GA Airfoils
GA Airfoils
GA Airfoils
9TtYL€ V9
ssn Nor,tvr^v lvusNs9 uo,!I sTro,lurv Jo goTvrvc v
SIFOFIV VO
113IqTU ') L86T 1q6rr,{doC
^rfpH
.sTroJ,rrp prTE^ , alp suorf.rpa
Foo6 TTrls
snoT^aJd ffP lo sTTofjrrp aq,l .ubfsap lJe.rcJrp rreq:l raq:Iel u6rsep froI.rTE Jo uJaJ
-aq]
-uoc E {frrEurrJd sr sn{l .ro] , pezrururu ua€q sEq , pTftluol: 6e:p uq.f.l Jo uorlp^T
-.rap aq] pue ,6e:p uryrf ]o uorssrcsrp aqJ - 1)ra1 arfJ c, suorsrna: 1e:aua6 sapnTcur pup
'sfroJrTp (xx9-) p€raquJpc-q 6rq ar11 Jo uorlrppp ar{l qlr,.lrt ,ute6e 1ar( xTl}au Irolrrp !9
arl+ spupdxa uort.TpJ q:lxTs sfw . xfxlplu arIl 01 sfTol:Jrp a.rcrl pappe suorlr$ luanbas
-qns 'sTrof-rre gg dluo paurpfuo3 ,196T qc:q^t ,,,sfTol:rw !9,, l:o uor?Tps fs_rrd aqJ
'1ou alP sauPTdrrp peJ-d..od TE3rd f fnq
, stuanD! 6urqclrd qSTq
fo 1up-rafo1 ere saueldlreg 'sluaicru 6tmrceTd q6rq {1aar1e1a:
a:npo:d osle .{arp fnq '.r€qrec Jo luncrup lspaf aql qlr,.tr lfTI lscrn aLIf acnpo:d s1ro3
-.rTe pepEoT-pTl^l 'slroJ-rTe PePBoT-PIur rl:tr!\ f=4feq op saueldlres ssa.IErqA 'sluanD! 6uT
-r{clrd loT ecnpo:d ol p€peol p-rpA-rol fTe arp sTrol:Tp 99 'saueldl'res :o; aler:do:dde
lou trfqeqo:d osTp are f,aq1 'slro;::re luannr: 6urqclrd o:ez fETceds arrnber qcTqA Jo
qaoq . saueld:'ge 6ur,r.r 6ur.{1; :o sraldo:r1aq :o; aler:do:dde lou aJp sTTo}.rre \19
.
f JPrJf TP
pareaod JoJ sp6eu q9 g:o ,{1r:o[an et|f .:aAoJ pfnoqs xr.rlai attrl '6T-III arn6r] uo ur'.oqs
( E.9T9-0E \i9 pup , S'tT9-08 v9) aJdu Z snfd 'sTroJ.rrp 96 sr.lrE?uo3 xTrlEu aqf snql,
'.rdquec alerrdoJddp at{t EuTppp uaqf
, ssaulJTr{f pe-rrsap a({} crf u}rop :o dn uorSnql-Tlsrp ssauj{JTt{l atlf,
burTpJs Aq palelcdrelur aq .{an a6trer srt41 url;}r.'r sassau{JTql raqlo
'rp6uaT p.roq3 aq1 lo a6efuacrad P sE ' I3rq+ SgT puP 'tET 'XZT 'G@T -
' sr{f,ua1
ur fuarsr]l:s3 IJTT
ubtsap alanrrxo:dde aq1 6uy1eclpu1 ')(x9- pue 'xxt- ')o(E- 'xxz- 'sTaAaT J€qrPJ ?-
'(dsno ou) noftc< gre ' (ad ? dsn3) :o<x--xx 'Fffll -
'fiol€q 1)cal arFl uT passnssTp
suosEe.r roJ 'pasn aq 1ou pTnor-{s sTTo]rrp 1T6lf}-S \D\iN aqtr ',{larrrlcadsa:
'sfroJ:rE sar-res 99 pue 's€tJas t9 'sarJas e9 '1T6Ip-t \pVN eqf aJPTd
-aJ pue urcr; p€^Trap are .derL1 'p:oqc ]o luac .rad ur 'ssar{Jrql froJ.lTe
unmxEu Jo uorlrsod aq+ ol .r-a]3.r s:squJnu asaql ' 0t pue ' tt ' St ' 0E \19 'sar-:aa t -
I9TwE S -a1du:ere I :lo slsTsuGl )rlrlsJ rTof-rTE ,,v9,, aql
.,,aso uoTleT^v TE.rauac Jol: sTroJrTV 1o 6o1v1e;r e
-sTro].rTv 1l9,, sT lrnsal aql 'esn atreld:re TpnfsE :oJ 6urpue'fslno 1nq ' aTqs+rns uEq|
5ur1ar ,{1uo lou Jo asodrnd aql :o; sfroJJTe tDv!.I e(I} ,,dn ura1c,, ' spoltrta,l \l'Jvll aql
uD.rl: :eJ oo1 bur.{p.Ils 1ou aTTqA 'pup '{:o!r agl ur suorssruD puP salElsru ro[EUJ a({l
ssaf,ppe 01 p€ATosa: J 'acpfd oluT fTP] of u€6€q s6urtll rr€qf pue ',{af 6usl 17.Jvtrt :oJ
6urlJo,*r uaql spa oqr'r reeur6rre-:crFErluG) P Iq urrlEqrel au o? palonb seA aACCIe luau
aJaql letIl azrr€ef ol
-alsls egl '{:o{ TToJfrp vJvN alIl ur suorssruD puP $rPTf arar\ 3ruos-qns
ue@ 1 arrf agl lnoqe 'ltpl: uI 's,0E6T aql fo {-ro& TroJ-rr€ \DvN alF} fo
uorfe6rlsanur snor.ras trur 6uruut6eq ralllP os :o rsa,{ p '886I punorP srql azrrs€J o?
a rgc J 'enll ssaTa{Fauou sT 1I 'eq ol reedde .{eu: lua:e1e1s stldl se afqe^arfaqun
',,aJueurro]:ad fToJ-rrE uo s:alanre:ed fTolJTP snor.rE^
6ur6uEqc l:o slael:l:e aql a?e6Tlsalur 01 pesn ax€.r fRIl srranncads ls31 Tauun? pur^rr
{1a:ran a.::a.l anq 'asn aueld:te TenFe rol peu6lsap fou a.rarv\ sf rol:.rre \DvN aqJ,,
uoTlJnpo.TluT
il
zeT .lorFlnv arr] ?noqv
tET ,. slTX :OlEraua9 xalJo^,. I -raqutni.l
eeT ,,p€TlTTdxTs -,{1t1tqv15 lJe:cJTv,. I r€qur"lN
8Zr 966I/S/Z ,SSIN of ?aaTqTu'H 'ra11a.I 9 JaquInN
t(.7 ,,TTofrw ETI0-(I)JnN vsldN er{l ;o anbrltrg q" g r€qujnN
5TT ,,g-og aql :oJ aTrJo.rd--ag 6ur14 are;:ns reddn uv,, ? FquInN
r0T ,,TTofrTv dTotz VJVN 3t{? lo TTPIS d:eqs atn 6unue;,, e _reqrnN
66 ,,s6ur.g paredeg, ro] seloN u6tsaq,, z raqulnN
e5 ,. sT ToJ:W 6utpuY1sJ-aPu1,, T -r€qurnN
Effi'
t.L suorl€s 6uT.M Jo ssrlsrJaFPrEqc crupuipo:av AI xrolEddv
saTqeJ alPuTpfo rToffT-v III xlotlSddv
'E
z9 sauTl ueat{ II xlq}Eddv
6n sruod ssaulcrrdl sTsEg I xICIMIddv
8r-87 1E-1 sa:n6rg
LI sa3uaje]au ]o 1sT1
9T ldr-rJslsod
.{pn1s asec 'saorlap a6pa Sutpeal 'saupTd:Te
€3eJ.rns-€r{f ' spjptrpc 'sbur/'! 6urtrTl 'se6pe 6urTrErf lunfg
spa[qng peleTaa '6
'T JtTto-(T)srN'J9TZ0-(I).{-Si'T+*n
z1 sTrolrrv lsvN sn€ueTra3sTw '8
ZI s3:g:l a6po 6urpeal 95r ' 4
OT
.tce:nr:g rllPf6o-rd -ra?nArrca ' g
uorfJT-pe-rd ecuEuriro].rad l]:Ef3-Iw 'g
OT
faqunu spTou^au I ssalr){Jrq} ' reqljeJ 'sdsnJ 'sarJes lo ?€IlS
I suortEJrTddv pup uorf3alas ffolrlv vD 't
v luandoTalac Trol:-rrv vc J E
sTroJ.rre sar='s 9 'sf toJ.rre 1T6Ip-9 ' sf rol-rrE 1!6lp-t
)t_:otq lsatr, rTofrlv !?vN Jr-rolsrH '7.
T
I ,{-raurng 'T
'oN aE-Ee
SJNIJNOC CO g]gVT
ilt
'uJa11d rPratla6 allPs srt8 enoTTol: ,,sTroJrw \19,, 'slTnsaf lsa1 Tatrunl Purr|r
aqf sfuasa.rd AI r.rpu€ddv puP ' sTro]rTe poJ€quJ€J atlt rof saTcIP:l alPurpJo al{f surE/luo3
III xTprraddv / (salrrofd reqllPc ) seuTr Lrseu aql sursauo3 II r|pJaddv 'su:o] ssau{crtff
clsPq aq+ surs+uo3 I xT-puaddv : uratr1d sTtI+ s'4oTTof puE {Jo& 1sa1 TrofJtE vJvN al+l Jo
{-rpurms e sr T eJue.roJau 'fro]:Te p€r€q)lec p urslqo 01 ' ,,auTT Lrpau, aq+ sp u l0ll){ aTn
-peq.s.r€qlpc pau6rsap tr11n9:a:e: p t1trlr ,,uro; ss€${orlI} 3rs€q,, srql burA:nJ uaq} pue
, uorlnqrgs1p ssau:{Jrrrl
fecrr+anur,{s e 6unrnr:ra1ap fo po+tar 3rsspTc aq1 ,{q peu6rsap
sE.i1sTToJ.ITe?sa1JosaT.Iasels?sa1cTfEuIafstrsasaq1.Iodm
'[Jtrro-(T)JT{ pup'Jsreo
-(T)gEr '(ST'O-(T)SI pue Lf?0-(T)fI] a-+rvp puP T-1,!d9 \6rdN aq+ epnf3ur saldueq 'uer11
rucr!: sfrofJTe aTgesn a{eu o? uorlEcrfTpon a^TsualxE} arrnba: UEI+I Jo TTE ptrE 'slfnsaf
:ood peq TTp €^eq asaql 'sae{o1&e vsrat fenprnTpur uolJa leuosred e .{TTensn 'anrrl
ol aurl uD-rJ sTroJJrE pelpfosr urslr€J paqsrlqnd^qseq \6rdn 'srpe^d ?uea€-r eJcru uI
's-lTnsa.r paxTur rI+T,^r 'saueTd:Te go a6ue: aprA E uo p€sn
eq 01 ares sTroJ:re 1sa1 al.FI 'esn ll€.rJnP :o;: peu6tsap F11ecr3:rcads sTrofrrP Jo
s6o1e1ec qlrr1 l:orr ?,sa1 srln dn r,bllolr prp re^eu \f:)\d'I{ ecurs 'ra^a..1oq ' .{f aleuna;roliun
.aarssa:d:rr sr palr€fToJ sea lpr{] sfpp lsa1 fo funque aql pue 'Illulr{ q6nol$ penurluoo
I.rcr^r 1sa1 snonp:e sTqt 'sTroJ.rrE saTras-g puP '1T6Tp E '1T6rp ? aq+ a.ran srroJ.rre
lsa1' asarp Jo Lr ronl lseq aql 'acueu:ogr:ad froJ-rrE uo ueJJe aqf auruuelap 01 sJelai
-ered 1ro;;rre ursfJe3 6ut{:prl 'uJp.r6oJd lsaa TroJ:Tp c1lare1strs e dola.tap o1 spn dafs
l:lau atll '6ur1sa1 uppuef arn?eu .{q se/'r srqL 'q3fpas .{Eolouqcal E Fnpuoc 01 fnq 'sTau
-un1 arlt a?'pJqrfpc 01 1ou 'pT:olr eq1 re^o Tfp uD:l: STToJJTe unou:{ burlsal ue@
^fuo
\Dvl{ pue 'auTT-uo sea,{:o1e:oqeT srqa 's,0e6l-pTu aqf ,{S '(.Jvl,n) A:o1e-:oqr1 fErrlnEuo
-rav fprror€t{ Fa16ur1 ar{l 'E'rur6Jr 'IaT6usI le sarlrTroeJ lsal Tarrunt pura rrr€Fo!
l:o uorpulsuoo aql set'l lJolfe qc:easar \iJvN sTt{+ 01 TPf,.trIaJ ',,dn q31eo,, oe a6pe1
--Jtto(I){ sTr{] asn prnoJ.{rfsnpur uorlET^e 'S'O ?uEJuT erp ftql sadoq aql t{lTA 'Ip.rs€sa.r
.
fecrlnpuorae JTsEq frnpuoJ 01 u;p.r6o.rd P pun; 01 PePTcaP ssa:6uo3 'uossa.r srLFI .rod
satrrSrra pue saueld:re ecua:.r:o;rad q6fq fo a6ue.r aprA e pa1ca3:rad peq sarr+unq" raqlo
alrqr'r 'aur6ua,{f-reqq aqf pup r reure.rl T-Nf sll.mJ aql ua€q peq ItI4 urs'n aqt
Aq suorlnq-lrf,uoJ TPJT+nsuorae luecr;:ru6rs ,{1uo aq1 'lJEf uT ! fualdoTanep uorlPrnE ur
sar.rlunoc urapcru rarf+o purq=q pa66eT -s'n ewl lEt{} rHoqs pPq u,!M uoTlErAE
's'o ?uefuT ar{f azrprsqns ,{rlceJtpur 01 'u,${ Jo Era aql f.a1;e d11:oqs "dl-lsnput
's,9751 ,{1rea
ar51 ur ssa:6uc,3,{q peqsrlqelsa serr srrlneuorev Jol a€1lrurcC TpuorfpN aqt
^JoGTAp;
)ruCU TSIJ 'IIOJTIIV \OVN CISOISIH 'E
.sTaoJ:rE t9 -:no olur se:nlseJ 1s€q el|] eurgllDc aar 1et{1 os
,pe1sa1 1ou ara.'\ lprFl saf,nfEs] ursfJEJ ssnJsrp puP 'p61sa1 seJnlPe] aql^uu a1e: uaql puE
, e^rF€dsfad fecrJolsrq p UDJJ lJc{'l lsal troJJTE vjrvN al|} azr.rEurtrts leou f IT,q a!1
'stra:rrcads ?sa1 fauunl pur/'r se .{Taxai 1ou 'saueTdJTe T€€}r uo asn :o; pau6ts
-ep a.re sTTolfre v9 '{ aclr€r€la: ur pezT-rElrxns 's-relarErEd rToJ-Ire uTElxac 6uT6ueqa
l:o slJelfe aqt €urulra1ap 01 sasal TaurJnl pura :taq1 Surr'ro110} ' alsl 10u plp lnq 'a eq
pTnoqs v:MN qcrw dals lxau aq1 exe /taw 'peu6lsap r(laarlea"rasuoc are FxrP '.&euor?n
-To^ar uEr{f .Iaqle: .iranorlnrone e.rP srroJJTe w ar+J 'sTlol:f,Te vsd}.I,/\rw!i 6urpuodsa::o3
lecrdr{1 rprrr pe:edrnc pue paluasa.rd a.rP sffnsa: aql pue 'uJPJ6o:d Jalnd)Do 9 eJrra:tl;aJ
6ursn pet;:r.ral ue€q stq suorl3as TTe Jo a3ueuuoged pelcrpe:6 'su6rsap ad,{1 r'ro13r reut
-uref pup firaTnqJnf go aSue: apr$ E 6uruuds ).r.r+Etx pelp:6a1ur ' TPcTfuT e ur pelr:asa:d
er€ sTroJ.rrE ,,\i9,. / sTroJ-rle vsvN,/\od!l ar11 o1 d:e-nuo3 'e5lrEurrol::ad peeds q6rq ur
'
as€e-r€p ou qlfn slLrorJTJJaoJ luarn:u 6urq31rd :aool pue 'ssatralTlce;3:a de1; plrE uo=)
, aJuar.roJx€d peads-tto1s ra113q pu€ sTTsf s ral]os raf l:o sf roFrP ,,v9., s , ro
-f r€ -r-a11€q ,,,sTroJJrp,,
-qtnp srr+l lrarr:ano5 aso(n 01 peted.rcJ '.41a1a1d:rc sfrolrrE asoql aceld
-€J 01 p€ubrsap are puP 'sTro]-rrE Jruosqns Vsl/t{ ra?ET pue ',,suor1€S 6ur3 3o tr:oew,,
,T aJueJ€Je.r ut peluasa:d sfrofrrp \D\iN aqf ur peurErluoJ suorssruD pue 's6unn:1:oqs
, s-ro:-r3 aq]'
l€.If,o3 sTTOlrrP asaw 'l{dl^l OoE 01 sF€€ds 1}E:3rrP :oJ ptrP r uorrrru ua1
01 auo utoJJ sr€qunu sptouxau lio a6up.r atp ur tr11eJarra6 suo-rl1puoJ rloff Jof 'sT feq:t
'suorlEf,rfddE uoTlpTAV Te.rauec -roJ uraJaq paluasa:d sr e?sp uorFTEE:d acuan:o1:ad
pa^r.fE)p ralndrDJ qgrn 6uo1e SIToJJTP cruosqns ,,v:1,, Feu6rsap-A1nau 96 J:o Eofels3 v
IXV'{^IrIS ' T
2
. A des.i-gnaaaon systen was then devised, ktlerein the first digit inaicates the rax-
inum canber j.:r t of chorC. t'he second d.igit indicates the locaiion of naxj:nmr ca,nber
in lenalls of the chord length, and t}le last two dj,gits inAnca.ie the maximur.l thick-
ness of the airfoil as a tr=rcentage cf tlre chord length. For exanrple, NACA 241-2 indj--
cates a 12? thlck airfoil with 2? rnaximur camber l0cated at 40t of chorc. l,tote
thal the naxj:nun tirickness of this airfoj-l, as j-n al-l of the 4 and 5 digit airfoils
( turbulent )
i is at 30? of chord, as deterndned by the gerieral t_hickness distrib-_
ution formula givea belcnn:. with this systsn in place, an orderly fanily of airfoils
was ccnstructec alrd tested, with various thicknesses, canber levers, and chord.-
wlse position of raxjmr"nn canrber. other va:iables that eiere tested, not incl-uded in
t"!re designation systqn, included conditj,ons of "stardard roughness',. si:nulaled split
flaps deflected 60 degrees, and Reynolds nurbers of 3, 6. and 9 nrillion.
The first step was to desigp a ca-reful-ly spcified rhickness distribution. The
|':AcA researchers noaed frcm earlier tests that ihe best airfoil-s, such as the Gernran
Go ttingen *398 and tie l.] .s. clark y, had the cnximum tbj-ckness at 30g of chord.
Accordrngly, the thickness distributlon for the 4 digit serj,es was sefected to cores-
pond closely lo that for those wing sections, and is given by the fo.l1c'\ring equatj"on:
ly used on popular airplanes of the day, such as Cessna (241-2), f,usccnrie and Aeronca
(4412). and seabee (4415), etc.
B- liACA,?-dicir Airfoils. Fo]-lc'\ring the 4-di9i! airfoil r.rork, }.{AsA conducted a test
program i.:r the early 1930's to see j-f zero pitching rTn:rents coul,d be acbj-eved on
canbered ar:foils by using lead:-ng edge droop al-one, rather Lhan using a convent:i"on-
al camber schedule. ( Read Addendum *1 below. "Understind.:irg ei-rfoils,', for a d.is-
cussion of the relationshi"p betueen C1, d. and en in canbered a.d tutca,,nbered air-
foils). zero p-itching nsnents on ca.nbered airfoils had been achieved earJier by
refleci.ng the airfoil trailing edge, as on the M-6 and 2R11,2 ai:foi1s, bui this
r€thod has :he d-isadvantage of reducing the na:<irnLrn rift toefficient, T'ire ns, test
airfoils, using leading edge droop al,one, becalre knourr as the ,'S-digit" ser.ies.
In doing these tests. NACA shared the general fascinatj"cn of t!:at era for zero
pitching nsnent arrfoils, without critically evaluati-ng the need for t-t.j,s feature,
and without q,rantifying the effssts of dj-fferences in airfoil cln on ajrcraft per-
fonnance.
The only airfoils of any conseguence i-n th-is series are th€ 230)c{ airfoils,
represented by the ubiquitous 230f2 airfoil, r,irrch is 12* t-!dck, and has l.BtC
leadirg edge droop forward of the 15t chord posiiion. The ai-rfoil- has no other cam-
ber. and tire rean line is perfeeCly straight frcxn ,l_5C to the trail,ing edge.
The tii-ckness d:rst-ributi-on for ai-l of the 5digit airfoi-ls ( 0012 etc ) is the sane
as for all" of lhe 4digit airfo.|ls. The saqe faulty LIACA nethcd of ccrnbining tlr-i-ck*
UoTIPTAP Te-raua6 ar+l pazTuoTlnro^al a^pr{ pTnoqs sIToJJTp r,^or} JeuTLIrEr suosE€-t asaql
:oJ pue 'uorl3npsJ 6erp luecr;ru6rs e st sr.ltJ 'ZTote \DVII uer{] ssaT tSZ sr frol-IrE
ZTZ-,9 \p\dlf ar{f ro] uorl3nEE-r 6e:p uorlcas aq1 , aTdle:€ :oJ . afqeuosEeJun 1ou st 6utlr,
truaTnqJnl JETTturs P o? pe:dua $gZ f o -Gp:o aq1 uo s5ur,ra arlf, Jo] uorlonpe: 6e:p
' snt{L 'afqe:}s alrnb sr puE ,.ATsnorna.:d pe^arT€q tr€€q psq Ueql :a1ee:6 sr , saueld:re
uorlcnpo:d uo ue^a r sTTo]rrp esaq+ (Ilr^r I'roU JpuTurel Jo flr€1re aq? fprlt g aara.raJer
ut s1:oda: \i51tr1,1 '?cef uI ',,fur.cd uorfrsup.rf,, ar{+ tp noTf luafnq:n1 01 ,,sdrrf,. ioTf
aq1 a-:oJaq , SurA aqf Jo Fed lsa1Jrrll aq1 puolaq {1eq6r1s p{rp 01 XJEq e:e;rns do4
aq1 uo ,{1rsee -raqlpf p€Aarq3p €q upc l,4of J rpuruJpT , sTToJ-rTp ser-l:rs-g asarl+ g:}TM
. sJTf.neuo.rap sluepn?s .snorr€s TTE rof lsnu p sr qJTr{q ,T aoua.lraJal .,,suoll
]o
-ceg 6u144 fo s I lrJoquacq uo^ pue floqqv ur pessnosrp sT r{.ro 1 srrll 'saurT rJs€tu
(tT6T'tr-' JTToqPred
) ^f,o€qJ,,
aq1 acelda: o? , suortnqrrleirp ssaulcrrlt .dtTcofa^ fustsuoo arl] q:lTA
6uo1e o6 ol. sauTT uEa! lr.Jglsuoc Jo sarJ€s p [Esrnap osfE \DvN 'safp ?Er11
ur trcuarcr3:g:a a,ro:dun 01 ^lTcoTaA
'uoTlaes ,,tr.Iaaoca: amssa:d,, iro .Ipoq.TafJe J'eUor{s ITTEJT
-fTf,c eql ur .,sdsnJ,. asn osTp sadeqs asantr1 ',,fa{onq :gunueT,, arI} sp utnrD{ sr a6ue:
6e:p-rrro1 srrdl 'lueTnqrnl o1 sdlr+ xoTf aql efo]aq oJ€z un:; sea:6ap E snuru :o snld
Xlalanrrxo:dde ;:o a6ue: {cEl1p Jo a16ue ue qbnort;l surr: rpurutpT 6uo1 rrreluruu o1 .d1r
-frqp aI'Fl lTqTWcd sedeqs asaqg, 'X1anr1:edsa: , sadpqs sar.ras-gg EJe ,-99 '-Vg ,-tg
aql .IoJ p:oqc Jo tgt p{re ,tot ,tg.Lt ,tgt 01 p.roqr Jo t0€ uDfl ?Jp pe^su ssau1rlI1
ummxEru Jo uorlrsod at51 Euraeq puE , rTppJ aSpe 6urpe:1 prrp suorlcas asou
.d1anr1e1a: trq pazrretce:Eqc , sedH{s .,.d1rco1an 4up1suoc,,=lfTEurs
=)puaTs Tprc€ds pesrnap sJa
-q3-r?As€.r \n)VN ' e.rOJaJAql 'SfrtJrrT /noJ-rpu rExf+p.r UT[+IT1'1 ' ({lTcora^ fus}suoc ) asrApJoq3
se ffar'l se tr11ecr3:an , pa?erafaJJpun aq lsnu,.loTJ arll , rpurulpT urEuF-r qI,
'r!!oTf .rEuTurpT se Serp t{cnu sp acr,,.'1 r(laS.anrrxordde seq qcTqA ,reou laTnqrnl pup
'6oTf reuTuel o:E scnrpqJ€ru pTnr} lo pTaTJ ar.n ur suorl'.tpuoc oolg: :o[au o f aq,L
. sTTOf
--rrE 'aoTJ JpururEf fo ' seT.r€s-g aq1 ur 6ur11nsa: ,s,0t6I pue s,0E5T a4pT aqt lo lJortl
floTI rpurureT arlf seA ltrandolarrap ulvN lup3rlru6rs lxau aq,:EJJoJilv sar. Fs-9 vCvN 5
'AoTeq possn3sTp 'sanbTu{Jef reoT} -reurujpl 6ursn
Iq ,sr ?prt+ ,sTroJ:Te acuan:ogred -:aq6rq +e6 crl .{erq ra11€q e sr a.reql leql sr
sTToJr.re 1T6fp-g aql 6uruopueqe .ro; ErElcuoo ,{1ag:es arp aprseq uosEa: rat1}ouv
'pe16au Tsururrf3 uo srap-rcq 'asn
:ran{f lsure6p s6uru:en 6uor1s ansst 01 lsEeT ae :o ' aJT. oueda-r :Taql U,OJJ slToJ:Tp
TTE?s-dJeqs afrururfa of vwt{ Jo a-rnTref aIdL 'To-rfuo3 TE-ralpT fnor{lrjrr tr1a1a1drnc
sr 1Je-ro:T€ aql aTTlliru€ai 'acel:Is do1 lro]rre eq+ uo ftoTl: peqcsflE qsrTqslsa-a-r o?
