liftfor -Several Planforms (14)
air jet locatrd at the A
quarter chord position of
the wing root oriented to blow spanwise across the wing. Figure 57
shows a cross section sketch of the streamlines created by the jet at
an angle of attack well above the normal flat plate stall. The unique
characteristic o this two-dimensional view is the apparent stagnation
point that forms on the top of the wing surface just behind the leading
eege vortex sheet. When the overall three-dimensiona flow field is
considered, this point represents a line of division between the flow
entrained in the vortex and tite exterior field. Figure 58 shows the
result tng streamlines on the upper wing surface obtained by Dixon using
a slurry for flow visualization. The "herring bone" appearance of the
streamlines near the leading edge is like that obtained in similar tests
of delta wings (19). While the line of division between the two types
of flow is not strictly a stagnaton line, there is no streamiise com-
ponent of velocity ;nd the spanwise velocity is relatively smail. This
67
analogy ofesan ex planatison
of the unusual -1line of stag-3
nation reported by KIoteger
on the owl wing (6).
Dixon's -experiments
give some ,Idctof the
non-lVinear lift values
acieedbyspnwsebleFigure 57. Two-Dimensional Stream-
ing. Figure 59 shows the lines with SpanwiseBlowing 016)
lift curves obtained for
vaiu degrees ofblowing., h blowing cefintisdeidas
iet
q S
where S represents the wing surface area. To represent the Increment
*of lift actually produced by the spanwise blowing, an additional curve
has been superimposed to show the non-linear lift due to tip
This was comnputed using the method of reference 14.
Bastard Hing. The comb exper'ment also considered the effects
of the bastard wing on the flow field as It was previously assumed to
be working in conjunction with the comb, Including a small winglet at
mitd-span did result in higher values of stalling angle of attack in
the smoke visualization experiments. This was unusual performance if
the winglet is considered to be functioning only as a conventional
leading edge slat, In addition to the slat effect, however, a vortex
f-ield is produced by the bastard wing. This Is shown In Figure 60.
The forward sweep of the winglet produces a leading edge vortex sheet
68
dAWIAE/13-4
whbse sense of rvtation. is -
counter-clockwise ooking
upstream. The conventional.tra'111hg-edge vortex system Ai
Is seen to have-the opposite Jet
sense. Considering its - /F-small relative size, It is.
possible that the bastardFigure 58. Three-Dimensional Surface
wing operates in conjunc- Flow Marked with Slurry (16)
tion with its vortex field .to produce a spanwise fence to oppose the propagation of separated
flow from the root area of the main wing. At any rate, the unusual
counter-rotating flbw field described by Kroeger (6) is likely due to
the presence of the bastard wing.
1'AA
44
........
.4OA
Figure 59. Lift Curves Obtained for VariousDegrees of Blowing (16)
69
I GAI{AE/734-6-
i rai jins Edge
Iire
j
Figure 60. Vortex Sheets Generated by the Bastard Wing
70i
VII. Cohclusions and Reconmendatlons
The results, of the study of the owl wing and the iexperiments with
the hooked comb sug4gest the following conclusions:
1 1. The hooked comb present on the owl wing works In conjunc-
tion with a sharp leading edge to produce non-linear lift
on the outer half of the wing.
2. A working model of the comb system can be made from sheet
metal. The critical parameters In the comb construction
are thought to be:
a. A comb blade length to wing chord ratio near
b. Slanting the comb blades at a 25 degree angle
relative to the wing leading edge. This slant
Is toward the wingtip.
c. The individual blades must be tapered.
3. The comb turns the flow only at high angles of attack. At
flight angles of attack, the comb prec'ents an extreme.y
small drag profile.
4. The small relative si.a of the comb (1.8%) makes It
attractive for use as an aircraft high lift device.
It Is specifically suited for aircraft possessing un-
swept wings with sharp leading edges..
5. The flow field created by spanwise blowing Is very
similar to the flow field created by the hooked comb.
6. The bastard wing, extended on the upper surface of the
owl wing during gliding flight, may act as a fence to
71
GAA E n4-r
4elay outward pro.agation of f lo 6Searation-fro.i.
thick and highl.y camberedj wing root sectibois.
7. Potential applications of the comb's principle ofoperation ca. be-envisioned in the fields of fluidics
and cotabustion'ochamber design. Severar fluidic de-
vices already developed: use separated flow to alter
the performance of a nozzle. In turbojet engines,
shorter combustion chamber design hinges in part on
improved means of combining the air/fuel mixture. A
The hooked comb deserves consideration in both these
applications as it appears to bethe first unpowered
device reported in the literature that creates a
* stationary vortex sheet oriented at right angles to
the direction of flow.
The following recommendations are:made:
I. Further wind tunnel tests should be conducted using a A
hooked comb model to measure aerodynamic forces and
refine the concept of non-I infar lift.
2. Larger models should be considered for any new wind
tunnel tests as the small relative size of the comb
makes accur-te fabrication difficult.