(apnfrtTp Jo ssoT luanbesuoc q+T/'i ) {lressacau sr {oe1tp lo aT6up Jo uorlJnpe-r aTqpraprs
-uoc E ' F€:n3co ssq TTsls drpqs p ra1;:g 'saueld:re atrr.Srra-u.rirr1 uo dlletcedsa 'sa,rr1
Iuprx peles alpq pTnoc pue 'sTroJre 11e1sdrals Jo arnleu snorprsur al+l pefpr-lsnTTT
a^eq pTnor! qJrr.ln 'slsal eqdle 6wsea:cap TTP"s-1sod qcns pepnTJur :1.roq 1sal fro].rrp
\D\dI{ ar{+ Jo auou 'FEf uI 'TTsts aW} un:;,{raloce: pellllalle ur sP ' panlcco sEq TTsfs
d:eqs ar41 ralJe lcs+lE ;o a16ue E@ I{fra srnf,so ?sr+ d@r srsa-ra1s.Aq aq1
apnrcur lou prp sTToJ:rre asetd+ uo E+sp ?sal \ovN a$I ?eq+ a?PunFolun sT fI
'11€f,oJre fe{rtrp uo pesn aq crf 4ou ,seTlT
-sor:nJ ,{:olefoqeT uet{+ alou 6urtgou sP pep.refu.r €q pf noqs sTroJ:rE 1T5T-p-E \P\dt{ 3ql
, 'suorleJTTdde uorlprae le:.arra6 ro3: alqe'ldanceun ,{p4a1d:m sr (sTTE+s
^I6urp-ro€ld
dreqs 1 pred act:d arp lnq 'euofe doorp a6pe 6urpeal dq peaalqce €q pTnoo slustcr1;a
-oc 1JTT q6rq dlqeuosea: sE TTar.r sP slueurau 6urg31rd orez 1eql perro:d slua:rrred
-ra rro].raE 1T6Tp-9 \D\iI{ elll 'f,ST ' lP aurT trE€ l aqf tlr (}iu'rJ{ ) xlrntrr3uocsSp arp ^dq
pesnec, TTE +s dr€qs aTgstdaJJEun up -e6elupAppsrp :ofE! p seq TroJ.rrp aq+ ' r.a^a/qoH
.1,4oTeq
f,X unpuapgg ur passncsrp uoueunuaqd lJTl,alqqnq uotle:edes,, erll 01 enp
' sTroJ:re 1T6fp,-t alqeredrnc uet1+ ra?ea:6:o o1 lenba firarorJ;mo lJrf uruxrxeu
p sEq
ua^a TTo]irre aq;, 'anrlce[qo uSrsap aq1 Eurfaau 'u{).^'toT sfr .IoJ s-luno33P auoTP
doorp e6pe 6u1pea1 un:y 6ur,peo1 p-TeAJol: anFrf).e alIf Wlr,t'l p€ur. ^r€Aqu-Dc srlll pue r pex€qusc
,rc1 dllercadsa sr Tro]:re ar{f 'snt1t 'Ctg'I trel4l raqler t?'T .d1uo sr 7I3EZ \6rvl.l
-rol: ,,-reqrE3,, aAr?ceJJa atll 'snrpe-f a6pe 6u1pe'a1 arFt fno 6ur,{e1 :o3 poqla[ ,,snrpe-r
pup edoTs,, arlf puP sTqf o? anp puP 'pasn osTe sPr'1 saf.suTp-ro ( doo-Tp ) -Idqujec puP sssu
I
4
industry, especially considering the fact that these aj-rp.Lanes are in a Reynokls nunr-
ber range that is especial.ly favorable for lanrinar flow aj-rfoirs. Furthernpre, the
constant velocity rnean lines devetoped by I'lAcA for these airfoil-s includes one
(a=0.5 nean lj-ne) that produces lc'qler pitching ntrrrents tlrough forward loading, with
no drag penalty. a well-proven technique that was used in the prior (turbuleni)
airfoils. Ttris mean line is designed to hold constant pressure ( therefore constant
verocity) back to only 50t of chord, which is the practical lilrlt for larninar flo.r
anyway, and then it tapers the J-oading linearly to zero at the trailing edge. For
anyone acquainted with the t{AcA airfoil design methods, and wirring to derive their
onrn airfoirs, ttris seerns to offer the "best of both r"orlds", that is, low Gn as wel,r
as ro^r drag corpared to the publisbed "base line" 6-series airfoi.l-s, v,'l-rich have
relatively high On due to nid-loading. In fact, Cessna did just that, by deriving
and using the unpublished I'lAcA 64Mr5(a=0.5) for the root.. ;nd the ,rnputri=n.a naca
64A412(a=0.5) for the tip, of the Model 2L0 ,,Centurion', wing.
spite of the potentiat l:erforrnance jmprovernent that could be nade by using
lanirnr flo* wings. they have never gained wide acceptance in the cA fierd. I'rost
snarl light planes continue to use the 4-digit and s-digit turbutent airfoils, and
the great nrajority of light twins, and even ccrmuter transports, use tl,e obsotete
and dangerous 5-digit airfoils, a horrible mis-nntch considering the consequences
of sharp-stall airfoils in an engine-out situation, The I.lAcA 6-ieries airfoils are
often avoided, since they have gained a reputation of being "kilrer airfoils. with
sharP stall characteristics and 1ow naximum lift coefficients, as r.,eII as re]atively
high pitching nErnent co€fficients. Read on. and you wi.rl see that this unfortunate
situation has nothing at alr to do with the fact that these airfoirs are lanrinar
flovrtype airfoils. IiACA sinply scre$ed up.
3. CA AIRFOIL DE1/EI.oPMNIT
A. ordinate ccrnltrination t',tethod. AlL of the NACA airfoirs, sunnrarized on figure 1,
rierE-GEFgnd-5lEe-EIisEIE-EEtroa of ccrnbining a thickness distribution with a
camber schedule to form a cambered airfoil. An obvious rnistake in t}Ie |IACA airfoil
develogrent ,nprk was the method used for ccxnbining the thickness distribution ord-
inates with the camber ordj-nates. There are trno wiys to do this, the right way and
the wrong way. von l'lises discusses both methcds in reference z, uut unf6rtunalely
he does not indicate a clear preference for one or the other, The first (correctJ
metrod is sinpry to add the ordinates together at each wing station. The second
( incorrect ) nEthod, hthich for scnre unknorr'n
reason appealed to NACA, is akin to rolling
a. circle of ever-changing radius aLong a curved nrean- line, top and bottcm, and then
the area swept by the rorling circle beccnres the airfoil shape. This method. sho/rn
on figr:re rrr-r belovr, is quite curplicated mathernaticarly, rrtlich ironi-calry nny be
the reason that |lAcA picked it. unfortunatety, it leads to a distortion at the read-
ing edge, h,hich is tben $rcthed by the so-carled "srope and radius" method for for-
nting tbe reading edge - rhis taulty prccess is used on all l|AcA airfoils. un-
fortunaLely, this has the bad effect of super-erevating the leading edge above the
originar chord rine. Figure rrr-r sho^'s an exanpre- with ttre NACA ;4ls-airfoir the
leading edge is lifted .486c above the original chord line, creating a new rnean line
aborre the originar rean line. Tbis de-carnbers the airfoir, as a funition of the air-
foil thickness, as shovrn by figure 5. As a result, the thinner t.lACA ail:foil-s have
rDre camber, and thus m3re lift, than the thicker ones. this nlay explain McA,s
fascination with thinner airfoirs, especially for wing tips- anittrer nlistake. see
Mdendum Nurnber 2 belo^,. l^lorse yet, the incorrect rnetir"a ccnbining ordinates
frattens the initial slope of the rnean line, and this detracts "r frun ine row-speed
perfornance of the airfoil. But NACA never investigated this effect.
Figure 2 shcnrs the beneficiar effect of using the correct rnethod of ordinate
ccrnbination on a typicar }{AcA airfoir, the ubiquitous }.IACA 64-212. when tbe correct
method is used to form the airfoil, redesignated 64-zrzc I the stall is softened
appreciabry and the naxirnr.rn lift coeffi-cient with and without flaps is raised
'
srightl-yf ccnpared to the originar airfoil. Accordingly, the "dirdt addirion"
rnethod of ordinate ccnrbinaLion is used airfoils.
5
B. l"'lean Lines. Eigure 3 sha^is the wide variety of rnean rines that have been used in
llAcA'/MsA airfoils. from extre{nsLy forvrard-l.oaded ones (230pr) to aft-loaded ones
(cAw-r'-2). l'lost of the popular llAcA 4-digit airfoils use npderatery forward-Load-
ed nean ]ines (x4xx-see figure 1) with naximum carnber at 40t of cho;d, since it was
kaown prior to 1930 that ion{ard loading produces lcw pitching nsrent coefficients,
and this had been established in general telms as a deiirable airfoiL charasteristic.
In splte of this, only 5 of t}le 52 6-series airfoil-s listed on figure I have forward
loaded nEan lines (a=0.5,0.6), while a1r the rest have tl:e exaccly rr_id_loaded (a=I)
nean }ine. !'lhy the contrad.iction? Srnply because these aj-rfoils are n'erely test air-
foi-rs, not designed for actua-l- ai-rplane use. onry enough fon,Tard-]oaded airfoils
(five) rrrere incl-uded in ttre test program to confirm the effeses of that variable.
F\.:rther, the reLative inForbanc€ of the alrfoil pitchi,ng nE rent coefficient (en) co(rF
ryred to the drag coeffj.cient (cd) has never been studied nor guantified by NACA or
NASA- Horv can one intelligentty desigrn airfo.j-ls wj-thout knc,v,ring this relationshipu
For exanple figure 6 shorrs tiree airfoils with various crrnbinations of drag co"ific-
'
ient and rnrrent coefficient. Holi inportant is it to have lorr on, relative to @? rs
it al-1-irportant, or just a nicetyz Do he need zelo qn?
ans\"€r these guestions, ,,.E investigated the drag that is caused by airfoil On,
and- derived
-Tb.
a formula ( figure 8) for calcuLating this ',tr:m arag,,, or drag associated
with trirniring out the airfoi] pitching nErEnt, ior a typi.calty ;onfigured light p1ane.
hltren tri:n drag is added to the airfoj-l seqEj-on drag, arrivl at a icorrected,, drag
cefficient, Cd', which can nc'$r be used directl-y for 'nle- ccnparing candidate airfoils
for a given application.
the problen arises due to the fact that the center of pressure of cam-
bered-eriefly,
airfoil-s npves aft frcrn the c/4 position as the angle of attaci decreases- see
figure 7- This causes an. aircraft nosedown pitching nc.rint at higher speeds that
must be rea*ed b1r negative Iift in tie horizontal tail. Remqnber that Lhe aj-rcraft
cG rnust stay at or near the c,/4 position for satisfacLory r-or-speed pitch stai:irity.
But the negatj-ve li,ft in the taiL causes induced drag. e-lso, tne wing must nole gener_
ate an add-rtionaf incrsrent of lj-ft equal to the taii doern-Ioad, sin6e the sunrnatj,on
of vertical forces on the airplane must be zero. Ttlis additional increnrent of wing
l-ift also creates an incre{rent of induced drag. Add these tr,,D additional drag incie-
rnents (wing and tail) togetber, and r,,e have trim drag. l.totlce frcrn figure g tlat the
trim drag coefficient for the very high-on }{AsA GAw iirtoil= is approiirnateJ,y 20 drag
c€unts at cruise for an application having a design rift coefficient of 0,2. when
this is added to the section &ag coefficient of 53 drag c€unts, re obtain a correct-
ed drag coefficient of 73 drag counts. This renders the t.{AsA GA!,1 airfoiLs unsatis-
factory, or at least verlr disappointj-ng, for this type of application.
As a result of analyses Like this, rde have concluded tfrat forward loading is
necessity for general aviation airfoi-ls. Hoirever, it is not necessary to go to the a
ridicu.Lous extrenE of the ill-advised 5-digit airfoil test progran {;airt6its f,rrri.,g
the nExrmum camber unusuaIy far for^rard" ). In short. it is- noi necessary to have
zelo qlt, nor is it necessary to reduce on to the point !^'l)ere other vital perfonrnnce
characteristics of the airfoil. are curprcndsed. rhe IrAcA S-digj-t airfoil,s' achj-eved
1"t9 Gn, by! the price pa.id ( sharp stau) is unacceptabl-e. th&efore the S-digit air-
foiLs should not be used,
therefore, have selectd the IiACA (a=0.5) nean line to be the basis for the
universal family_ '.,eof only
four canrber profiles for alI of the "GA,, airfoils, turb_
urent as 'nell as laminar frovr type. The a=.5 nean line, h'hen substituted ln the NACA
64-2I2C airfoil. produces 64-212(a=.5)C. Figure 2 shorr"s that this results in a wider
larLilar bu!\9t srightr,y higher naximum lift clefficient with flaps, and nnst impor-
'
tantry' a 20t reducLion of pitcNng nrnent coefficient, with no oiher penalty.
The next step is to contror the i-nitiar slope of the r'l3an line, an irportant
parirEter for good slorspeed perfornEnce, Ttris was ccrpl,etely nr-issed in the ltAcA
work..aLthough scne later IIASA airfoils have lead"ing edge droop, h'hich corrects ttlis
oversi-ght. I{e have found t}rat the initial slope should ne a nLin:mum of 12 degrees
for any airfoil, thus our GA-2 nean line inccrlnrates .3* drop, and the GA_3 rnean
line has .2t leading edge droop- see Appendix ir. ne effeceiv;ess of the drmp
6
can be seen in the perfornance of airfoiL C.A j7- .2, which has a much softer stall-,
ar even wider laminar bucket, Io/r On, and yet higher Clmax with flaps, with no drag
penalty. as shcl'Ir on figure 2. For this reason, all cA airfoi.l-s can be described as
" pitching nsnent, soft-stal-l- airfoils". In fact, rTt3re]-y by selecting a GA air-
lc'e/
foi.I, an aircraft designer can be assu-red tlEt the pitching nsnent coefficient has
been held as lc",r as practical, short of degrading other perforrnance features of the
airfoil, thus for the rna jority of cases a trim drag analysis is not required. Hor-
ever, in ext.:rslE cases, such as a very high perfornance application with the design
l-ift coefficj,ent less than .15, a fornal ':ri.lrr drag anatysrs may be advj.sable.
Fort. ard loading has one possible disadvantage, descrj-bed on figure 10. Since
forHard loadj.ng increases the velocity ratio slightly in the forward half of the air-
foil canpared to nrid-foading, the critical Mach nunber fo! these airfoi-l-s is degraded
slightly. Ho\^rever, this occurs in the 580 MPH range, far above our needs, hence this
is of no concern to us.
C. Thickness Distributions. For our ce arrfoils. only ninor changes have been made
to ffibutions. The cA thicknes; distributions are listed in
Appendix I belcr,/. For the turbul-ent airfoils (cA 30), the GA 30A015 shape is ident-
icaf to NACA 0015. Hot^rever, this strape has a slightLy convex, or "boat-tailed" after-
bcdy, rrtrich is rather unusual. Therefore, we have generated an alternate tu-rbulent
shape, GA 30-01,5. with a strai.ght afterbcdy, that reduces the prof.ile drag frcrn 75
drag counts Lo 72 drag counts, with no structural disadvantage. One concern here, hov,l-
ever' is the change in the tlaj,ling edge angle, h'hich affects aiLeron effecti-veness-
see figure 9. Since the trailing edge angl-e is reduced in cA 30-015, the control-
effectiveness nny be slightl-y }ess tlnn wj-th GA 30A015. Hs,Eve.r, the trailing edge
angLe rsrlains generous. Further, there are better '.Eys to get effecLive ai-J-erons,
such as naking "fat" or even "crowned" ailerons, which can then be used with the
cA 30-)oor airfoils.
For the GA 35 and GA 37 larninar shapes. no changes were nEde to the }IACA 63
alld 64 series shapes. which are conservatj.ve, utilita.rj.an l-arni,nar sections, having
wide lamrnar buckets. Thj-s alloers airplanes with these thickness distrlbutions to op-
erate over a wj-de range of flight conditaons, while still renaini-ng "in the bucket".
This feature results in a profile drag penalty of perhaps on],y one or tr,ro dfag counts
carpared to sqne other so-calJ-ed "superior, lore-drag " lam:-nar thickness distrib-
utj,ons that have been develo@ since, ard the drag penalty is a snrall- price to pay
for the wider bucket on the c"l-assic t'lACA shapes .
For the GA 40-015 shape, r,E use the NACA 65-015 tiickness distribution as is, Tae
profil,e drag coefficient of this shape is also slightly higher than recent NASA ,'slar3"
shapes, but tbe drag bucket is crrrrespondingly wider, for olErational flexibj_llty.
For the cA 40A015 shap€, we have carefully avoided the }IACA 65A015 shape, which
has a poorl-y designed, straight afterbody (high drag ) and should not be used, W1th
the point of nraxirm:m thickness ncved aft to the 403 chord position, the design of the
pressure recovery section krc.rEs critical, and a straight afterbody is no longer
suitabLe, due to the unaccaptable drag penalty. Accordingly, l,ve settl-ed on a reduced
cusp shape by ncdifying the I'IACA 65-015 to fj-]] in approxlnately one-half of the cusp.
This gives the deslred strucEural and control effectiveness advantages of a true no-
cusp shape, with a drag penalty of .Iess than 2 drag counts.
ttotice that the basic thickness distribution for the c"1 thi-ckness distributions
j-s I5S thick, and then the 12t and 18* thicknesses a-re scaled proportionately up
or dorn frorn the 15t thick shape. this is a departure frcrn the NACA schen-- for their
landnar thj-ckness distributions, wherein unique syrnretrical shapes have been calculat-
ed and published for eacb thickness, and these shapes are not guj-te exact nruJ-tiples
of each other. The difference is very srnal,J-, ho^'ever, and does not affect a-irfoil
perforTnance, therefore ue have ignored it to si.nplify the process of generating fan-
il,res of airfoils.
llotice tiat the ninjmun sestj-on thickness offered j-n "cA Airfoils" is 12?, for
$e do not reconrend the use of wing sections thinner than this for any application,
even for racing airplanes. In fact, the optjmrn airfoiL thickness rega.rding lift to
7
drag ratio is probably closer to 13? or l4B thick, for any airfoil systern. Sirply
put, you lose nr3re than you gaj-n by using wing sections t-hinner than l2t thick.
For tail sections. synnEtri-ca.l- shapes such as NACA 0009 or I.IACA 63A009 are ccrm-
only used. Ordinates for these shapes are l-isted in standard I.IACA tabLes ( refer-
ence I)' or they may be scaled dcrrn frcrn the corresponding C,A thickness dj.stribution.
D. ordinate Tables. Notice frcrn figure I tlEt the NACA "ordinate tables" are rnerely
a-fGrino of-test sanples tJ:lat r,prt included in IIACA' s wind tunnel tests , and are
o-ng
thus incfipl"ete. In contrast, the GA airfoil ordinate tables. Appendix III| are a
ccnplete matnx of 96 airfoj.ls, in four series, tiro styles (with and without cusps),
four camber LeveLs, and tlree thicknesses, offering the designer a ccmplete "cook
book" of airfoils fron \,tdch to choose,
MSA's official position i-s that a ccnprehensive catalog or " cookbook" of air-
foil-s is not necessary, and suggests that a unigue airfoil- should be custorn-designed
for each and every nan airplane, one that "exacely fits the specificati-on for the ne\.r
airplane", otherwise ttle lErforlnance of the nevr airplane will be ccnprqltised. This
is, of ccurse I nonsense. there j.s no single, unique solution to a particular airfoi-]
design problern, since there are so rlany value judgenents. intangibles, and perfonn-
ance tlade-offs invofved. For exanPLe, hc,r "soft" is "soft". for a soft-stall airfoil?
Therefore, there will always be a need for a good catalog of airfoils, if for no
other reason than to serve as a standard to ccrnpare custcnFdesigned airfoils against.
FlrCher, airplane designers are by-and-J.arge not airfoil designers, and ia^culd rather
select frcrn a good catalog of standardized designs rather th,an take a chance on a
custcm airfoil.
NACATS thj-nking. evident frcrn reading reference L, was that "custcrn" I{ACA air-
foj-ls could be s)mthesized using l{ACA's data and nethods. In other hDrds, IIACA assund
tlnt a designer coul-d select a particular thickness distribution and a camber profile
frcrn tlle text and then ccrnbine thern by the I'IACA rethod, to obtaj,n the desired custcrn
airfoil, just ' as cessna did for the !'todeL 210 wing desigm. This is probabry t-trc reason
that I.IACA was content to publish test data a1one, rather tllan to publish actual- cata-
logs of airplane arrfoi.l-s. Hot,rever, to thus s)mthesize custam NACA airfoi-1s is a
dauntj-ng task, beyond the capabiJ-ity (or pa.tience ) of the average designer'. Further r
due to the nfstakes and onissions in the NACA work, the airfoils vDuld still be less
than ideal. For example, Cessna,s airfoils for the lbdel 210, r,,trile a step in the
rj-ght direction, do not contain leading edge droop, and the ordinates are ccrnbined
incorrectly. which is both rronic and unfortunate. liorse yet. on rrlf,st other airplanes,
Lhe test airfoils r,rere used directly. warts ard al}, because of t{ACA's nrista.}<en assum-
tion that designers r^rould slmthesize "good" airfoil-s frcm the test data.
The GA ordinate tables are arranged with the ,,cusp', tlpe airfoil-s at the top
of each page, ajld the corresFonding "no cusp" airfoi]-s at the bottcn of the page.
The ordj-nate schedules generally follora the NACA forrnat, hov,ever an additionai statj-on
llas been added at .25tc, to herp define the leading edge. unlike the !,IACA airfoils.
GA airfoil.s do not have a tlrue reading edge radius, and the readj-ng edge is j-nstead
a curve defined by a series of discrete points.
co],uru''s #2 and Jl3 of the ordinate tables are the thickness distribution (sym-
nEtrical shape ) and tbe nean rine, respectivery, frcrn which the tirree airfoj,rs 1n
each table are fornred. corunm #2 and +3 are listed for reference only. they are aLso
usable for forrn-ing airfoils with th.icknesses other than the three li;ted in ttre table.
The Procedure for this is to scal,e col-umn *2 frcrn I5t to the desired thickness, then
add col-unn *3 to obtain the desired aj-rfoil,
E. Aerodvnandc Characteristics of winq Sections. The aerodynaninc characterj-stics of
each l-2t thick arld L5t thi.ck "no cusp" type cA wing section, as detern[ned by the
"EPpler" prcgran, are given in Appendr-x rV. for Relmolds nunbers of 2 Irtillion and 6
ndLlion, corresponding approx-i:rateJ-y to stal1 conditions and cruise conditions re-
spectively for tlpical light airplanes. The characteristics of the 188 thick wing
sections are not shcrrn, but they are approxjrately the sarne as the 15t thick segtions,
witn 2lr drag counts addltiona] profile drag. See fign:re 15. sjnr-ilar]y, the charac-
teristics of the "cusp" t)pe sesLions are not sho^rn, horever: they are approxirnately
the sare as the "no-cusp" shapes, wi,th 2 drag counts less drag. Sree figure 14.
8
Tbe characteristics at Reynolds nurnbers other than 2 miU-ion and 6 rRil-Ii-on rIEy
be estuated by interpolation and extrapolation. and by reference to figu-re 17.
The data in Appendix IV j-s for the snDoth condition onJ.y. Pigure 18 shovrs the
effect of surface roughness on the perforrnance of a tlpical GA airfoil, 37A31-5. The
prjjnary perf orrnance degradation is ]-oss of lajrlnar f J-o,r', causing a substantial in-
crease in profile drag, and there is also scan-- Loss of nraxirm:rn 1j-ft coefficient as
wel-l-. Ttre data r-s for roughness conditj-on !=4, b'trich is approxirnately equj-valent to
the NACA " stEndald rouglmess" condition used in their tests. According to reference
I' this is eguivalent to a wing ",,e11 contanrinated with bugs". Ho+rever, the concensus
is that the ''standard rouglmess" lrrposed in the I{ACA tests is unusually severe, and
is not likely to be encounlered in actual alrcraft service. F\rrther, due to the super-
j-or leading edge shape of the cA airfoils, the perfonnance degradation of GA airfoils
due to surface rouglness is proba-b1y not as severe as with llACA airfoLls.
TLre perforrnance data in .\ppendix IV is for wings without f1aps, A]though rnuch
experinentat raork has been done in wind tunnels with high-lift devices (see reference
If chapter 8), there is no readj-l-y available ccmputer program ttlat can accuratel"y
predict airfoii- section perfornence with flaps, due to the ccrnplicated flc,ld patterns
in the flap systsn. AccordingJ.y, very littl-e data is presented herein for GA wings
with flaps, except for figure 2, which illustrates the principLe that cA a.irfoi]-s
have significantLy better fl-ap effectiveness, that j-s, increase of nraximum lift co-
efficient with a given flap system, due to the cumulative effect of the cA jjrprove-
rpnts, than thei! IIACA counterpart. airfoils. However:, raE offer the folloering as a
"ruLe of thunb" for estiratj-ng the naxj:num sestion l.ift coefficient achievabl,e with
flaps, for cA airfoj-ls. For large, ,,e11 desi-gned f1aps, the greatest expectable fac-
tor of jrprovenEnt in section Clnax for sinple flaps is 1.5. for slotted flaps 2.0,
and for Fc&vler flaps is 2,5, thus, for a cA airfoil with a no-fIap C]JrEx of 1.6, the
Clrex with sirrple flaps coul-d be as higb as 2.4, slotted flaps as high as 3.2, and
Fo/rler fl,aps as high as 4.0. ltese are section (2 dilrensional ) coefficients, and
appropriate reduction must be made for ttrree-dinensional effests. AIso, they a.re the
estirnated nnxirnum values for the best fl.rp systens, whereas existing production flap
systers probably fal1 seII shorL of these nunbers.
the ccnputer program used to generate the Predj-cted performance data of APpen-
d1x 1y is tni: eppfer program "profil", published in Fortran in reference 5' and
adapted tor perional 6c.teut. use (itS-DOS ) by referenc€ 6. Ttris is a sfuple program'
and does not iterate on itre tunOary tayer, thus it gives accu.rate results for only
rrclldesigned airfoiLs such as ,,GA" ai-rfoils, that have slrpoth, contj,nuous nean
U5es, roi porly desigrned ai-rfoiLs witi discontinuj.ties in the n6an line, such as
t]le !,IACA Z3-OIZ aj.rfoif ; it gives inaccurate results, especially for Clrlax ' because
11
it cannot fol.Lod the ccnplicated effests of premature setE-ration bubbles at higher
angles of attick ccrnrpn in those aj-rfoiJ,s. But htro wants to use poorly-designed
airfoils any/€y?
A cqrParison of cdrputer resu]ts and wind tunnel data for a t!?ical airfoi]-,
NACA 64-215, j.s shcr,fi on figure 2j. Note Chat the accurasy of the cdrputer data is
guite good, all tfrings considered, and is a credit to Dr. Eppler. The program
appea-rs to be especiarl-y accurate i-n predicting laninar/turbulent phencmena, separ-
ation bubbles, etc. Ho^rever, the program aplEars to overstate drag by at ]east 109,
ard to overstate, lift by about 53, The r€rst disagreenrent appears
to be in ttle pitching nrnent coefficient, btrich tbe program overstates by at least
30t. For thi-s reason, as absolute data, t}le ccnputer re-
sults shouLd be used cautiousLy. Ho\rEver, for ccnparative purposes, such as shqn'n
on figure 2, the program is exc€.]Ient.
The program appears to be cal"ibrated for best accuracy for airfoil thi-cknesses
of about L5t. ltre program shqrs a considerab.l-e loss of Cljnax for thinner (12t tnick )
airfoifs, ard a gain of cJ.lrEx for tldcker (18t thick ) airfoils, as shc,vrn by figu.re
16. Ttris does not agree with wind tunnel data, v,hich tlpically shows negU,gi-ble
differenc€s of cl,rnax as a function of airfoil thickness. Dr. Eppler cffrents on this
iI his book. reference (9), and suggests that, due to the lijn-itaLions of wind tunnel
testing, especr-a]-l-y the change in the ratio of aijfoif cross-section area to the
tunnel tlEoat size witi different ailfoil thicknesses, Ferhaps his qxrputer dat! is
fipre accurate than wird tr.mnel data. In any event, lrE suggest that \r,hen ccf{Ering
the perfornence of different airfoils with the cc(puter prcgram, one shoul-d use
the sanE airfoil- tldckness for both airfoils being c.crpared,
the Eppler prcgram has a routine for sirurlating the !=rforrnance of an aj-rfoi1
witi sjrpl-e flaps. Ho,$ever, tlre predicbed performanc€ is wildly opti-rdstic crcnpared
to wind tunnel data, because the program of necessi-ty, wfien the flap chord and de-
fl,eqtion are inputted, drav,,s a ccnpletely new, idealized airfoil tlrat is unreal-
istically snEothed so that the nenesis of the prcglan ( prenature flc'gr separation )
will not occur. Furtherlrl3re, tlle program caffiot begix to sjmulate the cffplicated
flo\^' patterns of npre elaborate flap systens such as sl-otted and F* er flaps. Ttrus,
rile did not spend nu:ch tirrE with the flap routine in this prcgram. In spite of this
shortcurfng, ho$€ver, the flap routine is valuable for predicting relative differences
of flap trerfornunce betrcen sj:nilar airfoils, as shGln on figure Z, Another use for
the flap routine is to predist On vs clnax for a given flap systern. In other
rnords, if a sirple flap systen yields a Clnax of 3.0 b}/ the prograrn, then the repDntd
On will be reasonably acsurate for a flap systern with tbat CLlnax, even though it nay
take a clonsiderably nDre elaborate flap confi,guration to achieve thatclnax in prachice.
The Eppler program al-so does not incl,ude any correction for ctfi[)ressibility
effects. treaLing the airflorr as bei-ng inccnpressible. Ilo€ver. this effect is so
nlinor as to be insignificant, for the flight regi.res of typical-,light general aviation
air planes. For exanple, the tdlng ]oading of a Piper cub, 1*/ft', yields an aEerage
pressure dj.fference of only 7 */ft-/L44 = .05 psi betr"Een the top and bottcrn surf-
aces, Ccc$Ered to the atjrDspherj.c pressure of 14.7 psia, this arEunts to a pressure
difference of only 0.3t, For higher perfonrEnce airplanes, such as a Bonanza with a
wing loading of 20 */ft', ttre difference is sLill Iess than 1?. thus, the assrnption
tbat the flovi is ilccnrSrressi-b1e is not bad.
ltre Eppler program for perfornance a''alysis calcufates ve.l-ocity ratios across
the aj.rfoil at each algle of attack, top and bottqr surfaces, then arints these ve1-
ocity ratios (Foint velocity divided by free stream verocity ) in tairurar and graphic
form. AJ-ternativery. it will calcurate and print out pressure ratios across the air-
foir- then, the program catculates boundary layer data for each angle of attack at
specified Reynolds numbers, including coefficienls of 1ift, drag, and pitching nrcnr
ent' the turburent area. ani the separated a-rea. Ttrese values are also printed out
in tabul-ar form, arld in a "boundary rayer swnary ptot". A salrple analysis print-out
is shorr'n on figures 24-I thru 24-4.
12
Several- yea.rs ago Dr. Eppler brought out an "ifiproved" version of hj,s program.
Ho\^/ever, the ner^, program seems to understate section drag, and also reports lalninar
bucket width as j-ncreasing with an increase in Reynolds number, htrich is not the case.
F\.Ether, the resuLts appear to be overly sensitive to data point (ordinate) accuracy,
thus I still favor the old program.
Eoth prcgrarns include routines for desigrn of airfoils by the nrcdern, one-step
" inverse" rethod, whereby idealized velocity dj-agrams are first calculated, and then
airfoil top and bottcm surface ordinates are ca.l-sul,ated frcm the velocity data. That
is OK, but the Fossibil-ity exists of "creating a nonster", such as the very recent
NASA NLF(I)-OII5F alrfoi], discussed in Addendun No. 5 bel-or^r. lbis is less likely
to lEppen witb the "classic" nEthod used by },lAcA and "cA Airfoils". In any event,
neither the airplane, the wind tunnel, nor the ccrnputer kncrds or cares htlich n€tiod
was used to design the airfoil, and the final perfonrEnce figures a-re the only thing
that caunts. Thus both nEthods are valid. Any airfoil, even if hand-sketched, can be
broken dov'rn into a symrEtrical thickness distribution and a camber profile (nean
Iine). and much can be learned about the airfoil- by inspeccing and analizj-ng these
tllo ccmponents sepajately.
7 . GA CIJFFS FAR WING RE\.{CRK
As noted earj,ier , all of the NACA airfoils have Iffr nose prof i1es. due to the
faulty NACA design nethods -The result j"s poor slcvr-speed Perfonnance, i-n rEny cases.
me niCe 4-digit ( turbufent ) and NA6A 6-series ( laminar) airfoils can be irproved by
adding a leading edge cuff to the airfoil, which drops the center of the leading
edge ipproxirmtaly tt to 1? of chord length. The rpthod for designing these "drooped
leiOing- eages,,, shq,n-r on figwe 25 for the !{ACA 64-212 airfoil, is to design a neur
nean line iorward of 10SC to obtain at feast 12 degrees of initial nrean line slope,
necessary for a soft stalI. Then, usj.ng the existing toP surface and the revised nean
Iine, r,,e calculate the ordinates for the bottcm surface.
Ttre perfornrance analysis for this cuffed alrfoil shclts a draratic jnprovenpnt
in ttre stall ccrnpared tso the original 64-212, as r.vell- as a slightly wider lan.inar
bucket- see figure 26. A-Iso shorvn is the perforrnance of an uPIEr sulface nbdification
proFosed by Hicks et aL at NASA An€s. Ttlis cutes the sharp staff ' but it also reverts
the airfoil to a turbuLent sectsion, causing hj-gh drag ' The perfornunce of the GA
37-212 is also sholvn, for ccnpari son.
This cuff nndification should be used only as a "guick fix" on existing wings,
not for ne$, constructsion- use a GA airfoil instead. A.Iso ' it does not \'tork on the
NACA s-digit airfoils, like the ?3012' which already llave too much leadj-ng edge
droop, and a different stall nrechanism, than the 4-digi-t and 6-series airfoils. See
Addendum #3 belov;.
8. }4ISEIJANMUS AIASC AIRFOIIS
!,iitb the cun-ing of the slEce age, NACA'S naflE was changed to NASA, and the agen-
cy's responsibiiities were e4xnded !o include national space Projects as r"eII as
the old-line aeronautj.ca.l, Prolests. Ttre total budget is currently about 14 bi]lion
do.llars yearly, of whi-ch about 1.2 billion is refated tg "aeronautics" projects, a
fairly cpnstant arpunt yearLy. Ho$Ever, over the past Several decades, '.Jork related
to @nera1 Avj,ation projects has dwindled to nearly zero, due to "budget restlaints",
in NASA,S hDrds. Ttre only general aviation airfo.il r.ork, for exanple, has been the
sporadic release of several randcrn airfoils. These have usual-ly been the brain-child-
ren of particul-ar NASA individuals, rather than any c,oncerted effort, and have been
disappointmmts.
A. N;\SA GAI,t-j- . In l-974, the NASA GA!.J-I and GAVFz "Whitc6nb" airfoils ( later designated
EtlFb'qffi 15(I)-041-3) vere released. Ttrese airfoil-s are characterized by a large
Ieadrng edge radius. and a slab-shaped profite having the maximum thickness at .40c.
this pioduies a long Iaminar run, at the expense of lanr.inar bucket width. The leadrng
edge has a droop -
of about .75tC, npre than it ne€ds to give the airfoil its soft stalf
chiracceristic. Ttre 1c'vr profile drag i-s enhanced by the pronounced cusp in the afier-
13
body, and the blunt trailing edge, The brorst feature by far, hcturever, is the aft-
loaded camber profj-Ie- an outrageous nListake for an airfoil that was touted to be
a general avj-ation panacea. This gives the airfoil a pitching rTDrnent coeffic.rent
three tijnes as high as it ne€ds !o be for the alrpunt of canber in the airfoil-, and
this produces very high Lrjr drag, rendering the airfoil unsuitabLe for genera.l av-
iation use, In effect, the airfoil has t'nlo notches of flap perrnanently buil-t into rt.
Unbelievably, NASA chose this airfoj-l for their "ATLIT" project, described belo'v.
Figure 27 sholas the bizarre press release concerning the NASA "ATLIT" (Mvanced
TechnoLogy Light 1\,rin ) project in 1974, r"trlch turned out to be a disaster, perfor-
nance-wise. The project was dccrned to failure frcrn the beginningdue to the seLection
of the hj,gh-Gn NASA GAvl-I airfoil for the wing. the predicted perfornEnce figures
\^rere no doubt calculated without regard to the high On of the airfoil. a rna jor blun-
der. The spoj.lers for rol.l- controL \rere also unsatisfactory, experiencing control
reversal at ]ow and mediurn deflections. Nothing new was Learned by using a Longer
sPan. tapered wing- these effects had been kno\^'n for decades. This merely contam-
inated the data, making it impossible to ccnpare to the base-l-ine ( factory ) airplane.
I'IASA \das so ernbarrassed by the poor perfornnnce of this airplane that the pro]ect
engineer never even bothered to hrite a final report on the projecc, but he got pro-
ncted any,ray. He no, has a top position at NASA Langley. In retrospect, I.iA.sA forgot
that their function is to c.onduct basic aeronautical R&D, not to "shcwr the industry
that they are \,rrong, and this is the proper way to build a GA 1i9ht twin". Ttris is
an example of hhat can happen if R&D is not rnarket driven, bub is directed by irre-
sponsible bureaucrats. Ttre airplane was bougbt back by piper for salvage of engines.
instrulTEnts, etc, and \aras then sold to a technicar school for students to drill- hores
in it. with the proviso t-tlat it never fly again.
Figure 28 shc'v/s the lA\Si\ !'A'ILIT' airplane during wind tunnel tests at NASA l-ang-
J-ey. this j.s g! npney.
The cAli.l airfoil was also used on the Beech Skipper and the Piper Tcmahawk,
wj,th dj,sappointing resul,ts on both airplanes, parcicularly at high speed.
A ccnnnn revrork for this airfoil is to fill in the bottcrn side cusp, as shcun
on figure 6. This reduces the aft loading scmewhat. reducing Gn and trim drag, but
the airfoil refiEins a rrr-ish-nash.
B. MSA NLF(1)-02I5F Airfoil. This airfoil, due to Scnrers, and descriH in Reference
T was a$n- oilg'inaTlt as a sailplane airfoil. thus it has too much carnber in it
for typical porrered airplane applications. Accordingly, the trailing edge must. be
reflexed approxjrraCely l0 degrees as shc'vn on figure 29, for powered airplanes, Even
with the reflex, high aileron hinge rrnents (stiff ailerons ) renEin, so a ccnrncn re-
r^prk is to make the ailerons and f laps f lat-bottqned, as shc^,,'n on f i.gure 29 , to
salvage Che airfoil. Of course, the better solution to the problen j.s sjrply to de-
sign an airfoil witl- less camber in it in the first place.
Anothe! strange feature of this airfoil is the canber "diP' .60C, as shc'vzn
on fj-gure 3. This is an attefipt to move the trans.itj-on point aft, "a extendj-ng the larn-
inar run to reduce profi.Ie drag, which it does. Hcrvrever, the fallacy here is that the
negative carnber at .60C causes negative lift, and this causes nore ( induced ) drag
than is saved by the extended larninar run, reducing overall- airfoil efficiency.
Al-1 in all, MCA,/NA9C has never reached a concensus as to the best shape for carnber
profiles for general avj,ation airfoils, See the "GA" camber profiles in A5pendix II
for a conparison,
C. I{ASA NLF(1)-04I4F Airfoil. this airfoil, figure 30, due to viken, was envisioned
as a very lo* drag lamjnar flovr airfoil for high-pornrered, high perforrEnce general
aviation airplanes. the Foint of rnaxinum thickness on the airfoj-l is at .45C, and the
nraximum camber, c.entrally loaded, is approxirnately 2.53c. This makes the airfoil
ccnpa.rable to the }IACA 65-414 airfoil , Unlike the }IACA airfoil, hodever, the t\T,F air-
foil has enough J.eading edge droop, approxjrrately .353C, to give the airfoil decent
slc'vr speed perforfiEnce.
Realizing that the price pa..id for locating the nraxi:rn:m thickness so far aft is
a very nafrc'e, l-arninar b:cket, NASA fliSht tested the airfoil on a Cessna 2I0 with
14
narro^' chord (I2tC) "cruise fl,aps", reflexed upr,rards approxunately 7 degrees at cruise,
and adjustable. Thj-s was an effort to shift the faminar bucket verEically as needed,
to keep tbe airplane "in the bucket" under varj,ous flight conditlons, tirat is' at
valious typical v€ights and speeds encounterd in aircraft service. lttis Proved to be
difficult and jrpractical. and the net wlng drag turned out to be nlf,re than antic-
ipated. The scheme was t]"ug considered to be unsatisfactory for the intended appli-
cation. This is an example of a "single design point" airfoil.
ltotice frdn figure 30 that the -7 degree cruise flap setting, necessary to reduce
the rnaximum carnber frcnr 2\\C to IEC for Cruise. cleates a region of negative carnber
at .85C. This causes a dlag penalty, simil-ar to what hE have seen prevj'ously fron
ref)-exrng the tlailing edge of the NLF(1)-02L5 airfoil ( ccrnblned with the canber dj-P
at .60C). reducing the efficiency of the airfoil. As before, the better choice would
be to desj,gn the airfoil witlr less canrber (NLF(1)-02I4), to keep tJ-e airfoil clean at
cruise- in short, a conventioal design approach. Ho^Iever, the airfoil ltoufd still be
a single point design. Accordj.ngfy, our concLusion is tllat the point of maximum air-
foiL thickness should be no further aft chan .40ct for typical general aviation app-
Iications .
This airfoil has been nEdified. however, by Team Nsrxesis ' and j,s being used suc-
cessfully on the Nernesis Formula I racer. this is a very special application' a
"single point" airplane, and thj-s suPporEs our concLusion that the airfoil is un-
satisfactory for mcst general, aviat.ion aPplications.
It is interesting to note that I'JASA's Cessna 210 flight test progran for this
airfoil, repofted in liAsA Tl,1 85788, was simi.Iar to the Poorly designed "ATLIT" flight
test progran descrj-bed above. On the Cessna 2I0, the span Iras increased frdn 39 ft-
to 42ft, nraking it :mpossible to ccnpare the perfonnance of the nelJ airfoil to tire
"base line" airfoil. ltre inescapable conclusion is that NASA was here again trying to
shov,/ that they can design airplanes better than the priroe rnanufaceurer, ratber than
bej-ng satrsfied with doing Iegitirnate research r\,ork. At }east, in both cases, llAsA
proved hc)\d not to build a general avj-atiun airplane, and this j.s lvorth sonething-
I guess.
9. REIATED SIjBJECTS
A. Blunt Ttailing Edqes. Al-I GA airfoils r,ave LEen desigmed with sharp traibng
edges, for uniformrty. Hor,rever, sharp-edged airfoils, on wings and control surfaces,
'.Drk bet'te! if they are cut off as sguare as possible at approx1rnately 998 of chord.
Ttre reason for this is that the trailing edges have finite trailing edge angles,
and this causes the flow to selErate before reaching the t.raiJ-ing edge anyway. So
you might as weLl cut it off. Ttre airfoil will stilL acts as if it ',ere I00t 1ong,
B. Flyinc Winq Airplanes. Due to the lack of a tail, flying wj-ng airplanes must use
zero-o'o airfoils, thus GA airfoils are not suitabl-e. nlrttrer, due to pitch stability
concerns. flaps cannot be used on flying wings. Ttrus, due to these tr.rc design con-
straints, the perforrr€nce of any flying wing airplane wiU be nrediocre ccmpared to
an egually sophrsticated canventional airplane.
C. Canard Airplanes. This configuration has sinilar design constraints. Fl,aps cannot
be used on the nEin wing. llrther, the forward surface rl[l.rst al\.Jays be rll3re heavi].y
loaded than the main wing, so tlEt the carlard wilL alvays stall first. Since the
rnain wing can never be allo*ed to stall. it never reaches its nexjmum lift coeffic-
ient, and j-s thus relatively inefficient. Ttrerefore, due to these tlro design con-
straints (no flaps. no-sta1l nrain wing), t}Ie canard configuration can never be as
efficient as an equally sophisticated conventionaL ccnfiguration. Flrther, since
full--stall landings are not possible, landing sPeeds are relatively higher than for
canrlxrable conventional airplanes.
D, Three-surface lirplanes. control surfaces placed in the taiL have longer rsrent
arrns than those p)-aced in t]"e nose, tius .re npre effective. 5o htly bother with con-
t.ro] surfaces in the nose?
E. leadinq Edqe Devices, such as slots and slats. th1n airfoils, and other airfoj-ls
15
that are subject to leadlng edge separation, can benefit frsn these devices. GA aj-r-
foils, hc,wevlr. ajre soft-stall- airfo1ls, htlich n*ans that lead-ing edge separation
jfiprovement.
is not a problern, hence slots and slatsS cannot offer much Ttlere are
tv,o exceplions, honEver. First, in the case of novable slats that nDVe out fldn t]1e
leading edge and increase the wing ;rrea I rrDre lift wil-l result srmply frcm the in-
crease of iing Secondly. leading edge sl-ats contribute a positive (nose up)
"r.u. offset the negative pitchj"ng nEnent due to large
-traiung nterrent that can help of tiese
pitching l-arge
of leading devices,
edge flaps. In spite tDtential advantages edge
with C,e airioils, the prir.nry effort at lift aug[rEntation should be \,rith effective
trailing edge f1aps.
F. Case Study, Rutan Canards, Figure 3I sho\ns that Rutan used fhe GAwt-l airfoil scaled
E 6TA{-rhick"==-?6;-6-ari-Eze. Ttre cAw-l airfoil, however, \^rith its high Gn,
has a considerable center of pressule travel, contributing to Pitch stabi-lity Prob-
terns. Accord.ingly, for the successor l-ong-Eze, Rutan chose a forward Ioaded ( 1q"er
Gn, Io,.rer c.P. tjcavel ) airfoil to j:nprove pitch stabil-ity. Ho^'ever
' The Long-Eze
airfoil is a turbufent airfoif- vrby? with such a 1ar9e, lightly loaded nain wing, the
natural choice Seenls tso be a fs"r Gn, soft stall laminar flonr wing. I donrt have an
ans!',er to t].is puzzle.
16
Postscript
As stated in the Introductsion above, tJre original NACA airfoils \,{ere never in-
tended for astual airplane use, and, according.Ly, "GA Airfoils" has been eritten
to address this need. tn retrospect, one must ask why NASA, in the forty-odd years
since the NACA airfoil uork was done, has rnade no effort sJmilar to "cA Aarfoifs"?
Hovr fiuch valuable tjne and effort. has been lost, and hc,vr nnny precious lives have
been needlessly wasted, because llA.SA failed to address this problernz They certain-
ly must have recognized the need, so why dj-dn't they acc? vitry did they fail to put
an end to the lrlaLh that the I,IACA airfoifs were astual airplane airfoils, and why
didn't they admit that l,IAcA's mistakes and crnissions needed correctlon?
It is trard to think of arr ans$er to these questions without beccnring cynical '
one possible explanation is ttlat NASA doesn't knctu hc'v/ !o do the job. TtEt is hard
to believe, ho,i/ever, in vielr of the astroncndcal budget tllat l.IAsA is bl-essed wi-th
each year to pay for talent and facilities.
Another possible er<planation is that, by the innEdlate Fost-I4MI era, the United
states had achieved undisputed rrcrld-wide suprsnacy in aviation affairs, in spite
of the "good advice" frcrn NACA, and we enjoy this nLmber one position to this day.
Accordrngly, NASA could have felt that there was little ne€d to do additional aviation
R&D hDrk, nor to reviev the old lrork, so they "lested on their oars", So then the
question arises, lrhy wasn't f,lASA's furding cut off at that point? And why has their
funding continued yearl-y to ttris day? After al-I, the agency was set up origj-nalJ-y
to subsid.ize an infant industry, which aeronautics r,ras in the 1920's. But U.S. avi-
ation is no\^r not only a matu.re industqr, but it enjoys a runarrday position of $,orld
suprsrEcy ! To c.ontinue to subsidize this industry with federal funds, at the expense
of starving deserving infant industries of the I990's, rnakes no sense h'hatever.
zurthenrcre, a nore basi-c, philosophical concern exists, In a free society, one
must al\./ays be on guard against excessive centraL planning, in rhich decisions re-
gardj-ng all-ocation of productj-on resources are nade by bureauclatic guesstinEtes,
rather ttran econcrn-ic realities. R&D !$crk is a legitinate production expense. and falls
in this sarE category, and rnrst be hel-d in check by the realj.ties of rnarket econcrnics -
For this reason, r{re must always be suspicious of gove!:rurEnt R&D as opposd to prj-vate
seccor R&D. Since the U.S. aviation industry is ncrvr undoubrtedly nlatu-re, they are
unguestionably able to do their ovm R&D 'aork, in the norrnal course of their business.