72
: II, - -mi. -R.R. The Si ient' F-lght .6 00|-, Journtl 6Vth.,. .al'M roha-,iial 'sodiit, .3 - '- -
Arnu-tca ' 8:. 83,'1;843 (-3)
2. aspet, August. "Biophysics of 'Bird Flight . sciLnci,, _132: 191-199-S(4Uly. !96 ). -... . _ . ... ..(qu ' 5d3, S1"ick, E.F. "Bird Aerodynamics". Soaring, 36: 26-31 (June 1972.
4; Soderman, Paul T. Aerodybami Effects of _ad___Ede Serrations@T'UngLg _X,____
Wa nto: aioalAepati-cs and Space Adi hi t ti f 6Sipte-mfer 1972.
5. Hertel, Heinrich. Structure-Form-Movement. New York: Reinhold Pub-lishing Corp., 1966.
6. Kroeger, R.A., et al. Low Speed Aerodynamics for Ultra Quiet Fl I ght.AFFDL-TR-71-75. Wright-Patterson AFB, Ohio: AF Flight Dynamics A
Laboratory, March 1972.
7. Shestakova, G.S. The Problem of-the Penetrability of Bird's.Wings.Royal Aircraft Establishment Library Translation ll : Farnborough:M inlstry of Aviation, December 1965. AD 804465.
8. Vinogradov, I.N. The Aerodynamics of Soaring Bird FlJ t. Royal Air-Royal Aircraft Establishment Library Translation 8. Farnborough:Ministry of Aviation, January 1960. AD 234631.
9. Gerardin, L. Bionics. New York: McGraw-Hill Book Company, 1968.
10. Kramer, M.O. "Boundary-Layer Stabilization by Distributed Damping."Journal of Aeronautical Science, 24: 459-460 (June 1957).
11. Spark, Johq. Owls: Their Natural and Unnatural History. New York:Taplinger Publishing Company, 1970.
12. Taylor, John W., Editor. Jane's All the World's Aircraft. NewYork: McGraw-Hill Publishing Company, 1970.
13o Jacobs, E.N. The Characteristics of 78 Related Airfoil SectionsFrom Tests in the Variable Density Wind Tunnel. NACA Report No.
Washington: National Advisory Committee for Aeronautics, 1935.
14. Nicolal, Leland M. Design of Airlift Vehicles. Colorado Springs:United States Air Force Academy, July 1972.
15, Maltby, R.L. and Keating, R.F.A. Flow Visualization In Low-SpeedWlind Tunnels. RAE Technical Report Aero 2715. London: Royal Air-craft Establishment, 1960. AD 245348.
73
GAgA/73-
16. Dixon, '.J,. If U ~ na~c -b Laer l Bowtn,4 Ove~r 4-1Lift'ing-of r ace. iM Pan e ab S6- 3 a h i g 6 : $mrca 4 s t a e~~naiztic~ ad Atona ut tcs, bur 99
1'7. Werele , Hs and GallTon, 11. ontrole dtcoulementi par Jet Trnvrsa10.Nctsk.A. Q 170. ChatT 16h: 0 ficii Nitional dftudesipt itdeReheches Aerosp.atia4es, -1-972.
18. Cae 11l, t. and Camarasescu, N.. NewResearches on Smal 1'. ab±Chord iRat-io Wings wit La te ral1 Jets. FTD' Transl)a t ion iC-23 -319 -71' -Wright;--Pdtte~rson 01ITh io. oreigni -R.chnology Division, -USAF, 191,.19. Mla 1tby., As L. "The D0evelopment 6f' the Slehdier Delta COnicept!'. Air-
craft' Engirneering, XL: Zl7 ach16)
20. Chang, Paul K. Separation of Flow. New York; Peegemon Press, 1970.-
21. Polhamus,, E.C. AConcept of the Vortex Lift of Sharp!-Edqe Delta 4Wings Baseid -on a Leading EdeScti Anlg. NASATND37Washington: National. Aeronautics and' Space Administration,December 1966.
74I
AMAEJ34 -I -7
9 Vfta I?OdGereV iar5-,Xt.eso wasorn on. 26 NRovember 1,041 'in GreenhVIIe
Pennsylvaniai. He graduated. fo ytinn.Joint ;High School, in, 4ames-
towni Pennylvania in -1959- and- eniIst;ted, in the Air Force -the iaO6 ea'
Af ter 06petn9biisc training,, he was ass-igned to-~the- United, tat~s,
Naval Acadey reatoy School .a-t- Bainbrldge aval 'training tenter,
MIaryland. In 16, hdehetered- th-Unlited States- Aler o-:Academy andI
Igraduated in 1964 with a Bachelor of Science degree and a cominissibn as4
a Second Lieutenant in the Air Force. After graduating from pilot trainm
IIng at Reese AFB, Texas in 1965, he was assilgned to the 29th M II tary Alir-I
lift Squadron at McGuire AFB, N.J. as a C-iO pilot. In 1967-68, he
*served with the 362nd Tactical Electronic Warfare Squadron at Pleiku, RYN
as an RC-47 pilot. From 1968 until entering AFIT in 1971, hie served with
I the 30th Military Airlift Squadron at M~cGuire AFB, N.J. as a C-141A
instuctr plotandflight examiner.
Permament address: Liberty Street
Jamestown, Pennsylvania, 16146
I4This thesis was typed by Katherine Randall
I7