For this reason, the aviation R&D budget of NASA should be terninated ifinediately.
And let's not be rlLisl-ead b,y those vrho !,Duld argue that it is necessar!. to continue
NASA's aviation R&D to insure tllat ttle U.S. aviation industry rnaintains its position
of leadership. That is a specious argrLrrEnt, and ignores the reatities of li:rdted
federal funds. ltard choices have to be made j-n washington, but fortunately this is
an easy one. llreher. Let's not confuse the i-ssue by clairning that tIASA's rdork is
vital for nati.ona.I defense. Ttre rnilitary can cronduct thei-r crvn R&D, under their q'm
budget. Accountability for governrEnt spending insists on th-is.
l,lo one will be hurt j-f the NAS|C aviation R&D shop is closed, and the U.S. Tax-
pa.yer wi-l1 be eteFraLly grateful. In fact. nor is a good tijrE to terninate this Fed-
eral jobs prcgram. the currently bocning air transport i-ndustry can easily absorb
the "out-placed" 1=rsonnel resul-ting fron "dosm-sizing" f.lASA. Now is the tjjle to act.
Aviation R&D should be rnarket-driven bry Adam $nith's "invisible hand" of free enter-
prise, not subject. to politics and bureaucratic r^trjms.
17
LIST OF REFERE}JCES
llr
' ,'c'
'. '€'
.0.
- .).
j
'. .i.C.(r-l.
''9 (('
.\L'
k t( '' '..?
q-
'c.,f- q'
I9939ni6i
Oi=qaqaqrT9+ "ittIt'i,j-s:
338333bFF
Z66ddddd.Add
o-
{ ree1Iss:r?:==1::3:::::::::::s:::E$ii$$ss$*
o
? =r 5
?
t-
o
? .s(
to i- .il,
tl
a
rpe I'l
F
-',
o
EX iF
l{!
c& . \. . Vl . r. .i.
. J. t
t-
(t
uO
>z r... Y.
.t:.<
o!o }l(J
l_. ,rr .
a. o.
OA o. .-. ). 'a tlP.
4 E c!.! . \5. >. lu. .:.
2 d oo . :_. ir. . l-a t.
.g
Jg.
rl
? r_r- . o. q4. ..<t l!-
a rI a'f :5 . t. .J
't
o
.s., '\}' .{
!r 2
lr,
lc'E . t_.
o o _oo
6bl: .
€;;
c-i-i
2
0
'arl
Qd6
--d.=9:i==F6: aoQt99
<ss
()AJ ga*";$$E$$cA;; ==&E(F
i88 s:: R 3:i I1*i:i $ qRF F HR ;i*S**
I
)r lv -9 Fr e,uR E I
19
!
1l ^=
,TJ
;- qfl
tt Fl
OJ .Q
>=o
.-i
po I
'! :r
50
ii li +l
I
:,i
?l
sl
il -t
:l l
: q-Y
l!
I L \! 7
:
2 ') OJ
- !l t- SJ
;lil :&
;.c
?l it ;l
: 1l-l i:
d l
Y ?
q I
-p
lf
!,\
d!l
\D oa I 6
o o
lJ>
l!
a 9:.'
I
?
!t-
:
-1
o
\l
I
I
rl .t
'il \
1
Le
--:lrl
!l l
"I
il -1:
il
+YJ-
trI t1
L€Lrl.liE
t.oo1 : ,lO
.a
.io c *.1
\' 44,
'' <
P. r,;8 (++i) \
L \,.--'-
_,-_r=-\:--
--? - \ rr\F^r{ L,{€
4411-
,+*
a
t'l A <4 j.f .A
4 r?1
q4-tL /,sb .Lo
-?-
+'!r I 2.+e
olo L 4-+re t.56
-!-
llSoctr
tl*l'iu
* le-
This shors that the a.'rount of canber in IiACA airfoils
varies with thickness. due to Lhe faulty "sl"ope and
radius" r€ttrod of leading edge design. Therefor, NAcA's
t'est datn rega-rding effect of thickness on airfoil per-
s. {an ( ei,tso,'l formance is invalid, or at leagt conf-anr-inated.
CA r({t €
il
I
;;.;n
"tl
_,;. = 3.49t1o
r
(4+rLT-
\l
1_l-
il t'
rl
*, f,r.i- ,l\
a
t'
3
|., (2
lY
t!, {lfl
2- --tl
E Jq)
'6
z
fi"
I
v,
t
I
oo
\ fi!
! ".t
10
t^
I c
I CE
(i!
I -o
I
+ -l
,t I
co
:
J
t.
t" Lrt <2
I 1l
1t
(x
o
l
(l ,gt
CJ
c {l <J
t ol \
I
vrl
€l
2l
I Io rE
g
fl
(t
o
(u
.-t z
o
) t1
o
6 l{
i!
*
t{ t4
rc,
'd
(.)
!
t v:
,
o
l', b
6
=
F
r.l
rd o
{.
_t (t'
c
I $10
) Io
d
H
)"
F,,.
23
(rrrei or Paq::grq Vt: tt
---.--.--.....-__ tSl r *r<r n,rro'Lr
c.f : ."tC - !'=
,A'
a
x
' -+iot'- I
.t€ -\- \1"-o-
z'
i
a'
<r'
lt<.Ix;-
.rt
?o, )9e?o
. ( llrtia aLrn^l
t.Z
F rouRE ?
\ - ._7
lor'
e"[-,"-r(ki#)] -. r[+-.t1ftgt
NAcn 1la6 ^
6 -.-.
r.a 16
?rt'trtil]< b
24
.ooa
Rer - NA 5t 4 e,- /q7{ ttrlgA /17.2-
4t-tr Nhat rN- tz*{' t?lta.tl t14 7
-?^t
^{
q.
ry.
q-
EP/ '/ '7
HinXz,
ahor\t 2^ L
(.zo-ror
'z .G
7
- 1o tnotc. -!q
ail zr ctt
-o16
Ttris sho$rs rhat ai -eron fr ts ar, r a funr tion ol the
trai Iinq ec ;e angl l. (Sna l.Le, he rvy ai}, rron..1o; ds) .
rO20
o it a t2 t5 20 24 2a f:r . 36 6
lr mrE-€d3! enrlc, 0, deg. Frl'rRF q
Hlnge aranent parameters of plain 2-O ailerons rlth various chords
and values of +. Gaps sealed; cu,/c=al leron to ring chord raf !o.
u tf4.tu,'q--J
ta- 0 R rrrc,+ c-SFeED- VAIz,ooe ^h'P'es1u5 l,
=&
o
t8 4fii'u( -\
NA.A 1,3<
tlttt(-
6
I la+'t lrs(4'.5)R4
I zsotE
Nqcn
y6nn ftr
This shov , *. ,"r{u loadins I :,refS the ar*t ** I
[r,*r * l*ort
u,.] t.4(
5<to slightly.r Hs€vef , U a eitical remi-ns lrell alove t€ raDge of nD6t
s -ed
-
general, aviation airl Ianes -
I ti{L - rclnfee
r-\r- I
' TuRdvtrtf Fq't 'La,..la^ Fuo',/, -Llt rrAR FLou) - LAmtvaA F!. s,
't.s' C ,3o t/c -t...** @ ,!5 r/e -t ;r1^ (-,515 xfg -trr,. @, +o r/C
A A A A
)r. r elL Tarr, OvsP No Cusp Cus P N. O.tt? Cu se Nt OuxP
{ ,l_lL
?ro
ut '2-tL 2tL 1- t?-
OE
,$ LtS zts zt5 ztt
ztg ?tt 2tg zto
t{
-. {.r
;s 3tL lt7- ltL 3tL
^< 3rt 3 ts 3t{ ?t{
urt 318 ?rE 3 tg 3tg
E\,
\D ar
.fo +r? {.rz 4rz 1tL
OE 4t{ 4t€ 4r{ 4t{
lu
7w
CE
4rB +tB +t8 4rE
btL
(ot3'r 6l't_ 6tL .6lt
=€'
lU
btg l.t 5 6tt 6ti
{E 6 t6.{ 6te 6t8 6t8
\J
totg
=, (;r6 ff-ra)
Exanple- "GA 35A315" - C,eneral Purpose, Landnar flow. llo cusp, !Fd. camber, l5t
z.o
Reqr.Jacos AJo. C*naf
?o
C*ono (er) t V(nPn) < .1ooo
l R=
I
€,<- ,Lhrrno^-ls'r-') "9
n
D
o
J
;fi ,/iQ'\
\2,7
{ r,o
?qtJ€ ALL OK
lu {rC 3.t ,.
c< |
+6Ne t/-;\
A, Ay'r' o 3"r lrA
ll - \.-/
,
l 40x zl1-
J
F
.t o.>
7 @ II-s.=i;,uo
c4 +o 50 7o
FlrNo C *acp - (^JCH€5
Fleune rl
c
t
o
.p
o
c
lrl
.-1
\o
o
! il
5 ( 6 dl
U' o
rJ +J
o I il
al
-a
\o f;l
I 4 .-r I
o
n
)
o
u
{
q
(
$
trtduq,E lL
{FEEcf oF 5€RrF5
"
/''
3o-3 t{
-, t'/
/i
.'/
-'s 37 n 3t{
Lpi
ili
'lt
?Ar '.'
I + P.=Leb
. t/s q
t+-
lf, i:
ii, 'uJ,' '::!::
^o''- 't
:ll l- t<tzt<
ch
,*/'
c{
\-r---
,f]-c
i-L-ttez'{
to'tt
o
t i- otro
o( -- D e6a FFS,
I.l, -
{ ,0 to ,ot< oLo ,ot{ ,d!o 6t <-
'6o (o tt L- zh4 lqf
tx - 7rf A tAFo re S - {A Rrors !et,e5 fth r3
ErrEcT OF Cusps
*
PeR Gsps4 fdoo*-, ?tai r." (A,Rro;. ii)
t---t (tA 3?- jr5t ( qusP) l
+-{ 6n 37 4t t r ( rrlo C,"s cl
,1.y'',
'./'' ,,
LL
THtcx '
AtRio,r- D t<T re A- li1s.9]
c. f a* 2-LeL
,b 11 /'q
ii'
il -tiAlt(
/t')..
I
ii"
rl
I ln
J1-JttJ
it
ll
V=a
tr (cus p
tlors
! -, ./- 37 -g \
t.L
t 37Aot{ ;
a jl4t-tf .a
- o+37ALtt
et -a
10.
I @ato'.- \
N ol/€
I
FF€cT
;J
oF (a^ B6R
*.2'
,J
.'
-.
I
,a
t(,
rL
+
3 7A4tl
+/,
rg
I 'fRo-,t' ,1."
t..
tt+
P€k E?PceA PRoaeA t
l-A,aro,,. I1 ') ,;/ --a \_\
| a-a- -a-
tt
t.+
,t\A
tb , *-';
* __ g"+
,,,s'
-T Aat{
t\
w ar,'
t.z /
+
,.\ .^
1,,
/
+ .--- -__-_-o
4, ,:t{r"j 7 Aott
Ilr i1 / t.q-+-.oJ!.L
t.0
,"t /
ol" _.-t /
lit( l'
o'
,8
lL
U
,tr
,tl p, ztloa
\,,
f; /
.L
c;,
AL /'b t,' /"
l,'
zl it -
/
A, bx
{ I
lA-
OL
37 Aor{ rls
tri* ha -1".,t+'7A2't
) t'f'
t+
II
I ta trt
l,'-'-'-'-l*, ."'"!'''
.L
l{
.; r
-11_!3!t
._1_
tl
t{ zl Azt{
So{l
d=
,.l1-- -.,1--- 4.., . -
_l:!\ Ll
v
t'
jj::---:
i
6
De€' R€E5
. rr Ar
tS
?^c1
'-J-
Lo
/zu i
-J ?
6A31 A9tO
'
(L
lli'o=,",
ioa -
llrr- I I
C".,r., !
,L
,/{ K\
''tl \:l-.*
lli'
,to,lLo
I - D ec.eee s.
t\
$:.-L
AE- +lzY/rr
,o tf LO ,OL< .ot o
(.D
GN 3-t A3l X ArRFotu5 ;th lb
t.e
, F. _J,,.robi lrrpE._ oF RE or-nr /o. I /6
o R= Zx to- ,, t? '- a- -
* A =6*too lruz' -"'- + -t .
--
a Rr to1<r6b ---- il/ -t" ,-'--;
't7
/1 I
,' ,/' ,t
r/i
"/"'
tr
.i),
t.+ o
-t ('
," tt tt
11= la "to - .l o,, ../
l\ t3
-/
,\",.' o
X ' 6x /o-\ /\ /',t
I. L
-t
)i'
.t /
''/
tt/
,/
o
,
,/
o./
_!,
./
'\
l\
\
'\_f=,,.,0"
t
bo ) -R=ltIdL
ll,' '
-t T
1 to.
l\i L ct+
illr
lrO I
d --6
i/l
a+o o
llr
,L zo
-5 6 t to tU
lO I .I
-) o( = PEER€Es
l.'[, 1 zV
{ Or.
U
'! L-o\. - zl
-.t rlllllll
O .oo{ .oto 6 rt n ,Oro ,tL{ ,oro ,oSf
\D lrA-zntlq{
hA 31 A 3ls 4rRFaru Ft6 I rl
- 32
t.b
f/c-
! n-
t{' c P I Le 2, l'ao 2.gAr4 li rtt.r or fo'o"{"e" I I
al
.
(A rc sor. ii )
I,t
R.6rrP-(-O
,A--
-(T",*)-
z* to6, y : +
'* ---
,d I
A
tD
ch
cl+ I
--o.&
lo rf zo
a(= 9F6AE€5
o 'ott
Co
FR. - zl|tl.1q__
r.A 4i--A -Vt< A
''a'cor- C',,..,oe ra
33
. KII'FOX SPEEDSTER
Wirh a look relrriniscenl ofrhe.rcing fil:ine\ oftlre.l(l s. the Killbx Spcedslers{ruls itstlrff. Fl\'orah[: itA
acrodynl].nlc churacleri\ficr ard an fi{l hf Rot!t c) l: rigin3 l:j\. it lol perfunnance. Chssic looks ud
leadin! edge.ngrreeing xllo*\ il top sleed oi 5lintes ir\ rornxl rtlll \peed. oilerinf ntarimunr 3OJ-i,tz-
eiljcjcncy and sllcl-\'. The Specdstcr refre\cnls aarl spo pe bnnrncc ;rnd lnluL :r, .r prirc r'rrr r.rrr
lflorrl. Worid renown lerobaljc pilorJinr F.! klin dcnto rtralcd thc Spe.dster xl O\hKosh 9l ard
srve ll rale reriews, crlljng il".'\ lrcrl jiltjr llj$hne. I $dnl onel '
?e,acen]
I arn intleed using thc GA 30ll-315 airf,oil on lhe Peliean < ll" > and wo arc vcry happy l:ith
ir-
Jcan-Iicnf Lepage,
P.ling^
Presidont
soN€{\Rl
FLETCHER BURNS REPORTS ON HIS SONERAI II WITH RIBLETT AIRFOIT
From lhe Soneror Newslefer ( '\e /.r' a7 uAci (l -ztz)
1
The wings ore thjcker thon normai ofter o beni stolic wiihoul a wrinkle. Thick€r ls strongerl This is o
\ring on my woy lo Oshkosh (which ls onother generol oviotion oirfoll not o pylon roce oirfoll. ii
story). i corne up wlth o 15% ihicker oirioil lrom lcnd! muclr softer now. climbs belter wlih lwo
Horry Riblett ond I love ll.
don't ihink it slowed me
I cbocrd cnd lhe roll rcle reolly improved (oihough
down ol qll. l've sondboQced them down to I G's i.i didn'i need ta).
hh 31 fr3t5
JAt/ tqss' € A n Technicot counselor News 7
f t&uR€ l1
?1 t
!t {+
ln Januarv oi 1993 Ken Fogers called me and iold me that however, Ken iell ihat in lhe long run it would be betler for
he was disaopo nted with the perlo'mance ol his Subaru him lo invest his etions in a new wing so in October oi 1993
Leoacv oowered Protech PT'2 (see Conlacll isslre t11 ior he tore the old wings apart, salvaged whai he colld and
en;ine rnstallalion oetails) ln thrs same issue of Contacll began bujlding a new set oi wings using the GA 30U-613.5
wa"s an a.trcle lhal I had written describing on of my airtorls airfoil.
GA 30U-613.5. This airfoil is designed io be a replacement
lor hio'r-li+t 'Cub" type applications and Ken wondered if it The new \ryings include a set of Hoerne. wing lips that I
miqhinot be a good choice for the PT-2. I assured him that sketched for Ken; lhey result in the wingspan being reduced
it ;ould be. but suggested that since his arrplane was by 16 inches. Since the new airloil has a di{'lerent lift curve
comDlere and {lying. there may be something that we could that of the old wing it was necessary to decrease lhe angle
do to "patch ..:p" the oid wtag Acco'dingly Ken sent the of incidence ot the new wing, and also the existing horizontal
details bf the old wing io me, and I made a perlormance slabilizer, by 4 degrees-
analvsis usino the NASA "Eppler' code
(pR6f trt, com-nercialized for PC use as'Aidoil
it" Tni. ir a simolecode, but il qives e\cel'e4t
resulls, especialiy concerning laminar,turbuleni
transition, llow separation, elc.
Sefore After
Cruise speed 85MPH 100MPH
Top speed 90 11o
Stall speed-no flaps 6tr65 45
St3ll speed-w/rlaps and power 50'55 35
Rat€ ol climb(FPM) 5m "betterl"
From thcse. quantitative valucs of lifi, drag, and pitching momcnt can bc calculated
as follows:
L = lifr = Clal Sw V2, pounds
D = drag = ta* Sw v2, pounds
M = pirching iooment = C.f 2 S* V2 C, foot pounds
Thc tcrm p n Vz appears in all thrcc basic formulas, and is somctimcs known as
"dyaamic pressurc", q, pounds/ft2.
Notc that thc pitching momcnt (in wind tunncl tcsts) could be mcasurcd about any
refcrencc point, but for conveniencc and standardization the quaner-chord point
(C/4) is used. A ncgativc pitching moment is nosc dowl. Ccnter of Prcssurc =
CI.=Cl4-[Csr/Cr(C)].
Thc flow coDdition (Rcynolds number = Rn) for cach wind tunncl tcst is mcasured
and cootrollcd. Although Rn can bc dctermincd cxactly, for practical purposcs it can
be approximated as follows:
Rn = 90ffi (MPH) (wing chord, ft)
Thus, an airplanc with a 4 ft. wing chord flying at 100 MPH has a Rcynolds number of
approrimately 3.6 x 106 (3.6 million), somctimcs writtcn as 3.6E6. As wing chord and
spccd decrcase, so does Rn, and performance drops slightly. Pcrformancc drops
rapidly bclow R-n=5E5.
Test rcsults are usually givcn for wings with iDfiDitc aspect ratios. No significalt
crror is introduccd down to an aspect ratio of abour 7, bclow which thc data should bc
adjustcd in accordance with proccdurcs found in standard lcxt books.
Most data is takcn rrith a pcrfectly smoorh surface (r=0). Incrcasing dcgrccs of
surface roughness degrade the pcrformancc significantly. A surfacc faclor of r=4
approximalcs a wing surface wcll contaminaled with bugs, Laminar flow wings
suffer nost.
Usc a "wing cfficiency" factor =.85 for non-uniform span loading.
FrcuRe L l
I'
',ll
' 'I
tt ,o
i
\, : :li
Ib .>r C!
(
'l:
q.
:uE
6
? -r :<t
r-< i=
A 2 <J: t6
t- a rll{
CT
( tl J$ra
\t e
U
t
jt
tU
,;
i; ,, *
!
L
II
€
vo ttlw - ozrJg,
"9t/r12
u
o/
rl.,
q-
)-
6
<.)
:l
v ;
(
'-f Ll
J t o
, a
I I
o !
q, ! ) b,
-rt
{J a t I
rt U o
s J c
5 2_
J
'2 t 7 Ol
2l' : I'.
<l t* t .: ir.?
:-l o '.,:l
to url
J o
t .) :
C
;:l
2 irr I
tr i!l'l
: i j
J,!
:''.'l
c : .'i
'G.r:
Bdw i .i-.J
z
I
fl
J
TI
,l AI
I
-,1>t
..1
a.
AE 0)
).
t
tr
al
!t I
I
"l
a cd lJ I ':l
E <r
,tal ll-r-"t:l
OJ
.tJ
Y o
I
.-t
J-l f ,l
\l
lf
5 ,)
dl <ir
*i 4.1
t
J ,t
0)
t)
5
o -rl\ -lt
.-l 3 f,i
-lo luq)
>5 t -tl ?'1
e
o
CJJ
0
t!:
i:-
(r d ,
-G;
J lc
( ro 31
u u
&
\U (f
!
; !l
'rt
t .-l
]J
E b lJ
v ol
.-t
.-t
lr J
\5 t,
o
d.
\- .tJ
trr
\u
t-
f,
q-
E
a
\J
rd
rr 0rl
cl
851
!ci6l
O..t'
-r 3i
q-i
(,
f
-T
38
YP tcnr- COilPUTER A
^JALYSIS
( Eppler program)
A I RFIII L- i i.
ERSIDnI 1- ?(l
proqram for design and analysis o+ air+oi1s
Fortions o+ this program are
COFYFIGHT 1985, 19Bb
by
A ] RWARE
A1 I r-r ghf-s re=erved
Ser I al l-rc-r - B6{:)(-,{:11:=
?1
A I F:FO 1LA a.l1 l: , ij(:i 7. TH I CF::NESS i:)(lil FLAF , OO DEGREES DEFL.ETT
nl.r
-1. i)(:) . {:r(:} 1- i:)r:) l. ii 4- (:,i:i 6. i ji,
B.':)i:' 1':'.i)i-, 1:. 'j(:' 1 4 . ,lL-t 1 6. ,jil 1{3. i:r(l
NIXY VELOC I'T'Y D I STF:I TJI,IT I NN]S FOR THE AET}VE ANGLFS OF ATTACI..
ELAT I I,JE TD THE CI'IORD I- I I.IE
(l 1, r:)':,r-,{-)i, . i-,,:,atr)al . E5A . S5B .457 . AEc) .849
,845 - gtrr .8f,2 ,874 .815 ..3(,i5
:l
.:-r+
, ?5{)( ,r)
.'!.__JJ
t-i
, lt'.t7 t}
.:._'-r
.:li
- /414,
' ::4 .::;:
. l{t
.94?
L a{){:rl:)if . {:)f 1:O - c7?r . 9{'1O . 985 . 989 .s94 ,998 1 . aii,,L,
1,(_)11 1.(-r15 f .i:r19 f .i-ii1 1, (:r?: 1.t)21
:i .851:irli:) -irtrl64 1 " r-i(:,g 1 ..(:,1: l. , (j: 1 1 , r'l; 1,i1trf, 1, r_'trE 1. (:r48
l.. | )1--ir' 1.tt7-j l.ar 7E 1.r:'81
1, (:i6? t..-t76
L. L),5.+ 1.r,l'r
4 , B{_){:r1ti:) .\) iija 1 . (:lI8 1,,14o L {:t54 t. t-rfr I 1. {:i8$
1,1(:r(:r 1.1li-r t.11q 1.1itA 1.1f,2 1.114
5 ,75i:rr-)r-r .r,)5:51 1..t7ii 1-(:)B(i 1. t:)89 f . il99 1. 1{:r7 1- 116
1.14b 1.15? 1,17i:' 1.lg(, 1,189 1.196
b .71:)(:)i:)i-r .':r6:BF 1.-111. 1.1:: 1,]f,: 1, 154 1. 165 1. 1B-1
I lr'ri 1 al? I :--'
7 ,, '55\..t1:ii:r . tJ77-=S 1. 14q 1. 16f, 1. 176 1. 1gB 1.:1:)1 1.?il r._ _
o;
,s x/c l.o
SUT4HARY AIRFOIL1 af,l5
: 2.79 DEGREES ANGLE OF ATTACIT1 FELATM T0 THE trHORD LINE
* ALFHAO
INDICATES BLIBELE ANALOG LONGER THAN . {)f,{]
R = ?OCICIC)OO HU =: R = 6l:)O(l{)C){:} MU =.1
ALF.HA =1S_1_Cro DEGREES
UF PEF
TURET S SEP
, 5{j45
CD
. crr]34
? 5. TURB
557f, S SEP
,
CD
. C){:r51
- C,C)O{:| C){rr_x_r
LEIIER .5865 . OO(IO . CrO?g . 5944 . OOC)tt . (:)f_125
TOTAL CL = .142 CD - . ol:t64 CL = .142 CD
= , Q(r56
cM = -. ('529 cH = -.t)529
ALF.HA = , {II1 DEGREES
l STURB
UF'PER .517E1 S SEP CD 2 S TUREI S(jtlaj(-)
SEF CD
-O{)(){r . (ittf,7 - 573f, . . (-){rf,f,
LOWER .541C) . (:lcx:r{:} , (tct?7 .5845 . OCI{-)C) . Qtr!4
TOTAL cL = .25? CD = . (i064 . c\:t37
CH = -. (1559 cN = _,(:,559
ALPHA = 1, O(r DEGREES
l STURE S SEP CD ? S TURE S(J{rctr-)
sEF CD
UFFER ,5f,49 . (){:)(l{:l , (:rcr4 (:) . SASS , , C)4r=5
t-OltEF .424(.t . C)C)(-r(r . C){)25 .3723 . Cr()OC) . c)o22
TOTAL CL = .362 CD = , ()a-t65 CL = ,f,6?_.._1589
CD = . ctct17
Ctl = -,{l=B? cM =
ALF,HA = ?, (rr_r DEGREES
l STURE 5 SEF CD 2 S TURB S SEP CD
UF'PER .549f, . OOC){} . o(r4= .=97t . OOOC) - oof,B
LOU]ER .3Ct77 . ._)r)OC' . oo2f, . 557r-) , C){rCrO , r-)Ol1
TOTAL CL = .477 CD = . 0.-166 CL = .472_.1:)619
CD = . c)c)s8
CM = -.('619 CM =
ALF'HA = f,. C)Cr DEGREES
1S TURB S SEP rJa_r46 CD ? 5 TURB 5 sEP CD
UF'PER ,565C' , Cr(,()C) . .6t.t67 , OC)OO . oQ40
LOWEF ,3734 . ()ClC)Cr . c,c) 1g . . OCIC)O , QC) l9
TOTAL CL = ,5AA_,CD = . oa-t64 CL=4?A .5El2 CD = . oc)59
= CM
CH = {:)69() = -. o55O
ALFHA = 4,Cr{' DEGREES
r STURE S sEP CD 2 S TURE 5r-lOOCr
SEP CD
UF.FER .=794 . CrOC)tr . of-r49 . E49t . oo45
LOWER .3777 . {)O{iO . (r(r16 .3772 , OOC,() -, oo 1B
ToTAL CL = .692 CD = . c)o63 CL = .692 CD = . C)C)64
Cl4 = --(t681 Cl4 = -. C)681
ALFHA = 6. O(r DEGREES
l STURE S SEP CD 2 S TURF S.r sEP CD
UFFER .6Ct7t - {tc).-){r . C)(r57 .9476 . (rO.-)O
196 . cxlB4
LOhIEF .24:E - CIOC)() .c)o12 _3756 . . C)CtlS
TOTAL CL = .912 CD = cL = ,89? CD = . t (t97
Cl4 = --8t744 CM = -. O696
Ftduae a4'3
41
SUMI,IARY AIRFOIL af,l5 ANGLE OF ATTACIi: RELATIVE TO THE CHORD LINE
* ALFHA{I = ?. :9 DEGREES
INDICATES BUEELE ANALOG LONGE R THAN . (:,f,O
R = 2(r{:)O(:,rf(j HU R = 6l:)(:)(:)0(-r( ) MU=3
ALPHA =
1 S TURE S SEF CD
UPF.ER 9812 . Cr4L9 . (rE
, a-) 1
LONEF .296L , (rcr)(r .(:r115 . Qcr 10
TOTAL CL cL = 1,._)86 CD =
CM = -. C)699
ALFHA = 1(J.
I c S TURE S(:)822 SEF CD
UPPER 1, O167 . . (11 f,6
LOTJEF . 2081 - (]{:,{_x:r . C)CrC)7
TOTAL CL cL = 1.?52 CD = . c) 14f,
CM = -.Cr631
ALPHA =
I 2 S TIJRB S SEF CD
UF,F'ER 1. 1. (,1E}6 . (:r(rC)C)
1294 . C)16f,
LOUIEF .1196 . . {Xr{:}S
TOTAL CL CL = 1.4(lO CD = ,.-1169
CM = -. OEB6
ALFHA = 14,
1 2 S TUREI S SEF CD
UF.PER 1. I.t)797 .1gB(r .t:r!97
LOUJEF . (r69C) . {)Orjrl , O()(-)4
TOTAL CL cL = 1.52? CD = .Q?(:rl
CM = -. C)5Of,
ALFHA = 16.
I ? 5 TURE S SEF CD
UF,PER 1, C)?10 .2545 . O2f,9
LONER . (Ja-)oa-) , ()()r-'a-) . otlll?
TtrTAL trL cL = 1.619 CD = . C)241
CM = -. (:)4?6
ALFHA = 18. C){] DE G
1 s TUFB ? 5 TURE S SEF CD
UPFER 1. Cr23{r 1.C):S5
()o(:)f-) .. f,f,l2 . Ct297
LOWER
TOTAL
(lalt)o
CL
_ o{x)o
CL = 1.6€l? trD = .(t299
. (lcr{]?
DH
0 DetrCH- =Z-?s-lr9{'
-, Of,7(l
6A 3-i A3t{
^L Atlo
KG
!,i ------ R: 6x roe
-
l.o
cL
Ct'tr
-a,lo
- o,o9
f taueez4-+
! ------ |
o oo
ta
| - €ar
"J ^/
ooo
l' r' I
a a c
a o.
U
I
tr
o
N
o J
€o q
t) a o
o o o o
]J
! ? +'
t\
otr
o
+$: J
x
?3i
I
ctL0l€ L€
43
'1 <J3
t!
E
o t\ r,
.-l 2 I
!t *.s;
LH
!r0
rd
C)
9l
il:t
!l ,-' ^r'
!t
lr(
.lc IE 3l i j
t .1 {- ':.
J- ts
F
" ri
o :l t: !
!<
El: I TJ
C'l il
:l 'i ia
N
!
.'t I o 5t
.l ilil .:i .l
"
il 3l :. ;l{5
ii
I
6 li
ro tl sl
l-c
i .tJ
,,
J-l
t
-|-3-
.l
ti: r€
.rJ
t.
o it I
r+ l,l
Ia o
o
o
6
)
!F
0.J -id 5-.
{J-+J
lro :
o.'Jl! I
cOiEO
u q-.r
d\
'.{!O t'
.]J +J 8
9
r.{ {,
O'E
a. I tF
UFr
T a
J
I z
'* {
.r.J 0)
lo
rJ
l.
16
3\
()
a
'{i
\i
..l
'
?r i
44
NASn an (w)-l ArAtrarr- fus (')- o+rrl
t_
2
CAr4B€A-'/- ..-
,5
__=i
I
j
I
.t :
I
-/-
U
><--
-.t I
I
-.?; ,l .2 .3 A.5.6.7 .8 .9 t.o
r/c
lq'r4 Figr.rre l.- Section shape for NASA GA(w)-1 airfoil. (r1% rr'rrc<J
IRST FLIGHT OF SENECA WITH MODIFIED WING SCHEDULED - National Aeronaulic: and
l$ad-EAtnin-istrarion will supervis€ the {irst flight ol a Piper Seneca next month modified with a
-Snew
wing design that promises major performance improvemeflt3. The wing assembly for the air-
craft - designated as the NASA GAW-!.- is being built and installed on the aircraft 8t Piper's
Lakeland, Fla., facility. The low-speed airfoil features full-gan power flaps and spoilen for lateral
control. Wind tunnel tests with models have indicated that the netv airfoil will provide inqqelq! of
90 oer cgnt in range,80 ocr-,cent in the single+ngine-rate+f+limb, 30 oer-ce?t in maximum lift,
g9_gg. c_elt-tgup and jgpzullt in cruise speed. Although the rving is being built by Piper
a;E install€d on a Seneca, NASA spokesmen emphasired that the program is designed to provide a
more efficient wing for a wide range of generil aviation aircralt, not iust the Seneca. A full flight
resr program will be conducted at NASA's Langley, Va., research facility after initial flights are
completed in Florida. " AT u, r' fAcLtec-
rruftG Ln
c
/}
-!)
:l1
.-,j
I
Ic
-.1
a
"!O
-u
I
U
io
J.
rJ
tu
..)
o
:
t-
J
P
!s! --
l' ILt L6
k-1
at
a0
!)
tl
+
,",1
() ?
CJ
tb
!!
CJ J
00
or .t
rd6 q"
\
5.
d^
!!r u ---
.t J
ort
r0) I
7 )L ,--
va
+.l '.r
']Jo .-3 o
\3
UO
a,
..11,
OX
\r
tc
CJ
Cr*
*
o., 16
1]U
6C o
l'n! c.'
J
x
\h
lr.
J
s 1
tt *)o ''
{1
€
.9
r0
u-.
t, x
tJ
-,". t-
,
._:il
uJ
i'*q
- ltt
'2
Ctr- ? 4
47 o
N d
tlo
t\ -
o
o- t--
|-\
I' ---'
I
4 ,
l, ,-
l'.
\-- /
L a
]4
G OF+
!|, :t- o
.J lz()
e o $E
U
q{ otr
J
€ !c
=- \9 ,e&
o
F
lt .-l lr .
4C 'ocrr
qc
ll
d-
od)0)o
rt >
0l rd
cE
..,i'4C
c o,o
otrk
t rd .,i @'
dFr
jrc 3 .cr'd
o!6
JIC
Erd
lO..l
.,t lr o,
lJ 16 o
6-o-q u,r
'{-r O
"Ef
rd!
_t. .,t-
.'.a 5
u)
q) I
F tlt J)
EOd
.-rl{o
"l
It (u-.{
'-{ ! O,
n
J
,dE
f, OO
cd .d
o,o-.1
! r-l _c
rlJ o'sFr
w o0J.
!oo
._;.1_E
t- u|r ..1o
tr lrO .lJ
_-l
t{ : .-i
l! rito o-c
o{Jo
.6CO
L qr
t a'c+l
0,.1
\T
lr
J E ti
q'l.!tt
..{ id O
E'A
{d >..1
or lr l]
OOJId
!>>
{
l.
t&uE€- 30
48
t'l-'1o T*,.( iJA cA 4 D'611- ANp t g|Air ArrFcret
oLD-E_L_AgEqlr
JE
of6 t ao&>
NLF(') - d?r5 F
l
Lo!6_E+€ fie (r/6
^J
4.O
I
s I
I
L(
I'o
NASA L!(t)-6+18
VARr E+F
( crau-r ,rr.o'r'el)
.-o I /C_ ,o t.
EtA
APPnDIX I - Bas.rc Thickness Forms.
49
T - 84stL a*rc<lveSS Foems synretrical- shapes
(
. l5t thick)
A ' "cusp' TrPe 7 Atc<+v€ss D tsrRrgurrorts
GA 30 Aotl
(nrRca oors)'
GA 3s-Ots
INACA Ll-ot5)
, QA 31- 615
(,vncn 6+ -arr)
GA 4o-ots .
{lncn al-ors)
-
o lo ?.o go ao ta Lo .ro go qo nc
llote- For thicknesses other tllan l5t, these sba[Es nEy be sca]ed d-irectly
up or dcnnr to the desired thickness.
f I6LJAE -L-I
50
a - BAsrc THrcKt.]gsl FoRmS ( symrEtrical shapes, 158 thick )
6A 30-o rs
GA 35AOl5
(ueca ut aors)
6A 37nots
(NAcn 61 Ao rs)
CrA 4oAots
Frh T'L
51
APPNDIX rA - ctrso Tv.pe Thickness Distributions (GA 30A, 35-' 37-' 40-)
Feb 94 0 015 63-015 54-01-5 65-015
sta Ord ord ord ord
0.00 0. o00 0. o00 0.000 0.000
0. 25 1.090 o.a75 0.865 o.830
0,50 I .204 1,208 L .724
o.75 1.856 L452 1.456 1.356
L .25 2-367 I .874 r.442 r.702
2 .50 3 .264 2.610 2 .528 2.324
5.00 4,443 3.648 3.504 3.245
7 .50 5.250 4 .427 4 .240 3.959
10.00 5.853 5.055 4 .442
15.00 6.681 6 - 011 5.785 5- s04
20.00 7,172 6 .693 6.480 6 .223
25. 00 7.427 7.L55 6 .945 6.764
30. o0 7 .502 7.42I 7 .379
35.00 7.436 7.500 7.442 7 -396
40,00 7 .346 7.473 7.49A
45.00 6.976 7.O99 7 .224 7 - 427
50.oo 6.518 6.665 6.810 7 .168
55,00 6. ie1 6.108 6 .266 6.720
60.00 5.704 5.453 5.620 6 . 1.18
65.OO 5.166 4 .72I 4.895 5.403
70. oo 4.580 3.934 4.1,13 4.600
75.00 3.950 3.119 3 -296 3.744
80,00 3.279 2.310 2.472 2.858
85.O0 2 .566 1.541 L.677 L .977
90.00 1.810 0.852 0.950 1.144
95.00 1.008 o.300 0.346 o-42A
100.00 o.158 o.000 0. ooo 0.000
APPENDTX rB - I,Io Cuspltlickness Distributions (GA 30-, 35A, 37A, 40A)
- a1.'5' -"
As:*A
\\r Fta
x/L ro .L 2.tah- I
53
APPmDIX rII - Airfoil- ordinates. \
_-- o I
i 0) ,_a
\ E
iE +J
.t
lg o8
8.f
o
c<
lu (,1
a
lE
II oo .-r
(to
lJ
o
' dd .ri k t J-
I +l
]Jlr
..r o :t
I
) ^'HY' ..1
6*
6
f
L)
! J 6J
I ?g
: !.r
!+ 4 r. :'
f t-
t cl r 8, ;l h
\l
+l t o OJ
\!;
E
2 sl'
a[c
/1,
'.c,
E
!o
zl l c) d,
c iz
2 2 I
+) .-t
Fl
?.
o
.l 9t{ -9
CO
1*u I
I
II I I
a 5ig=i gifrE= 0,
5S5t8 ]J
|!
2 zii-
!llo
o 3^
tlrrl ,o;E= i
fc
1) 9
=
so a
a l -Elr
F ,9 E d c:4 !Jtr
_o
.-l llct o . 5 6 0-- - g
!
.lJ
o
I
, i
g dHiS: !U)-t
o
Bl
ol ^Fit) tu
ri \'
OG
'z gigi II If
f
I
E
F o>
.iFtReda
-. x
5;;E;
F=rq3x
g!!
!r-O
of,..i
)Ll
cl
dl 6
Oo t
e333: .;. ; EEEE;
:: 5 c€. a.l
!,)
s8.I
irrd
+J
.
EI
., :6
ai f- li'
rfd
U
z sEe g 5i
9 o<
E .9o -u:
-.i I
.ll'' .i
tlJl o p
zz
\o e !-! s >r <'a s';
t! | aZE' A Idjtc
tJIO U
c)
ii
I
l! !.
El p t'> + \
I
PilFE
!J
l.l
.-i
tr I L,
d
OF E
lr I rt88 i I= E Er ol..r
o
ulc
c
fl'.g
jlel #u
ln s
Ec
oi !h
at
I
;3t t i:I E Eii:€ n
lo c
Irl-c .-l
.r I rJ
c
;ii ; EE€ E:
I
F a$ E:r r 8gle E
EI E
\' I
a!t t 4cl'n
!!t i i s:;€
T .P rJl
!' 1
3
I ql
11
Hl s t ele E
4 iE Ili dsl.r B
tli ;€Es,i
a
!t ,
Ia- t*[He
EI -. Cl-r.,
ot
rl l3$ a StEEB nlo r,
E[e 3
glt tFSFB
EI
F{'5'o!Il.
eta fr-I
q ...1
6 o
-c
o .-l
.lJ |!
o o
LN !
-1
ro
o l.l
q
IJ
Itr
U
o
.l-,)
c
a
a
rd
0)
!
0)
ol
16
o
1 o
a
o rd
u.
E 4
-.1
3
a
\t ]J
t1
v o
v
I a
!n
c F 0,
E E
-c U
U)
t) * -c
+, l{t !
+J
o o OJ
c u'
rtt lr)
a6
ct a.t
oI u')
o
.t)
1 o
o
) tl
o 6
$
(
I o
0)
_c
.u
OJ
r-.t
l[
U
( o
]J
ll I
t
8
I I
0)
.lJ
2
\i\
r\
Fnf,-7
55
Feb 94 30-015 Camber 30-2L2 30-215 30-21-8
sta ord GA- 2 upper lower upper lower upper lower
0.00 0.000 0.000 o.000 0.000 o.000 0.000 0.000 0. o00
o .25 1.090 0.060 o -9J2 -0.812 1.150 -L.030 1.368 -L. ?44
0.50 0.1,16 1.338 -1,.106 1.543 -1.411 1.948 -L,7L6
o.75 r.856 0.169 1.654 -1.316 2,O25 -1.687 2.396 -2.058
2 .367 0.263 2-L57 -l-.631 2.630 -2.1o4 3.103 -2.577
2.50 3.268 0.450 3. 064 -2.164 3 .718 -2.818 4.372 -3.472
5.00 4 .443 o.715 4.269 -2.A39 5. 158 -3.724 6.O47 - 4 .617
7 -50 5 .250 o.892 5.O92 -3.308 6.L42 -4.358 7,r92 -5.408
10.00 5.853 1.023 5.705 -3.659 6.a76 -4.830 LO47 -6. O01
L5, o0 6.681 L-244 6.589 -4.101 7.925 -5.437 9.26L -6,77i
20 - oo 7.I72 L .420 ?.rsa -4 . 318 8.592 -5.752 10.026 -7 . L86
25.00 7- 427 1.557 7.499 -4,385 8.984 -5 . 870 t 0.469 -7 ,355
30.00 7 .502 1.663 7.665 -4.339 9.165 -5.839 10.665 -7 .339
35.00 7 .440 r.737 7.689 -4.2r5 9.L77 -5.703 10.665 -7.r9L
40. oo 7 .265 1.780 7.592 -4.O32 9.045 -5.485 10.498 -6,938
45.00 7. O00 L.?92 ?.392 -3.808 a.792 -5.208 10.192 -6.608
50. o0 6.655 L.764 7.088 -3.560 8.419 -4.891 9.750 -6 .222
55.00 6 .240 L.672 6.664 -3.320 7.9L2 -4.564 9.160 -5.816
60.o0 5.755 1.537 6.L47 -3.067 7.292 -4.2I4 8.443 -5.369
5.190 L.374 5.526 -2.774 6.564 -3.816 7 .602 -4.854
65. O0
70. o0 4.540 1.189 4.82L -2.443 5.729 -3.351 6.637 -4,259
75.00 3.81-7 0.991 4.045 -2,063 4.808 -2.426 5.57r -3.589
80.00 3.053 o.793 3.235 -1 .649 3,846 -2.260 4 -457 -2.A7r
85.00 2 .290 0,595 2.427 -l -237 2.885 -1 695. 3.343 -2.153
90.00 L -527 0.396 1.618 -O.426 1.923 -1.131 2.?2e -1.436
95.00 o.?63 0. r98 0.808 -O - 412 o.961 -0.565 1.114 -0.718
100. o0 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000
FIAUR€ fi.|1
LA55r( * rCrF{- LrFr A rRFatLS 72
C ( F oa. (o ^^?ARtso N a^iry)
NA CA 4 4I'
FlQvP,€ fl,--zo
APPENDIX IV - Aerodynanric Characteristics of wing Sections ' 73
A FPelv olx U'
a
8.
rd
-c
U)
q
!\i
.-l c
oc
fd
--lr
c{ L
-t
o q
ltt
(!
C
a6
!{
qJ
.t)
&
6
o
U
IJ
q.t
t
o
q)
(J
: ":l
3rl
Yl
o
g
4 :. iltt
o
q-]
_ ,il
t{
0,
i.r*! .sl
I 'l
E
.lJ st
O
..t (l I
5! i! "ol
6
I
!.r
9!ltJ
<
:l N,9t
0rl
!l
di
'dr
Ft(19-l
;
II
El
:.t
"l
rl
sl
il
11
*l
!l
;l
l
5l
'g ="t
"{/ t
-.J",
JJTS
Yc(:-:
!J*
r,.- 19-?
75
.ib
l-i I o
;o
c*
tt
:I
r"Y
I
t_ el '
"t o ol
I
ol
\_!
i ol
l.ll
5l
tl
,!I
€: g
.""--. -
".=: t:il
:r,-?,l
.'"\..
<< ft ; o
U
*l
!:o'
a
76
/4
,i -\t
6 u .x
'rL
e1
6-ai-
1<'
\9
9+'
t; c
.l
:.e
ll
o
aE
!,
;
10
\5
, i2 )
.r -t I
n.-'I
cv,<a- *
:l
il
hl
i:o
? I ,.r" i o il
;l
i.!l
oc H i
..<i
_9 .,1 C
o
6
-----+ f
lr_
"o €l:
,i,':!$.r l
!{
{
!"
:
l
.a1
'ti
:l
>l
tl
('
u_
Il
: "l
OI
al
sl
_-
r, -$ i
!
o
q
a
. -l c-u
9Jq
:q
'l
;o
-<(
-oO
x! i.j
.ls
ltc.fl'b
79
:
JI
fl
o
.t
"!l
"il
i ;"i
;l
.2v'i,6
**€
o?.ra
!
e
!
;-g
<id
L<
;14
\'
-it{ l:
9
lJ^
,. 'e It
i <i r13
li
cr .**
80
\'
il 9
:;
<
gh
-,0
{
9Y -
. +o:
t+
LI
lt 1q q
t\
o
o
art i
J
,) X r,.
\.. (
.t
;
t
g
i-r
''t'-^'
i (
.-;=
"l 1G-
81
:l
-j
",I
\n !
v{b :
-a
Y a
\5
tl
rl
J^
NJ J t
82
e
:
\6
A
i.
6
\r \]
F------.r---- ,l
.-i
\'-
1i ;t;
\5
I
-'\ -
li .s
Ir 6
:3<;_ "-'../ .,
)-:
l!
-l
U-e
;
$
.t
ul JJ
sl o'
+ A ;
vlrr
!
-s
t_ {
i \5
e€
cJ .5
o((
^\-Elr
...83
_9
;._
s
tl
I
;
t, .:
9-r it
}.
i !
t
!: 1(
P
)
6 tt
!,
FJ .9
f Y
UJ
l
:r
:
l
'/l
;'l
t-
E
t
4
o
t-
I
,5
.9
H
-. s q1 6
d 1,, g
.lOs. -! <t
:
\'
I
o
:i! ,rr
i
s. t_
-'
:
<l si\l
zt= .l
>r ;,,:
Y
JI
al
U
I
I
ql.t o ^ {l
dl
oa
1"1
Fl
\-
:l.,1
a
tt
cl
!-r
1 . t'l o
, ar ti
\;_
tl q'i
8s
{=.l
:J
cl
//
JI ? ht
I\
)
-l
.lal \t'. (!
ul \ cl
,5
.ro,yS
tj-
'J
i
1
q
r
d
v<f
-{ tt
':'l i5
'+ ir
86
I t
c
" rl
.ri
.-l
:TI
'fl
,l
T
:. -
9r
o
) o
o
!
{
t
(
.q
*-
'e--{-
J-
87
{
!L
t)
dl tJ al tl
ol
jl :r"l ,!l
* il
.ir -'l !
I
6.S
_:t_9
j;l
- cl
: Ft
-t
$t
r-e u:
-Jl
,, ,.:r.
cr(u<
-tr \t-tb
fj
-l
tl { ;l
I
:l.l
S
: :l
;l
-" r^l
-t I
:;l
-ql
ol
t :l
tsl
"" {
u9
F1
i: i i
.+J
rl
5l
ol E
{
Io
o
,tJ
5
H
!
.!
89
,-r
<i
----
1': !
tll
.l )
!
d
4_
{'l
i;Yt
**; :
ath tv' l a
J
v;
o
5 .5
+
9 t
.t I
aJ
r.u
'{l !6
\'
JI
a
at : il
ql
s
f ''-l
o <l
-l-.
,51
- ".: l
:3i f
)-o-
,.1n-ta
a
,.9
tf
-ldl !
1
a :1
o
:tl ;
c€
U6
.Js -! o
\3
.i5w,'a i
,' '. i
q(s!E
r!g
- cl
",
t c_<
!
{l
cl
-?i
o
a- ,-9
;+
\'
:l
ro'9s;i
: r< !
.!:
"92
:
\.- i?
i?
tl li
\: 1 rl ?/
-.,
a
d t
\:
I oo
a o
\.,
)
\: \D
{
l r-
lf H
\'
d.,
dJ,:
(g
s l\.
-;.--.j\
t
'i,
-;
--,
...i3 -s :=:-- :,/-
I
9. t
ho
JI
d
!l
:l
i i-," F -l
{l
Jo <l
ol
+l
<l
.Dl
99.r:
r{!8
,? {,tji
"rc('!{
Frt's'-10
ADDE 'Jl-.ri'r LrUi'rEER l 93
Underst€.ndinq lirf oils
Or, '' nverythj-ng you always wanted to knolr about airfoils but \,rere afraid to
ask'r, for fear of feeling stupid, perhaps. Donrt feel bad- there are scme confusing
aspecLs about airfoil perfonrErce, and a surprising nw cer of stupid airfoil design
mistakes have been rnade in the past by people who should have knoarr better. 9're will
point out a ferr of these, but the prirnary purtrEse of this artsicle is to "get back
to basics" so that you can better understand the nain causes and effests of subsonic
airfoiL trErfol:IrEnce,
SwrrEtrical Airf oil Perfornance. l,ie ' 1l start r./ith a sinpl-e case of b^'o syllnEtrical
( uncambered ) airfoi.Is, I5s thj-ck, sho,.rn on figure 1. The first of these "basic thick"
ness foms", NACA 0015, is considered a "turbulent flor^r" shape, while the second,
lrAcA 5rtA0I5. developed later, is a "la$inar flovi" type shape. we dril,l a hole in each
of these at the guarter-chord Foint to receive an axle, and then nDunt then (one
at a tinE ) in a wind turlnel such that tbey are free to pivot on this ax1e, and then
bLcn on thern. The airfoil is then tested at various angl-es of attack (C ) frcrn zero
to about 18 or 20 degrees, plus and minus, in one degree j-ncrenents, and lift and
drag are recErded at each angle of attack. Lift and drag are easy to visualize, rneas-
ured in pounds, aqtsj-ng at the pivot point. Lift is neasured perp€ndicular to the
air strean, and drag is nEasured parallel to the air stream. The }ift and drag values
at each angle of attack are nohr converted to dirrensionLess perfomance coefficients,
CJ- and cd, consi-dering t}Ie wing area, air density, and ai-r speed. lie nexts plot cl
versus Cd to obtain tne "Iift curves " sho,in on figure 3. the Cl vs Cd cr.rrve is
sFnr€trical about alfa = 0, r*rhere Cl- = 0, since the airfoil itself is s1'nnetrical.
the zero-lift drag is called "profile" cf " form" drag, and the total drag increases
as angle of attack increases, due to "induced" drag, or drag due to lift. As tie
angle of attack increases bq/orld about 15 degrees, the lift begins to decrease as
f 1ov, separation occurs on ttle top surface, @inrling at tie trailing edge. ttris
area of separated flonr progresses forward frcn ttle trailing edge as the angle of
attack increases, causing further loss of lift, and large increases of drag, until-
the airfoil evenlual-Iy "stalls".
i,ie are not finished with or:r !.rind turnel fieasurenEnts, ho^Jever. on scarE arrforls
the lift and drag are not exasll-y centered on the guarter-chold point, t}lat is, the
"center of pressure" may be sdrF distanc€ forward or aft of the guarter-chord point,
and tlre airfoil terds to rotate about the pivot point. For this reason, rrE [ust al,so
nrcunt a sprjng scale at the trailing edge of the airfoil (see figr:re 1) to neasure
the pitclling nrnent in foot-pounds about the quareer-chord point. A nosedorrrn
pitching tendency is, by convention, a negative pitchj-ng firctrent. !,ie nc'v, convert the
pitching roxrEnt sprillg scale reading tj:res .75c) at each angle of attack to a
(
di:rensionless croefficient, On, sirnilar to Cl and Cd, by considering the wing area,
air density, ai-r velocity, and also chord lengrttr. vihen r,,e plot the p.itching mcflEnt
coefficient curve, qn^ /d vs alfa (figure 3) for these tvro syrrr€trical airfoils, we
see tiat ttre On is zei6'for any angle of attack up to the sta].l point, ttEt is,
the center of l-ift center of pressure ) passes directly through the guarter-chord
(
point. Inde€d, the center of Iift for anv s}'nretrical airfoil. is always at the quarL-
er-chord point, regardless of tie angle of attack. ltrus, this point has special sig-
nificance, r*rich is the reason tbat we picked it for the ncunting point for wind
tunnel tests in the first p],ace. the explarEtion for this phenc.rEnon is beyond the
scolE of this article, which is anotler my of saying tbat I don ' t kncr^, ra,hy it j-s .
Regaldless, the pitch-ing ncnent cefficient about the guarter-chord Foint for any
synrnstrical airfoil is always zero.
Ianiner F1crd. the next thing to look at j,s landnar f.l-or./. The ai-r alf aror.rnd us in
FnaturiFinraisturbed staie is laninar, that is, in layers. rf you ncve your
hand slor,ly in fron\ of you, holizontally, tie air renains in layers, and your hand
passes ttrrough snrcQy with mimmun drag, without disturbing the layers. Tlr-is is
knorn as " la,rLinar flor.r". If you rrpve your hand rapidly enough, or if your hald is
rough enough, tire flow over your hand wil-L "trip" to turbulent flor*, rrrhich has rpre
than twice as m uch drag as lanjrnr flow. Ttre reguirerent for .Larnj-nar flo\^' is that
the flcrvr over the surface should not be accelerated beyond cerbain lirn-its in any
d.irestion- fore and aft, sideways, or verticalLy- tlEt is, we need c\onstant velocj-ty
94
fl-op over the surface. If the flov/ has tripped to turbulent, \,vre need only to slct/
doh/n to the point h'here the acce.leratj-on lirlits are not exceeded, and the fLc'vr will
once agaj-n revert jrnrEdiately to tanr-inar.
!€ se€ that the 64A015 shape has a region of lovr drag betvJeen -4 and +4 de-
grees angle of attack. this lo.r-drag region is knov,'n as the lanr-inar bucket. Beyond
I 4 deqrees, rr)St of the laninar flow is lost, and tl.e ftor over the entire wing
beocrrEs turbulent. Iaminar flovr is easiest to maintain near the front of the air-
foil vtrrere tie pressure gradient is highest ( "ncst favorable" ) , and is hardest to
maintain beyond the thj-ckest part of the wing. Accordingly, t].e " Ia.runar" shapes
have been desigrned for constant velocity flov,r over the wing at cruise, and the thick-
est part of the wing llas been IIEved as far aft as prasticable, to maxjrize the ]-arn-
inar run. The net resuLt is that the laninar section has approxirnately 20t less
drag at cnise tllan the ccnparable turbulent sestion, a sigrmficant FerforlrEnce
advantage.
ScnE recent
ients than the I{ACA 6-series shaFes. but tleir larlinar buckets are carrespondingly
narroi,er. They camot tolerate nuch change in angle of attack without tripping to
turbul-ent flc'r.r. At higher Reynolds nunbers (higher speeds, larger airplanes ) larninar
florr is npre dj.fficult to maintain an!4Jay, so ttle bucket becqres even narrcr,\Er
with these airfoils, wtrich rnakes ti€fn jjrpractical for nrany applications. A1l in all,
tbe NrcA 6-series tllickness distrj-butions are not bad, having a ccmfortable bucket
width and depth, perndtti-ng wing twist if desrred, cl"jjnb at relatively slo,r speed,
etc. withouC running out of the bucket.
So nuJch for synrretrical shapes. turbulent ard lan$-rtar. |lexb ccrrEs canrber, or
wing curvature.
cambered Airfoil Performance. If rre take t-ire tr€ synrEtrical shapes d,iscussed
in the nr-idd.Ie, r.rc obtain the tvp canrnerea airfoils
"!o"e, ""E,ffi'EA-ij@
shonn in figure 2. The camber schedule is carefully designed, and is called the
reference (A). The canber schedule for GA37A3I5 is a nr:dification of another nean
l-ine frcrn reference (A), and is especially designed for lanr-inar flctor. The maxjjrrlrn
canlcer in each ai-rfoil i-s about tl:le saIIE. Also. both ajrfoils are forwarded loaded,
that is, the maximrm camber is located forh,.ard of the airfoiJ- rnid-point. at a.bout
.40C.
The perfonnance curves for these tlrtc airfoils is shornn on figure 4. Cc.Tparing
with fi$Ee 3, the first thing we notice is that canber produces a sigmficant in-
crease of naximum 1ift, with no change in crui-se drag. Second, tfie stafl becc.rEs
softer, that is, the top of t}le lift curve is flatter, and lift continues to be
developed at higher angles of attack than with the uncanbered airfoils. This is
prirnarily a function of the inj-tial slope of the rEan line ( slope at .25tC), an irn-
portant paraleter for good sloH-speed performance. this inj-tial nrgan line slope
has been optirlized at about 15 degrees ,rn a1l "GA' airfoiLs, correcting an unfortun-
ate crnission in the NACA r€rk. Ttre third thing that we notice is ttrat canrbering
Iifts the vrhole performance curve (Cl vs cd) vertically upmrds into a nrre usable
range of CI, with no increase irt drag. Airfoil perfornance is neaningless belo*
C1=.15, except for a few ve-ry high performance airplanes with extrenely high pcner
to v,eight ratj.os. AII other a.irplanes olErate above this point: for exajrPle, Irpst
Iight GA airplanes operate at cruise lift coeffj.cients betrreen c1=.2 and .4. I€ see
tiat our cambered aj-rfoil- perforrnance curves, and especial"Iy the lan-inar bucket.
are nohr centered vertlcally on ttre desigrn lift coefficient (C1=.3) instead of being
centered on Cf=0. thus the top edge of the larnina-r bucket is raised to nearly C1=.8,
i.€f I into the cljjlib range. the bottcrn edge of the bucket is at Cl-=0 , renainhg
ccnfortably belov, tie lo,"est operating CI nrini:rn:rn of C1=.I5. A11 good stuff.
ltrere is one big disadvantage caused by the camber, ho\rever. tlctice that the
pitch-ing nsrcnt coefficient is no longer zero, but is Or=-.05 for both airfoj,ls.
Increasing the camber, or rpving the loading aft, both result in an increase in
cln. Ttlus, a pitchijlg nsrEnt is ttre prj,ce lie pay for carbering tie airfoi], and it
requires us to invesligate tie effects of On on airpJ-ane perforrnanc.e so that ue c€n
design our airfoils intelligentl-y.
']
95
Ttjfi Draq. As shovrn on fj-gure 5, the center of lift (center of pressure ) for carnbered
airfoils does not alr^rays remain at the nragic guarter-chord point. It does so at the
stal]", but as the angle of attack decreases, the center of lift IIEves reandard. But
since ne must always locate tie aircraft C.G. at or near the stall c.P, , so
that tie nose of the airp.Iane will drop at the stall, the wing lift vector at cruise
f or carnbered airf oi.Is is af t of the r,,Eight vesEor, and the airplane has a nose-dc'rrn
pitching tendency. The fix? I,fe put a horizontal taj.l surface on the airplane to react
the pitching nE(IEnt, holding the tail dcr.n. But as tle l-eamed earlier, negative l-ift
in ttre tail surface produces induced drag' *hich adds to the wing d!ag. Retrernber
also that for every pound of negative lift in the tail, an additional pound of li-ft
must be generatd by the wing so tllat the swnnation of vertical- forces rernains zero.
For exarple, if the airplane r.ireigls 1000# and the tail do^,n-load is 40*, the wing
must generate 1040* total lift. Ttris additional 40+ of fift, above hhat rapuld be
required for our sl4flretrical airfoil, also creates induced drag. so vte see that t.rjm
drag, or t}le drag associated r,rith trjJrrLing out the pitchlng nEnent of a canbered
airfoil, has trrc cc([Dnents- the induced drag of the tail, and the additional drag
of the wing in producing an increnent of lift egual to tie tail doum-load. For our
tr,.E canbered airfoils, fi-gure 5 shcr*s that the trim drag at cruise is appro).ijnate1y
5-8 drag counts, which is about I0S of the sectsion drag coeffici-ent, and addl-tive
to it. conclusion? Unless e€ can shctur at .Ieast lot i.ncrease in rnaximum lift frdll
cambering the airfoil, it isn't r.Jorth it. Ifappi]-y, the increase in Clnax for our
airfoils is about 30t, considelably greater thEn the drag increase of 10t, so kle
have a balgain. But thr-s is not always tj€ case, especially with aft-loaded 3irfoils,
so t}te bigger lesson j-s that r*e should always consider the effect of oj fferences in
on on airpJ-ane perfornance when ccrnparing airfoils. vse also see that it behooves us,
when designing subsonic (1ight cA) airfoi-ls, to keep the On as lcnr as possi-ble, other
things belng egual. Using srmth, forward-loaded nean lines, and using only as rnuch
camber as necessary, does ttris. This is a rrErjor critj-cism of the l|A6A 6-series air-
foils. which are nid-loaded, and tie infanpus N&sA GAIAI (I^5-1) airfoils, which are
aft loaded I an j-nexcusable mistake. Don't use thern. For ccnparison, OrF-.05 is lovr,
CrF-,10 is h-igh, and the Gn of the f.ASA GAI., airfoils (npre than -.I5) is outrageous.
Ttrere are other th-ings that can be done to ninirtrize trim drag, such as using
a long tail length, at least three wj,ng chord lengths long. A high aspeqt. ratio
horizontal tail. surface also helps. Anot-Ller effestive technigue is to have a r,,eight
shift to the rear at cruise, follcwirg the C.P. shift to the rear. Retracting the
landing gear to the rear does this, as on tie Iancair rV and the Cessna 210.
Sunnarv. ltrere are thre€ crefficients of airfoil lErfolnEnce- CI, Cd, and Cm- equaUy
jnportant. the effect of airfoil- On on airplane performance mtst be quantified and
not neglected in desigrning airfoils. orrly forr,€rd-loaded nean lines, such as the NACA
(0.5) nean line of reference (A), should be used for csrnon subsonic C.A airfo.il de-
signs, to rnillinrize Gr. It is not necessarl' to reduce t].e airfoil On cc(pletely to
zero, rrrhich r,.,a.s done with the }.lAcA sdigit (230rc<) airfoils,at tlle unfortunate cost
of degrad"ing safety, but onl,y to keep the On as lcnr as praclicable, other things
being equal.
Disclaj-ner. Ttris paper is adnittedly a sfuplified explanation of airfoil lnrformance
and design. For exafiple. wind tunnel force neasursnents are now ccnnrcnly taken with
pressure rakes rather than old-fashioned spring scales- Al-so, for satisfactory
horizontal aircraft stability, it is not necessary tiat the taiL lift coefficient
be negaLive, nelely ttlat it be l"ess than the wing l-ift cefficient, provided that
the nuny olher factors affecting horizont-al stability contribute in a favorable
(stabilizing) direst:-on. Hou,ever, at slo\^' ftight, which is norrnally the lrrcst crit-
ical conditron for longitudinal stabj,lity, the center of pressure of any airfoil
approaches tne C/4 point, and tie c.onclusions presented herein regarding "trjm drag"
or drag associated with trinning out t].e wing Qn, are not invalid.
Revised 5/I7 /95
Ha.rry Riblett, 416 Rilelett Iane, wilrLington DE 19808 302/994-0479
3
5Y--c';..". A,Rio,L5 - tt*
96
-".^ Laa,.o< Ru,JG o{- o',
-lL =
10. BuL E*r ' -9r'€
fl< _
LrnrN^q Tvti
Lrlsjg6__&:af )
5 f.r., q sa4rE ( I
ilr.,h,'trL
Fl(t
Y
Y<_
)1 ,te'b Retca* v.oe
{/ t-
LA'NI'\.AA
'YP'
?th1-
97
:a
€
q
',,1
3l
rl
:il
.rl
rl
rt
it r43
*l ;: br ,.
F
ol r! .,t i
r. !l
>ll-
;l
3la
l
<U
il
il ! s. .r! r E
:tol
F 9
rq :l
ilEI
{l
;l
98
..Lo
c.fl {s c-
4,.
,.r )
.t-r .Sa
( . P,
x
I '+(
=
"-(+)
Acq 2-{rl I
Sro N
I
8<
'{ Jt
o
3o
rt
l-
r?5
r!
ll.'
rlt_l
1-
t! ,c
-'[".,= cF'i-;:f{.4'-i)l-
= ___- r'f-c.-c.F.,,.t1'
r/.rl-'r\ '.\c-r..,j
v
L
D
,oo
|
,.'l
*Dr +
'r
ltt.- |
ll ,il I rA'. Lerare' Jq
.q ,60 tO
( \\ ','''rN.'l
A'R' = 6
---i | searv = NtNl.5PA^r
c(
l_r -\o'
,il\ I -tA''-
Il- i
l- oool
I i-_;:.:+;\:rl\,-
.-.-{ 1...o-o'---
"n
/
i-- ------l-
-l
no r,'- o.,f,' i:-;^l{,.'j* I
|t.z 't-,ti-^l
II,,.}\^riIrl"lq'-
t.6 ).+ I.t z.o
,*
ADDSNDUM NLD4BER 2-
99
Deslgn ltotes for Tapered wings
!,lhen designing a taperd wing airplane, the designer is faced with certain choiees.
including:
1. !,lhat airfoil secbions should be used for the root and tip sections, especial-Iy
concernj-ng [Er cent thickness?
2. Hctur much t\,rist (washout ) should be put in the wing to protect the wingtips frcrn
prenEture stall, and to preserve ail-eron c.ontrol tlEoughout the stall?
3. !,lhat taper ratio and what aspest ratio sbou.l-d be used?
Hj-storically, designers have teen guided (at least partj-ally ) by a series of taP-
ered wing tests conductsed on turbulent secti-ons at NACA I-angley in the rn-id-1930's.
Figure I sho\ds a tabulat.ion of the strEcurEns used during these tests, and the tabulated
test results . D/pj-cal performance curves are a.Lso sholnm on figr:re I. l.Iote the sketch of
the initiat stalf dlstributions predicted by l.lACA for "restangular, tapered, and sr,€pt-
back wings". This sketch is especially botherscme, since it il.l.ustrates the NACA con-
cl-usion that "rectangular wings stall first at the wing roots and tapered !,.rings stall
'
first at tl-e wing tips"- hence, supposedly, rather drastic llEilsures are needed, such
as use of considerabl-e rrrashout, or use of sections with rore camber at the tips, and
then even nrcre uashout, to ccrnbat these "bad tip sta1l characteristics" of tapered
wings. The bothe.nscrre aspect is that the planform differences betrnteen the straight and
tapered wilgs are not so great as to account logically for the n'Erked differences in
the location of initial stall reported by I{ACA. Neverthel-ess, generations of aeronaut-
ical- students have been conditioned to accept these resuLts as gospel, and to desigin
accordingly. ltre ccnnonly acceptd washout recqrrrEndation is about 3 degrees, in spite
of the fact. that twist reduces t}le overalL efficiency of ttle wing.
!et's look a Iittle cJ.oser at the test s1=cimens used in the IiACA tests, hodever.
Incredulous.l-y, of t].e 22 specj:rens tested, a1I but tLlc tEve a t thickness taper frsn a
r@t thj,ckness of l5t, f68, 18t, or 20t to a tip tiickness of only 91, in addition
to the planform taper. The nexc to last specimen uses a 23013/43010 ccnbination. that il
nrrre canrber at the thinner ) tip- all t}.e rest have the same canlDer, root ard tip. onl-y
(
the l-ast specirren has the sanE per cent thickness at the root and tip, IZt, but it uses
an odd (eUiptical) planform. Evidently. l.lACA ( erroneously ) assurEd that the preferred
way to build a tapered wing is to taper both the pladorm and the per cent thickness,
and unfortunately the test results are contarninated by the effest of these tr.ra variables
ccrnlcined. casting considerabfe doubt on the conclusions. It is curj-ous (and unexplained'
tllat NACA never dj.d test the tHo variables separately, except for sarple #22, htrich
has an odd planform that further contaninates the data. In fact, if they had tested a
2415/2415 sampLe for exanple, with 2/I planform _taper ratio, they reoul.d have found
tllat the initial- statl, distribution is the sane]for a straight rectangular Hing, that
is, benign. Further, if they had tested a straj-ght pLanform wing with a per cent thick-
ness change frcm root to tip, they rrculd have seen the locaLion of the illilral stalf
shrft to the wj-ng tips. thus. c.ontrary to NACA'S conclusj-on, the spanwise location of
the rnitial stal1 j-s pri:narily a function ot the per cent thickness change frcrn root to
tip, and is not due to the planform taper. Prmf of thrs is shor*n on figures 2 and 3.
Itote that the thj-n (9t) tip sestions stall about 3 degrees ear.l-ier than the thicker
/raot) sections, for both the 24>o< and the 23olo( sections. Flrther, the mlst efficient
'secLion thickness in tenns of naximum L/D is LzZ, as shom by figure 7. Therefore,
there is absolutely no reason ever to use a tip thickness Iess than 12*, and NACA'S
design of the tapered sestions was a rnf,nunEntal mistake.
UnforEunately, follcrvring NACA'S psr exarple, a r.hoLe generation of tapered wing
airplanes was built with wing tips that r^,ere t@ thin. Fign:re 4, c€urtesy of I'lr. ceorge
Copland of Duncan, OK, is a partial listing of representative planes frcn the I930's,
1940's, and 1950's. Note tJee ill-advised use of 9t thick tip sections on nany of tlrese
ccnrrDn airplanes. It is also interesting to note that a fe!e, of tie nore sucessful air-
planes (P-38, P-5I, Dc-3, Eonanza ) did not fall for the trap of using 9t thict( tj-p sec-
tions, but used 12t tip sestions, contrar)' to I,IACATS reccnnendation. l.lote a.l-so that the
Pi,per PA-2+ cc.manche, desigrned i.n 1954, rrras a historically si-gnj,ficant airpfane, for
it "broke the ncld" by using only planform taFer with no [E! cent thickness taper, with
100
no washout in the wi-ngs. The resul-t was a very docile ai4ll-ane, Unfortunately, hor"ever,
too rnany nndern airpJ-anes. such as the Venture, Malibl, and high-perfornEnce E€ech
ncdels (see frgure 6) continue to be buj-lt with thln wing tips, unnecessarily causing
control prob.l-etns and ccrnprcrnising safety. 'Itle 23015/23009 cc.nbination is especially
j-11-advised, due to the additional problsn of sha-rp statl cha-racterj-stics of the 230)0(
airfoil-s. Figure 3 sholrs the mechanisn of this sharp, dangerous sLalJ-. The 230xx carnber
profil-e has a discontinuity (ki-nk) in at at 158 chord- Tl-ris causes the airfLsd. at
higher angles of attack, to selErate at that point, resulling in a separation bubbl-e on
the top sLrrface. foU-c'\^Ed by reattach(Ent before the no.nnal trailing edge separation.
The added thickness of this separation bubbfe on the top of the wing causes the wing to
act as if it were thicker and nrere ir-ighly canbered (high li-ft) at high angles of attack,
i-n spite of the re.l-atively lovr gecnEtIic camber of the section. At lc,v, angles of attack,
the bubble disappears, and the section leverts to its lop--drag profile. For many years
thj-s airfoj,r was touted as "lnving the best of both r,rorlds", but ttre disadvantage is a
dangerously sharp stall. As the angl-e of attack increases further, the Pattached sec-
tion gets shorter and shorter unti.I the separation bubbl-e suddenly expands to cover the
entire top surface. resufting in a sudden toss of lift. This is bad enough on singre
engine airplanes, but on twin engine airplanes with one engine out the airprane miy
slap-ro1l suddenly and unc€ntro.Llably into the dead engine, if the speed drops tcrl lorr-
(King Air crash € l,filmington 5/2/93, Fig 5 del) Recent1y. vortex gen;rators have been
used to rmprove this situation, as shc,wn on fign:re 6. the vc's are instaued at l-0t of
chord just before the discontinuity at 158 of chord. They function by creati-ng a thi-ck
'
turbulent ( energj-zed
) boundary layer i-nrnediately aft of ttre VG's, preventing ttie
ation bubb.Le frcrn formrng. Ttrj-s forces the flo,.r into an even rpre frl-gUy ca.nereO=epar-
snape
(nore lift) than with the bubble. wi,thout tlrc disadvantage of the bubbl; ( sharp stalfi.
Ttre Price is a slightly ]-orarer cruj.se speed due to the srnall- anrcunt of drag frcrn the
vc's' but overal'L this is a very good guick fix, uproving safety. of course the bette.r
solution, next ti.file around, is to use nDdern soft-sta}l airfoils not subiecL to the
separation bubbl-e probl€rn. Also rerernber that the vc's are not a parurcea for aI} air-
foils' but rather a sg=cific crutch for the poor features of the 230>o< airfoiLs.
one finaL ccflnrent is needed regarding l{AcA sarple *22, with the erliptical plan-
form. This slrcirnen has the Lrighest CLmax (I.8I) of any tested, and is aclordingiy often
cited as proof of tlle " inherentLy superiority" of the Lltiptical planform over t-l-re uni-
forrnly tapered planforms. I{ote, ho^rever, that this specijnen enjoyi tro distj-nct advan-
tages over the other specjrEns- it has a unique high-fift secLion (4412), and in addition
it has the same per cent ttrickness root and tip, l-2s, htlich is, as r,,,e have seen, the
optimun thi-ckness in terms of best L/D. In vierrr of this, we vDnder if the legenda-qr
rnystique of the ellipti,cal planform is justified.
Regarding planform taper ratio. it is r+et1 to renrenrber tllat a ve4/ short tip chord
reans a 1on' Reynords nlrnber at tlre tip, and ai-rfoil section performanci drops ofi at
J.c&r Re)mords nurnbers, ltlus it i-s best not to taper the wings tcro much. cccntcn taper
ratios go frcm about 3/2 to 2/I. Snal-Ier tip chords rEan $naller tip losses, but also
lq.€r ttn, as stated above, Of course a higher taper ratio (2/l) alsi tras a structural
advantage since it pernLits a greater spar depttr at the root for a given per cent root
thickness,' other things being equa1. Alternativery, a higher taper iatio may permit
use of a lo"rer per cent root thi.ckness (15t vs lgt) for letter i/n**.
Regarding aspect ratio, the hj.gher the better, Ijnlited only by structural- consid_
eraLions . AsIEct. ratios betr.reen 6 and l0 .rre ccnnErn for porered- aiforan"..
rn sunnary I for turbul-ent wings, the cips should not be Less than lzt thick. For
Laminar wings, notice that uEre is litt1e difference in L,zD beer*een ]2t and l5t thick
sections, thus r5t thick tip sections are preferred. rf thj,s is done, and if nrcdern
sections such as GA airfoi.l-s are used that have forward road-ing and lead:-ng edge dr@p,
the tip sta11 probrsn is eurrlinated, or at least, is greatly reduced. use of additional
canber at the tips is not needed, nor is vring twist. A flapped section starls at a
loler angle of attack than the sanE sesti-on unflapped, ana tfris in itsel,f provides pro-
tegtion against tip stall-s. For twin engine airplanes, to herp the engine-5ut situafion,
the above recqlnEndations are especially applicable, particularly with regard to the use
of soft staLl airfoils, so tllat assynretrical- lift c-onditions are unlikely to develop.
ttarry Riblett - 4/I8/I994
101
NA CA TRP€(t:D uJlA/6 Tesrj
oul o r..,.i.-
o6.1-0 o{ +o @ iF
o04t]-ot1-o'!oI
o d.t od_o$, !.
o orol 06-dq.0r
.. i..L,'ad... b, trF d c. t.t trd .tn wtptbct
o @'l-o e.l o orr ! n
odq-od' 0onl0ir
o6ri-o@ oor..o.t nr. rr:to. lntttrl .L.u dl.ldblrro tor Er.4u*, t^Ffd,.d
Do?rl-o@ oororo.o n Dr,b..l vwr.
0 ootd-o or! o od o
oc,l-odl - (l)S.t-rr. o,, tl 5!F1. rppdtmtlo. t{.!nod t.. Gnrlnt.a t}. s9.n-
od, o'r
o oo..i-o orr r o!. o r.
vlt. l.In Dttlr1iD11on.r Nr'r lx 96' r9'o.
(2)rEry. \rid, J. -llEnf! si*tu .r }teu-f,[!. rtro.
t.a tt y'rml
zo ..0 €.oui
re :e .E o,cl-
t.6 Sz !F
>.Ot.l
G'9 F
tt -aei iotzl
r.r!i.a.! l.o'eL
q
'"!-ii4
; G_
.r B.,5l
e
-c\.," i ,p.-
: .. .06 i @?:.
2 .o. rg-
e
o a u.-.,.-
) /Et6t Q.-o-o, Et/tt/t .t.6t9l
lsn t*-Feco. e^ ;F. I-. '.? g -.:F
I ,r., .Jt^aora t &.. Rrl ZO9QW|'
1rut .frtt d:Ltt.L. Lt uQf tt73',t' r .L
E-rl-
1c*ccpa tor t*r-a.t
o . a ..tQ.r.d
ldr9..etl
t6 m z.2a 5z
.ft.<t
to tut ctt'tt
'
t-
.'l Ft6uRE
102
0
i I
,t
:,1
.::
i--
l- d 'ul
:j
: aY a- ry\ e ;
st".. .
Yi1"''^,"2'
'-; z .:
€
!i 1
t
q
' t./ L',/
,/ -t 2. 'nf r-. l |. o\
,.s(r
1; {
.q
t-t
at .{
-l
9;
J U^
a -.
q |l.l<r
'l=l I
Qo "
,.\
>l ; I
t,v J
l^ ri
lll
{
t-
l^) oj
si
/
,o$
'v ./-"-/
/' ,+/ ./;
i tli {
,rl
,.- t'
:iul e
.F :G .) ii
]- zt' "'
' $r.,' o \ lo
F
.t./' \j{
"- ,\s
\l Fl{
Fr Ll-
qIJ-- l-
5
$
d tlg t r.{J
\
S
:.]
! ^o
-o
i1
{ \r
I !lit
ql
ti ai rrt
t.
re
;l -{l
rl
ll
rl
+ o rl
d -.1 I
F-
.t
N
+
103
Airfoil- Appl-ications
Airpl,ane Root IiP
l4ilitarv
Lockhe€d P-38 Lightninq 230J5 44L2
BeIl P-39 Aircobra 00I5 23009
Cureis P-40 l.lalrhavrk 2209
l{orth Arnerican P-51 Mustang 63-215 plus strake 66-2r2
Chance Vought F4U Corsair 23018 23009
Boeing B-17 FJ.yi-ng ForEress 00L8 0010
Consol idated B-24 Liberator Davis Davis
ctlrtis-Vgright c-46 23017 4410.5
Douglas c-47, DC-3 2?L5 44L2
Ccrrnerci-a1
culver Cadet 34L4 3408 plus sJ-ots
Ryan Navion 4415 6410
Beech tdel 35 Bonarza 23016.5 23AL2
B€€ch Skipper GAlr-l
Ces.sna 210 64A215(a=0.5 ) 64A412(3=a.!)
Cessna 310 23018 23009
Cessna 337 Skymaster 24L2 2409
Cessna 402 23018 2300 9
Cessna 500 Citati-on I 23014 23012
Piper PA-2$ ccnrmnche 54A215 SanE, no twist
Piper fA -?t ch€rokee 65-415
Piper /d -rg runairar"t GAI^I-I
l4ilsubishi MU-2 6/lA4t5 63A2L2
Gnnnan Goose .
23015 23009
Hcnpbuil-ts
Fal-cp F-8 64-272 54-210
Questair Venture frcrn t4alibu
( ) 23017 23010
I-ake Buccaneer 4415 4409
Bushby I{idget Mustang 64}',l.2 64A210
wittrrEn Tailwind 4309 top.0005 bottcrn
classair C,A!{-,' l,lcdif ied
Lancair, wheeler D(press, Prescott Pusher NLF( I ) -0215F l,4odif ied
Pul-sar lrs ( 1)-0313
Jones lihite Lightning 66-21s 66-215
lhorp T-I8 53-412 lbdified
trtGUAe 4-
104
./
ts TAUKC (a
{,
105
3oo
a
tr
L ID
Rnlo
Loo .(, uu*@ (4- 3rror.
{or,,l O ig= 6r/oc
- Zq-X\ AAJtr zjdkx DArA riAcn ag 631 * 6p,
La €..eerc, ?es r& gSl
,l
y'ar Doeelrlar=F,
- L*-zxx DATA f:p.am Ae6orr + ':,72^I*::rr,
Ptr' cePr AlRFor TH,a*U-t, ,to -'r-lrc'l'11-
lfF -,,t,rta.r
J5 ; te,lP-P 1 l8
106
:l i:
!
'e4rM r. !rvd rrt,t)rr.f t !rr. -rortn'ot
@
!
:l t I
'.1 L
tl t,
il
t^l
!l
'l
2l
tl
I
{ .l
.J
:l
Y
i
l-
a
-t
)
t!
3 :,
rt ,i
-o
d
! |-
il
1
F
1
:l c
o
., il F
a
:l g
c
F t I
c !
! .?
g -! I
i !
{ 3l J
'rb t-
t
: ;
)
!n
crI ! o
F i'"
.r- 1
tr GI
,I.'l
5l
lr
tl ll
t; ; /! $l
l
T
]-: I
: o
)?
(a
I i:
*.1
r-l 2
\ ;l
/l-.. o
IIO. 3 nar"v niurett 1 07
ADDn{DUI'1 /il6 Rib'en Lane
(Irl/JtlI995)Tandnq trre sharp stal] of Lhe NACA 23012 A.irfoil Wilmington, DE 19808
30?934-0479
Srnnnarv. This paper discusses t'.r) methods tllat carl be used to soften the notoriousl-y
l-narpstaff of the neCA 23012 airfoil, that is, use of vortex generators, or alter-
nativeJ-y an upper surface build-up (reprofiling). l€ also discuss the nechani-sm of
the stail on ttris airfoil, and explain wlly additional leading edge droop is ineffect-
tive for i:rproving the stall on this airfoi-I.
Discussion. In the mid-I930 ' s, follorn'ing tie r.prk on the 4digit airfoif series .
NACA aducted a series of tests on airfoils related to the 4-di-git airfoils
( salre
thickness distribution), but with the rnaxjmum camber placed unusually far forward.
The ob ject of the tests r,,ras to see if the aj-rfoil Gn cDuld be reduced to zero, Iartdch
is not a necessity for conventional, airplares any'\r'Jay, as long as the GIl is kept
reasonably lcrhr by nEderately fo4rard loading. Anyhc'vr, these " zero on" airfoils !,rere
designed ana esiea, and the nrcst widely used of these 5digit airfoils is the ubiqui-
tous and infanrcus I'IACA 23012 airfoiL. It is l2t thick, with Peak camber of 1.8tc
Iocated at the l-5tc position, rather ttEn tl]e usual 40tC Fosieion us€d on nDst of the
pozufax 4-digit airfoi-ls Q4I2, 44L2,4415, etc-). the nean line aft of .I5c is a
straight 1ine, thus the only camber in tl:-is airfoil is in the first 15t of the chord
length. Accordingly, rrie can describe the airfoil as bejlg the DiACA 0012 sYrlretrical
secaion wj,th the first 15t bent do,mward5 (lc:ding edge droop ) approxirrrately 1-8t.
Actually, the effective drmp is only about l.4t due to the faulty "slope and radius"
nrethod of leadil]g edge design used on all of tjte I'IACA airfoi.l's , ho/,,ever the fact
renains that the camber profi,le of this airfoiL consists of leading edge drmp only,
with no conventional camber. the result of tJris is an airfoil- with near-zero Pitch-
ing nErrcnt coefficient (Gn). Fr:rther, sirtce there is no negative canber jn the air-
foil, the nEximun lift c,'cefficient Eemains high and the induced drag is loi^r, cdrpared
to the best knorvn previous zermn airfoil-, l'4ax ltur)l('s 1924 |+6. The [&-6 achieved
zero cln by reflexing ( negative carrber ) the nean line frcrn .60c to the trailing edge,
effectj,vely kj,lling the nosedcnn pitching nE(Ient by applying a dcarnload on the trail-
ing edge of the airfoil ( figure 9). this, hokever increases total induc€d drag, and
reduces Clnax. the carnber profile of the !F6 in the first 60tC is conventional ' ho\^r-
ever, so the M-6 des have a nice. soft stall-
wind tunnel- test results of the "ns^," zero-on ai-rfoit, 23012' are sr.nrnarized
in t{AcA IR *537 of l{Ey 7 , 1935 (see figure I). Based on the faets ttlat 2301-2 shc,LEd
a nEderately high Clmdx, very lol Gn, and Cd no greater than the 0012 syrnetrical
section. NACA pronounced this airfoil to be "markedly superior to l€ll-knovm and
cc6rpnLy uSed sectj-ons',, and reccnnended its wide usage in glorn'ing telrn5. HouJever,
the airfoil has a terrible sh;rFstall characEeristic. wtlich MCA TR#537 failed to
d.issuss, and that i-s its domfall. thrs was noted briefly in Table II Airfoil Data
of IR #537, So IIACA knevr of the existence of the sha-rp staU, ard chose to ignore it.
SlErp sta11 airfoj,ls are bad enough on single engine airplanes, but on prop driven
twini with one engine out these aj-rfoils are especially letha1, causing accidents
such as described in figure 2. In fasE, sharp stall airfoils are the rnaior reason
that cA light twin fatality rates ironically exceed the fatality rates of GA ]ight
singles. [,le will not achieve true twin-engine reliability in prop-dri-ven twins
untit h,e get rid of these sharp-sta1I airfoils. ProP.driven twins are ljjrlited by
Iateral cont.rol authority near vtIE due to the effect of a "blovtn surface" aft of the
operating engine - with shatp-stal-l airfoils, sudden unccnmanded and uncontrollabl-e
upsets occur at relatively high vlo'lc. with soft-stal,I airfoils, vtlc is much ]o$,er,
and in addition the roll tendency is control,lable. Knq,'/ing wllat r.,e knod today ' 'e
realize tlEt NACA, as soon as they learned of the bad stalL characteristic of the
sdlgit airfoils, shoul-d have terafnated the project. Accepting the sharp stall
rnerely to achieve zero Gn was a Poor trade-off.
There are t\^D tl/pes of airfoils that have sharp statlsr tllose with too u,tt.l-e
camber in t-ire Ieading edge, and those with too much cafiber in the leading edge.
Exarp.l-es of tJre fj-rst group incluae nDSt synrnetrical sections, and l-olv-carnbered
air-
ioi:"! such as ttre- l-ater 11gcA 64-212, These airfoils experience ccnPlete and
suoa* rlo', separation frcm the very leading edge at the stall-, and they can usually
G-i-pt"""O by adding a sITEIl anpun! of leading edge d"rmP ' Arrf oils of the second
108
group. ircluding ?3OI2, experj-ence ccnpl-ete flo* sepa.ration on t}re toP su-rface at the
stall, frcm a poj.nt near the end of the leading edge droop, that is, at about .l2C
in the case of the 23012. the resul-t is the sane. horever- a shaq) loss of 1ift,
usually acccnpanj-ed by loss of lateral control, and a hysteresis loop reguiring a
substantial decrease of angle of attack ( wrth considerable altitude Ioss ) before
attached flc'vr can be re-established.
Arrfoils of tfre second group cannot be irnproved by adai ng leading edge drrcp,
since they al-ready have too much droop. For exanple ' adding npre droop to the 230]2
airfoil results in t}Ie 33012 or 43012 ai.rfoils, and these have stall- characteristics
as bad as or vrorse t-lnn the 23012, frcm wird tunnel data ( check it out)- I'itut is re-
quired is to ease the cransition frcrn the leading edge droop to the rest of the rlean
Iine, reducing the discontinuity in the nean line at tbat point- one effecLive and
proven rrEthod is to install a cqnplete span-wise array of vortex generators on the
top surface at alout .10c, (Fiq. 3, delered). Ttlese function by fi11in9 in the "low
sFot" on the wing dch/nstrean of the vcrs with a thickened boundary layer of energized
turbulent ai-r. discouraging flol separation. Ttris fix is cheap and effective, and
should be reguired on all twin-engine prop cqnnuters using 5-digit airfoil-s, which
is the nrajority of the fle€t. AIso. don't forget that that single enging airplanes
with 5-digit airfoils can benefit frcrn tj:-is as rrell.
Another possi-ble way of acccrplishing ttrc sane fix is to reprofile the wings,
filling rn tie lor,r sFot aft of the leading edge droop on the top surface with soli-d
materj.al such as foam and glass. This is quite cqrrrDn on experinental (hcnebuilt )
aj.rplanes, wi-th both nptal and ccrPosite wj-ngs. Fign:res 4, 5, and 6 describe this
nethod. l,lotice that the secej.on drag astuafly decreases, in spite of ttre increased
sestj-on thickness, due to the prcnption of Iaminar flovr. Furttrer, the zero-lift Gn
renains about the saITE, so top speed is unaffected. Ttris nerr profi-le could also
be used to nlcdj-fy existing [email protected] on production airplanes. Of c-ourse, t}le better
soluti"on aerodlmarnically is to discard the ?3012 airfoil crcrTplete.l-y, and use a rndern
lol.F-gn, soft-stall airfoil, such as a "cA" airfoj-l.
As stated above. the I{ACA s-digit sestions cannot be inproved by adding leading
edge droop, tius the tlc State/l{AsA fj:< on the "Venture" airplane (figure 8) is a
poor solution lo the prob.Len. I suspect that any jjrprovenent of the stall in this case
is nereJ-y the result of the consj-derable aerodlmaruc twist that was introdticed into
the wing by this fix, delaying the tip stal-l. Horever, tll]is causes a considerable
loss of efficiency at high spce d, and aLso raises the ]anding speed of the airplane,
so it is not a good solution to the problsn.
4301X Airfoil-s. The NACA 4301X airfoils are the sarrE as the 2301x airfoi.Is, but with
twice as rm:ch leading edge droop ( figure 9). The stall is as bad as, or tanrse thar.
the 230IX airfoils. Fortunately, the 4301x airfoils are rarely used. one exception is
Fred l^Jeike's " Etcoupe" ( NACA 43013), holrever on this arrplane the elevator travel, is
ljrnited so that the stall angle of attack can never be reached, Still, the 43013 was
a strange choice for this airplane.
Anotier airplane that uses the 4301X airfoiLs is the Ftench ATR-72 twin turbo"-
prop c:cnmuter. and tlEt story does not tEve such a happy ending. In october 1994 a
crash occurred at Roselar,en, Indiana kitling a1I 68 people on board, a direct result
of the NACA 4301X airfoj.Is used on that airplane ( figure I0). As confirned by sub-
sequent in-flight spray tests a 3,/4 inch high j-ce ridge forms on the top surface at
.09c, forfidng a very effective' spoiler. vffren ccrnbined with the already terrible stall
charactseristic of the airfoil, this caused an unccnnanded, uncontrollable roll-over
and irrec'cverabl.e dive. Ttre ice ridge forms at the very spot on the airfoil- \rtrere
a pronouncd suction peak exists ( figure 7), since here the tsrperature depression
frcm adr-abatic expansion and cooling is the greatest. The ice ridge is aggravated by
tlre strange ptacglrent of the vortsex generators on the ATR 72 wing. The "fi;<", longer
de-icer bcots, is nrerely a bandage, for the basic probJ-ern rguins. Ivlodern c-onstant-
velocity airfoils do not have this suction peak, nor the problan.
Conculsions. Ttrese interirn fixes should be used until the t'IAcA 5-diqit ai-rfoil-s cbn
( }IACA TR
537 ) r.b. 109
ITRTOILS ILII'L\G ?EX I.A..XITU){ C.{.I(BEA ST-OSI]ALI,Y FI..R
FORW,I.ND <.),
despo calcu,latioos. Tbe simple meenJi_oc rirloils terlstics over well-knorrn and commonl,r used s.irfoils o,
gste erceptioDolly lor+. pitcbiag DoDeDts, soElesb!,t,
lhrs clo.ss. Ir hos a hish ma_rimum hft and a lor
l9*". th- the theorelical vaiues bss€d on the mear
1",
liae. moheDt- Furrbermore, the niainum drng is
BotI the rueasured aod theorelicsj curves for tle procticrll_v o,s los as thst of tbe correspondiag
simple mean-line air{oi.ls are given in fgrLre t6- sji,-
Betrjcal oirJoil, the r*. J. C. -{. 0012.
_-
TJre anolysis of ihese chorts o'Dd the" dstr of toble I!gre.9e^neraU5, other secrions of tbis gtoup- sucb
IJ shor that tbe reflesed !-irfoils, altbouqh co[rDlr.irs ,
the N. .'1. C- A. 2lOt2 aad 29012 baringL e.cn loou,
as
fnror.rbl-r s.itb otber ref ered airfoiis, uJ;;J;;; pitching moment thllc tbe g30l?, sbo'uld supply thc
the simple meaali-oc uitt"ils_ f*tf,.r.o"r1il" need, of 'nan5
foils corerirg r range of carn ber locations lorro,rd ^r'- -applicarions ,"q"fiog a "figl,iJ-"'cam_
bered section of modernte thicl:ness f,uring"a
ot Dormal positions possess improved characreristics- ,er_, lou
pitchiDg EoDeDt-
-{ con:parison of tbe N. A. C. A. 24 012 ,ittr the
\.
--
-.1'. C. A. 2912 i-Ddicates the differea""s'tba; y;
attributed to tb€ dif€r€nce betweeD th€
forms. These airlofu harbg rhe same -.-_tio" LexcLsr ME}ronl,'l AERoliAvrrcal, Lesonaron):,
""-tJ;
tloD but difl'erent Des'!.li-ue foras possess aooroxi-
N.r.:rotr-e.r- ApvtsorI Conyrrrrr ron _A.unoraurrcs,
ntrtel_v tbe same lift and drag characteristics.' Ler;cr,er firr,n, 1-e., trlay 7, f $i.
The
angle of zero lift and the pitchirg moment, lo"r"""r,
are quite di{fereut. .Especio.llv notevrortby is AEFENENCES
the ven
r0uch loEer pitchiag moment produced b-. tbe airfoils 1. JTobs, ETlDel li., Itard, f,eooerb E., rDd pi.Etrenon,
reported herein. Robert trt,: Tbr Cbr,r!,.!.ristics ot ?g R.Irted
-Urfoii
ScctioD! froE T€srs i-b rbc !'t!iEble",Deasit-" Wiaa tumd.
T. R. \o. 460, li. -{. C- -{., 1933.
2. J^cob6, Easroao \., lDd _{bbott, Ira, E.: Tbe .\. C. A.
Vtiable'Deo:ity lfild Truael. T. R. r-o. 416, -{. C. ;.,
-\-. -{.
1932.
3. Jrcobs, LstEaD li,, rod Cl.af, \fillirm C.: Ch.rict.risticr
of tbe N. A. C. A- 230f2 Airfoil from T.!ts i.D tb. Iult-
*dl I'erirble-Dcasitr Tuoaele. !. R. .No. SJO,
N. a. TO. c- A-, t935.
TABIJ II._AIRFOII, DAT-A.
I Clrg-6at:nn r De
r---r tl
irr
lbdrE a!.t t.cno dl|tllr&BE
I "* -o,i$*-.ffijgr6
t!'r Ert !'
I
_
p.rr wur *r.
?
g ,t
r.
:';i,J3"f,:' ;ir..
i;.;. ol!d.<
;1311;-rrP6 "gP,n,,r" !
=- : _
ao
i:i:s:itlli:: ;I iii
\ :ii #;
c.a..1''2... ig jlfi ;il :lj
iil iil
ii ; ifr ,g ,m.* ii i * i l* ,H ill ift ii ii,_l :iB
:,H :Sl j.co, I $ 1 .@iJ
3r ,@i5
r0 ir
ro _-,1
6.26 12 z.
l.A l2 2i
itif ;; ilg;.i$f;''"t{r .Dord 1.5o.d &5 r i,a. ro,D,o8 !!. cr,ltD,u.. or r!. E-!riD..
t t-or ot i-r. rrccr o l-afi,m
! I rrb or hlr{urr. rsr N loJ-3 ro
r,rr
!!. tt.rcba.
EtAt)Qc I
c)
o)
c'r
(D
o.
E
O
5l ;; :i a;iz: l+: li:-,= - ii
A z;=ii; : l;li=; i;
=: l=t= E l;.r,-
110
o
,: E' I -.g:i;lE --! i!
<.)
o)
ot
ot
c
E
ic +f iietl'E:il1i ;; ! i
-
Ii iE
a.)_
(6
=
(!
c
f
:ecri;;fE;Flii k;;fii*
*(! E
E
co ; ul
S ig!i{+;:;,iri*l;i;;c;;
I/';vlg
o
zo)
E A,'
F tr::Ec€bE'Eb4:i:g;€t 2
UJ
.
4-zE+ia:cc i E *'-:
*: o
t-
\l
lb
J
fl il;e ;?iE=fi;*cli!';ii (U
an
t Eii;:ii;iiiff sg€Ei?cii
"
J
0
\J.
v
E iititi =:ei5E'ertii*;ii
:
td .E
o
r.o
I <
.AEi+i:;t ;{;f: F*€;;il;
o
q-
l,lJ
= J
c(
(-
TE 0
{
? () d
0
U
t E:iiiiff iiiiitiEiiii*
o
d.
(.r.t
J
t-
0
q.
.H \
J
=J
:.Ji
..,::
:1,.
s' iiel*i;gi?i3Fi ed
i ;":dg
$I ie's iiij;.EEIatqFe fu,.;
€
'JJ
':. tl EeE:$iEi?iesi€ T'8,"t
.EH":€ tI
),
E: gE:*ifiq!*.!; Ti ::.9
a,ifl -g€er I
ll -r,sil
I
f
:- ftiSEs=fE'is,HEE Ctt"rt R.tr 7-
111
UPPFR SURFACE REPRO TIUFl ?'3oxK Ar(F-orLs A3,-3hal9t
^JAcA
llar 95 NACA Reprof i I e ?f,(:) 1 O-UF if,{l1 :-uF if,(J 1 5-UP ?:(i l7-LtF :f,(118-LJF:.
O(:r l : over (-lr:) 1? tap ord l: LrF €rd lop ord lop ord top ord
4 . 2?..) i. I h'/ 6- 9?5 7 . 628 7- 9BO
1 {:) . {_lr:) 4. 6A: L97t-) s.?9f,
15.()q' 2, 5(:){) a. 954 9.181 I (:). ijTi 1r:r.518
?{-i. i)i:t I . 9l(:) /. /tt A. boE 1 I . (-)59 !I.-J-:'/
?5. (-) (:J
5, 94? ;r. ?4{ ) 8,I91 1(1.668 11.658
trr_). i_)(:, .5 _ i:, i,) 1 ;i . 4?{:r 8.4?1 9.421
_:Cff)
1r:) . ?? I
f,.: . (1(:) 3. 949 .l , 45i:r a. 4Lt7 a 11.478
4(:). {:r ij trl 1aa ( 1{1. 644 11-611
45,i](l 5. 581 :. ?60 7.911 8.841 ll,.li.6 I t.6i.i
51:). 4:r{:) 7.46t s. f,44 9. 668 I i:) _ 55(_r 1(,), 99 I
55. (_i r:) 6 ,477 7 -7rtj B. -i41 c.767 1rl. I E(,)
6i:1. {-r r-,
5.91f, 8- O54 4.814 9.195
6,C - (:) r:)
3. t._.-. 1 . g9(:l 6, rl?f, B. {19{:)
7(l) . i:rl) f,. 564 1 . 4?1,1 4.47i 6. {,r{)i-r 6.91,5
75-t)ti ;:. l cri:) {-r. qgr i 4.11{, 4.9u{} 5 . 69r:)
Fth, +
s ---=-
o,: L.ao,^,/E Xd.,t.
+
-
NAC4 13otz-'ArRFott--
F= 3 < toL.- .--SEaarN
S xacP
\ 7z sre"
Z-50tz -U P 4E lAoF,Le __
(onFu7t4 AtfrLY'tS
d.
.A - - \\
.\
I
11 .R,a.e'rr-i tlnll(
' -l-\
o
//c
7.3o, L- u p
tl
4
Vo l.e
*lc
,t
114
_!r i: ;iij=;^r-:.i: "-1{=} :i i:: lj:= !:.
= ::.:! iiii;=;: ,.
\'
x :!::;;;:,:g;F!:a;: r!i:+:;::;:!i =
a
€::;;;;1=!1i;iit; isi;;;i;j:;;! !r;iE:i;i;iiE, :
t-
j jEif iFiiF:;i=i:iiiriijiiff
q. ;;iiiliit;i:r: t; i:
)-
{(.\ ii iliii!lff i iiif ;i! ;;;i ii ; ; rai:;,
; * EF; i r : i€ ii
;;
clq
c)
I
! I cn
g o
-tt
a-
=
N
3
{
l--
--t
O
N
t 16 ii('F E-\
115
Ttris is qarbage
NORTH CAROLINA
._:r
-:
STATE UNIVERSITY
@,r:a :
. STALL STFIPS
Stall strips ar€ widsly us€d to force inboard
separation b€tors the stall celts fully develop.
t'.' ta' tfri' However, the r€sutting aircraft pitch_do*n firiis
th€ useable litt ot the wing.
Stall Pattem w h hboaftf Sla[ Strips
. TEST FESULTS
Wind tunnel data indicat€s that the adclition ot lh€se
wing enhancements provides more useable litt lhan
the industry-standard stall strips while draslically
softening tho slall break Subsequent flight tosts Modiflad V€nturo Stall Pattem
have further demonstrated th€ departure r€sistanco
ol ths conti
't.5
t.5
1.0 1.0
0.5 05
0.0
0.0
-0.5
* 8:sclinr .{1.5
oloops E Slotr
-+
-1.0
-t-0
.0 0to2030ao
Fl.ps - 15 d.gr. Fla9s - o d.Ct ALPHA . '.tt,
Wnd Tunnd Rquhr
Ct Ltto=4 - 1.
117
-,t'.
'.1
s
I
lt
F.
5l
xl :
n /
s
(l
ol^- +
av
AJ
I
:l
(f-lt^ ql
/
t-
Hl * /
t
/
\ .1l
i
.. --...-- >l .. .
Gl -'
?
0
/
'l : / /
6
r&
: ,l €
'' -t !
(l<l r! ---+---* \0
t
r
----
" /
J ,
/
4
Cl
rr
.rt-
\it /-A
1---\J--;-
Hl ,,
t.^,
A
Y
II 7 / l;*' v
\I i / II I.l
-s
I
.9
r\
+
r .c1i
r't.i f
a
| / / :i
It /
$
2
"l ,tr
--l . ?l
l/ I\ :j
,l\'s
t
.9 J -' U
t{ 6ar
llJ 0e
Y o_ Frq
-+-
q utt
ru u
v
U
ef
tJ'J
tl
-S-/o
118
ccnpleted airplanes, ho,tever, seve-ral, nEticds of re'profiling the wing shatrE have
been used witn good resu.lts. Seti Arderson has nodified hrs rel-l--knom BD-5 with an
ulper forl ard surface re-profile of his crldn desigrn, extsending back to approxirrately
40t of chord on the top surface, aJld ti-is h3s been fl-yjjlg for several- years with
excellent results. 'Ihis npdrfication is mini-rral and relatively s jrrpfe, and, Iike
otfler upper surface build-up schenes, stiffens the upper skin to prevent wrink-Iing.
Unfortunately, hordever, this partial re-profiling essential-Iy reverts tt.e wing pro-
file to a turbulent seceion, resufting in a slight drag increase. this is only a
nr-inor orsadvantage, and is of lj-tlle conceln, considering the nnrked inprove<rent in
flight safety tllat results,
Anothe.r schene for utr4=r surface re-profiJ-ing is shown on the attached sketch,
dalLed II/20/92- In this schenE the cqpl"ete up[Er surface is built up frcrn the lead-
ing edge to the flap,/ai.l,eron yELI lip, at approxjratefy 80t of chord. The flaps and
ail.erons regui-re no nadrfication and are re-used as j,s. A-lso, the exrsting tip sest-
ion, NACA 64-218, hrldle inprovable, has acceptable stal} perforrnance due to its
I88 thickness. 1trus, a re-profi.l,e tenplate j-s furnished for tJre wing root only, and
the re-profil-e br:ild-up feathers out to nothing at the wing Lip, arld at the }eading
edge ard at the flaplaileron r+e1l lip. I'4a)dnu.[n hrild-up, at the root, is I inch
additiona-l ttrickness.
The "GA" re-profile irproves tne alJ--i-nportant leading edge prof iJ-e, adds a
Slnall amcunt of thrckness and camber, and mcves the loading fonsard slightly so
that tl€ pitching nEnEnt coefficien! is not affecced. ltre result is a softer stall,
higher nraxi.mm lif t cef f icient with ajld h'j-thout f laps , and a wider larninar bucket,
In addj,tion, the new profi-le has been designed to nailltain c-onstant velocity flcw
to approxlrately nid-chord, preserving larn-irnr flqr ( ard lcrvr drag). CG l-ijltits for
the ai.rpl-ane are not affected.
Flight experienc-e with this nrcdification has been excelLent. A typical exanple
is the attached report frcrn Bobble Parcr of Odessa, Texas.
Another option for BD-5 projects under eonstrucb.j-on j-s sirnply to assernble the
wings as designed, and then re-profile the top surface befor€ ffigbt-
In the interest of rrProving flight safety, I offer a ful-l-size tsrplate
drawing tori lq1o to crf,ver the cost of printing and naiting.
8/16/1995
Harry Rr.bleft, EAA *29576
4I5 RibLett Iane
wi.Lrli.ngton DE 19808 302/994-O479
Note. The attached standard table of ordinates can be used to generate the profile
for-Tle BD-5 root terq)late ( 31 inch chord) , or for any airplane using the IrAcA 64-212
airfoil. I
\;
:
ac
:i /
J a =
,l<l .J
,
."
Ls
JI
,Jl
'r rl
$l z
\.!J ,
sl
; JI
ut
6 ol
0 \jdl
@ rrl
t
(o
r(l
)
tr
g o
q
a
I v
c
rJ llt
+ 7
s J \
o ?
c L1
q
\
IIJ
A
o
z o
r_lJ
J
0 l
1I
o
-{t
II o f\
\, OJ
c
o t-
c( v, Q- -t
q- !J
lJ l- J I
'E
o
0l q. -j
c ro
]J
(E
\5 U
\\
a \a
12'l
\-v ri
v o
'-z
rJ
;
iF :- I
I]
'I
I o
i' n ._ L
.-.-. i'. +
I
I I
s & F.l | /,- 'l l
ic
U
{
t
$n i i rl
''=1--'*---,* '-- .r-*-T ^0
IIt
.q.
ql<
d
"r
l, '',
'i* if ui-
'".,
.; ->
) L-
l',1 iI
.{)
l.l(
0
c( .; UJ b
u-
o- I .A
\U ho c-
t1 g\ \_y ",
I lfr
I
I
qr
a o :s(
o I
_t J '!-
!q
.i \5
j.1
o
\,a -1v'0
J
->\>o <.)
tL
I :,
't
( d f- l,i -,: a
IU
q.
+:ifa-':
.:, !l a ii: €
i: N trGOF.:,
\D
2
122
R. B. (Eobby ) Farner
P -lJ- rcX LZLtb ( Ret!"ped )
Odessa Texas 79768 915/563-2732
8/LL/s{
Harry Riblett
416 Rr-blett I-ane
wikilington DE 19808
Dear }tarry:
Like I told you, I \"pu.l-d let you kno\^' ho^'the (top surface ) scahr-on wing air-
foil re{iork calrE out (on my BD-58).
It lceEred nv stall sFed f rcm 80 MpH to 67MPH. The stal-I .is nq^, sc(ethj-ng .l-ike
an ErcoulE, kind of "mushing" like. lbp end, the s'aflE- no loss there. The warm day
take-off d"istance was 2500 ft, and is no,r about 1600 ft. I stiU hold it dcldn until
I get about 85 MPH, ard then lift off. HcrA,ever, I novJ notice that the acceleration
is guicker frcrn 60 to 85 MPH tban it used to be. probably because the wing is no$/
generating nrcre lift on tlle take-off ro11 than before, thus the take--off ro11 is
shoruer. TjnE to rotatj-on is nour only 16 seconds e 1600 ft, with OAT € 86 degrees,
density altitude 6200 ft. Our airport is at 3000 ft.ASL.
For the rer,rork, I left t}le wings on the ship, and cleaaed the top surface of the
wing of aII paint, to the bare ah:nr-inun, and scrubbed this with APX. I used 2 l-b.
dersi-ty fcElrjl, I/2 in thrck, cqvered with tr.ro layels of glass cloth. I put on the
first layer lengthwise, and the seoond layer 90 degrees to the first, and finished
off with tr.D coats of polyester, I then reinforced the leading edge with t\nlc layers
of fine hre-ave 3-inch glass tape, staggered, wtdch extended back about 2 inches
frcrn the leading edge on the loler surface. Itle tota.L cost of all Irlateria,]-s t|as
$105.00 frdn wicks Aircraft, and it t@k re 44 hours to put it al,L together. The
airplane enpty Height was 467 lbs, ard is ncr^' 471 lbs, so the rebork added only
4 lbs to the enpty ,reight.
t"ly ship is a BD-SBDH, s/N 19, N-6782F, '.rtrich I built myself. It has the standald
( non-stretched ) body fength. the wing spa.n is 2L\ ft. It has a non-turbo Honda
engine wi-t]. a 46x66 H€y prop. It has Matco wheels and disc brakes.
P.S. I also riirote to Rich Perki-ns and gave him aU the i.rfo and pish.ures for ttle
ED-5 neb.€rk n€rJsletter.
ADDE{DI}I II\J},IBER 5
123
A critique of the I\ASA N'I;F(I)-0115 eirfoi]
The Ju.l-y- August 1995 issue of the "Journa.L of Aircraft" (AIAA) contains an
article by SeJ-ig, I{aughfiEr. and Sc(rers, that presents a unew NASA" arrfoil, NLF(f)-
0II5, a "Natural l-a.rr-inar Flovr AirfoiL for General Aviation Applications". As it
turns out, the airfoil- was designed by Herr Doktor Richard Eppler, and I have corres-
ponded with hjil about it. Ttre airfoil and its nrean l-ine ( canber profile) are shotnn
on figure 1. After plotting these, I saw notiing frightening about the thickness dis-
tribution, although the point of rnaximun utickness I-s relatively far aft at '42C,
but I was concerned about ttrree things in the nean li-ne:
.1, ltle nose of the airfoil contains excessive leading edge droop, scnewhat l-ike
the NACA 230L5 airfoil, vrhich has a sharp stal] due to excessive J-eading edge droop.
Therefore, I suspected that NLF(1)-0115 nEy have a siJrdlar sharp sta1I.
2. A dj-p appears in the canrber profile at .60C, sinr-ilar to the NLF(I)-0215F air-
forl. al,so due to Eppler, As hE have seen previously. this area of negative lift ex-
tends the lan:inar n:n slightly hence the npniker "natural " Iaminar flc,w ), hoerever
(
the negative lift produced by the camber dip resufts in a loss of efficiency, a point
that I suspect Dr. Eppl"er has not c.onsidered.
3, Pronounced aft camber appears in the airfoil fran .65C rearward, and I suspected
a rel"atively high pltching lrraftEnt coefficient (Gn) as a resu],t. Ttris also',Duld prob-
abfy cause high aileron hinge lrErents (stiff ailerons), a.Ithough thj-s rould probably
be hard to prove without flight tests.
A canputer perforrnance analysis confirned ny fea.rs about this airfoil, espec-
j"ally the sharp staU at 13 degrees ang]-e of attack, accordingly, in an attenpt to
track dob/n the cause of the por performance, I designed a ner,r, airfoil- for ccnpar:ison,
c,A 42E-215, figure 2. Ttris new airfoil looks like cA 40A21,5 on figure 1, except the
point of naxj.rn:m thickness is at .42C rather th,an .40C. the new airfoil uses the un-
offensive Eppler thickness distribution frcrn M,F(1)-0115, and cqnbines it with the
proven GA-2 rean l-ine frcrn "GA Airfoils". This should give us a direct cdnparison
of the efficacy of the Eppler mean line versus the GA-2 nean 1ine.
Figure 3 shcrys that the perfornunce of GA 42E-2I5 is indeed superior to che
IErforrEnce of NASA Nf,F(I)-01f5, including a soft stall rather than a sharp sta.I1, a
wider laminar bucket, and a 30-401 reduction of pitching nr:nr=nt coefficient. The pro-
file drag as expested, is the sane, since both airfoils use tie sanre thj,ckness dis-
tribution. ' Ttre obvj-ous concLusion is thac the rean Iine of D.IASA NL!'( 1) -0],15 is poorly
designed.
If an even wj-der laminar bucket is desired for oFerational flexibility, GA 4OA2I5
rnay be used, but the tlade-off is a slightly higher profile drag coefficient. See. fj-g-
ure 3.
In spite of the ind-icated superiority of GA 42E-2I5, Dr. Eppler does not tike
1t, because it was not designed by the rEdern "inverse" rnethod for designing air-
foifs.
You be the judge.
Harry Riblett
416 Riblett Iane
wilnlington DE 19808
2/L4/r996
124 i
-:!--
'l,
t :l
ol
,t
v =l
i1
rl
I
>;
r:,
o. I
,-.1
...t
.o
..1
fo
O
I
tL
!l
z
o-'' f
il/
o
a
df
a.
o '4;A
? R t\
u.,)
o\
) J (n _. ___+
o -l
ll 2 .i\
..t o \
{ a s!f_H
s\ - ...-!
EPPLER
125
oct 95 4 2E-015 calber 428-2I2 428-?L5 42E-218
sta ord GA-2 upp€r lower upper Lorter upper lower
0.o0 0.o00 0.000 o - 000 0.000 o.000 0. o00 o,000 0.000
o.25 0.805 0.060 o .704 -o.584 o . 865 -O .'7 45 L.026 -Q .906
0. 50 1.loo 0.115 o.996 -O.764 L . 21,6 -O .984 1.436 -1.204
o.75 1.380 0.169 L.?73 -0.935 1.549 -1.211 L.825 -r.4A7
| ,25 1. 800 o .263 1.703 -),.L77 2.063 -L.537 2,423 -r .89?
2 .50 2.500 0.450 2.450 -1 550 . 2.95O -2. O5O 3.450 -2. 550
5. OO 3.530 o.7L5 3.539 -2. 109 4.245 -2.815 4,951 -3 .521
7.50 4 .290 o.892 4.324 -2,540 5.182 -3.398 6.040 -4.256
10.00 4.A70 L.O23 4.919 -2 .873 5.893 -3.847 6.867 -4 - A21
L5. oo 5.705 l.?44 s.808 -3 .320 6.949 -4.461 8.090 -5.602
20.oo 6.3t 0 L.420 6.468 -7.62A 7 .730 -4.890 a.992 -6 . L52
25. OO 6.780 L .557 6.981 -3.867 a-337 -5 -22f 9.693 -6.579
30.00 7.130 L ,663 7.367 -4.041 8.793 -5.46? 10,21.9 -5.893
35.00 7 -370 L .737 7. 633 -4.159 9.IO7 -5.633 10. s81 -7.LO7
40. oo 7 .4A5 1.780 7 -768 -4.208 9 .265 -5, 705 10.762 -7 .202
45.00 7.465 L.792 7.764 -4.180 9.257 -5 .67 3 10.750 -7.166
50. o0 7.300 I.764 7 .604 -4.076 9 .064 -5.536 10.524
55. oo 7.224 -6.996
6 -940 r.672 -3.880 8.6L2 -5.268 10. 000 -6 .656
60. o0 6.355 6.62r -3.547 7.A92 -4,818 9.163 -6. O89
65.00 5.595 I.374 s.850 -3.102 6.969 -4.22L 8.088 -5,340
70.00 4.770 1.189 5,005 -2.627 5.959 -3.581, 6.91-3 -4. s35
75.00 3 .940 0.991 4.143 -2.161 4.931 -2.949 5.7L9 -3.73'7
80.00 3.110 o.793 3.281 -1 .695 3.903 -2 , fI7 4.525 -2.939
85,00 2-295 0.595 2.437 -1 .241 2.890 -1.700 3.349 -2.L59
90.00 1,.495 0.396 1.592 -0,800 1.891 -1 .099 2.I9O -1 . 398
95.00 0.690 0.198 0.750 -0. 354 0. 888 -O . 492 I.026 -O.630
100.00 0. ooo o. oo0 o.000 0. o00 0.o00 0.o00 0.000 0.000
Airfoi I nurbe! 0 Drtr - 19-?6-1995
I\.\'.' 6A 4?-E -'Ll5
EFct--r2-
t.(
PLoc.nn*.
' :.,/o /D
.;
" /c-
hA +zE -z-re
^L t' v r0
i<:
------4=q<ro"
I'
-a
,AARP 5TI?
--t u-o--j/ Z tru,={r}-o,rl<
/"',r\
J, *rtrt----
--- -- I9 / tl -'
------
ll !'
o ;u ;,: i|
P iu (5o-. srA'-!, II Er ),,
()osr Stn'.) et Tqt<RN€ra
rr..* .t),.t<^rFl o,rr,\
4zL:-Lfr
I 1.,, ?--64 l-:::r::::_.: ^r€ J
0
\L
I -- I ..t-
*+Eb
) *; L,F
r o::L-
'.,.-.- {
!-A ut'l
i)et-o'or
I R''lrS t
I ; _,
a [i- _.1 _..
-": I __ 6A to*t-r(
\i c( - 4oe
"ll:'-A\f -( r "--*-dacn
"!: 11 7-y"*
4eoeo
Fe| - tofusllt +*-u{
U tlA- 1\'rlqr
-,L ,o7o ^ ot {
^
Harry Riblett
127
ADDSIDUM MJMBER 6
416 Ribtett Lane
Wilmington DE j9€09
3o'r-lat1$ -D +71 2/8/L996
National Ttans portation Safety Board
490 f'Enfant Plaza East
washington DC 20 594
Attn: Mr. John Clark, Chief, VehicLe perfonnance Divi-sion
Subject: tuneri.can Eagle ATR-72 crash at Roselawn Indiana LO/31/1994
Dear Mr. Clark:
Ity letier xo you 9/7/95 explaj,nd ,ahy the ice ridge on the ATR-72 wing, wtrich
was t-lle jfiediate cause of loss of contro.l , formed at exact.Ly the 9t chord position.
In you response Ietter I/25/1996 you ignored this point ccnp-letely, which is unfor-
tunate. for that is the key to the puzzle. Ttterefore. f arn enclosing additional data
showing that the cause of the ice rj-dge, r*rich caused the accident, is due to an
anc.rEly of the l,lACA 430xx airfoil, that is, a pronounced negative pressule peak at
the 9t chord position, exactfy where tJ.e ice ridge forms. l4ore conventional. airfoils
do not have this problern. Ttlerefore, t}re accident r,Jas caused by the p@rl-y designed,
unusual, and fortunately rarely used, }IACA 430)0( airfoils used on this airpLane.
This il.l-conceived airfoil design was part of a test prograrn conducted by I'IACA
Langley on "airfoils having the camber unusuaLly far: forw-ard", Ilere].y to see \{hat
effects the exErso-. forvard loading would have on airfoil perforrnance- hopefully. lovt
pitching nrnent coefficients, Figure 2 shov/s the forward loading scheduJ-e, and ccm-
pares it to a rbre conventionaL. unifoqnJ-y loaded ai-rfoil, the NACA 54-415, used on
the Fokker F-27 'Friendship". the theorcical pressure profiles predicced by I{ACA
for these airfoils, figure 4, shol't€ a pronounced Feak at .09C for the 430xx airfoil,
ard no peak whatever for the 64*41-5 aj-rfoil. The actual pressure distributions
(velocity ratj-o curves ) for these trno airfoils obtaj.ned by ccnputer performance
analysrs, figure G, confj,rms the predicted profiJ-es.
Ttre results of t.t.e 1935 NACA wind tunnel tests, figure 5, shoh/ that the arrfoiL
ncnent coeffici.ent ( unrelated to the Roseladn acci.dent ) is indeed loev, as anticipated.
However. the sharp and dangerous starl on these aj,rfoj-ls. as bad as any airfoil- ever
tested by NACA, is unaccepta-ble, therefo;c the tests must be considered a failure.
UnforEunately, ihese test airfoils canE to be used on actual" aj-lpl,anes, r.riti gener-
ally disappointj-ng resutts. espe.cially on propdrlven twins.
The ice ridge forms at the negative pressure peak (.09C) because that is hrhere
the adiabatic cooling is the greatest. Thls is analagous to ice formation in the
tbloat of a carburetor. In fact, tie top surface of an airplane wing is sirply
a one-sided venturi. Etcrn that point onrrrar:ds, further conjecture as to the exact
effect the ice ridge had on wing performance, aileron hinge nErnents, etc, a-re rrDot.
The point is, the unusual pressure peak on the 430xx airfoi,ls at .09c caused the
ice ridge to form at that point, and that dovrned the airplane. Ttrerefore, the IIACA
430)0( airfoil is at fault.
Accordingly, there nuy be scnre nerit for the FAA to ban airplanes usi-ng the
airfoils frcrn flying jrr freezing rain and &izzle ( figr:re I), but not
l,lACA 430:o<
all turboProp regional airliners should be so banned. To issue a blanket prohibition
is heavy-handed, irperialistic, arxi ignores the fasts in the case.
For confirriEtion of this airfoil data, check lrrith scrrEone other than:
-[{ASA, h,ho is responsible for the idiotic 430:o< airfoil;
-ATR (AIR), wlro made t}re trListake of ch@sing this airfoil for their airplanes;
-FAA, hrho made the nistake of certifying the ATR airplane with this airfoil on it.
128
.By copy of this letter' r request FAA'S Mr. salvano to incrude this letter as a
public ccmEnt in the docket rega_rding this accident, and also to include at in the
FAA'S tcing symposium schedu.ted for May 6-8 at Springfield. VA.
Harry Rib,.e*
,l11",^" [hAW--
CC: Venc]-.
,r,.
l!.^o,
n !:!_t '-,,
F
t. ;. ,-,,
lR - '1, ta 11.
129
tions ond tho.hey moke evey eftn o orcd ured for Moy 68 rn Soringfield, Vo. (ey os ;ce evidence probes mounted ouiside
inodvenently flyrng rnlb such reqions topics lo be discussed rncir.de how teez.
Tre c,ouA ot.rt ocnon stem-s hom ihe
ol'he copro,n's *,ndow Tne probe is de-
ing roin ond driz.zle should be definad, signed b occrete ice ord serve os o vjsuoj
c,oslr o[ on Americon Eogle AIR72.2OO oroblems ogsocrored wrrh forecosrrng jrs cr.Le lhot lhe oircroh l.os encounrered ic-
or Roseldwn, tnd Iawd,Sf Nov I4, I994. presence ond deveioprrent ol oclvonced ing conditions lor which it is not cerrifi"
p. 28; Nov. 7. 199A. p.361. Ahhough .ce de'ecl'on sysle.as copoble of prov;d- coted, Brodin soid.
rhe U S N,ot,o1ot T'onsporrorion Soie-v rng on eorly worning ro p'lots The ,rs€ ol The FAA rs sr'rl corducring speciol
Eoord hos not determin€d rhe probobl! orrsrde vrsuol cues olso w ll be discussed, ground6ored rests desiqned io oere.,nin€
couse ol rhe occidenr. onolys;s Indrcotes Solvsno soid. which regionol .ronspo s ore su5ceprible
,ho'on ce r dge fo,med oh of .ns w;pg "\ /e ore 1or expeoing ro moke ony oe- lo conl.ol problemg Under severe icirg
de.cers, coL:srng o.oprd rol onC suose- ciorarrons obour teez.ng ro,n ond orrzzle
q,renr loss of control by the tlignt c'ew
condit;ons. In.llighf,e5,rng \,srro o spe
oher rhe neetrrg concjuoes." ou'Florrici- cro,ly modihed U S. A,l. Foice NKt-t3SA
lAw&sT )o^ 2, 1c95. P 28). ponrs wjll be encoJroged ro openly d s. lonker oircrah is ovoiloble lor monr.rfqc.
fi,Af ACCIOENT hos not only rejnlo.ced cLsg lh€ isrue Fom o" operarronol rio.ro- lurers, bul the FAA ir, not mondo no its
conce.r obou he donqer o[ lreezino roin po'qt, occording ro Sorvono. He expe<fs ese. Solvono soid.
ord o-:2.? e. bur hos fo,ied rhe FAA ri cot the Russion deleqar,on ro p.o'rde,mpor. OPERATION Of Th! TANKtt ir funded
side'exponding furprendix C of Porr 25 bo bnt doto on Feez:ng prec p ohon becou!6 through Fiscol I996, "but we hove no
yond ;t! cunent b'eodr$. Port 25 qowm: of their extensrve llyrng experienco in s+. gvorontees beyond thi: yeor" thot the
cenif rcohon ol -ronspon Coiegory;i.croh. vsr€ weother cgnditoru. uniquo orrcroh will romoin op€.dtionol,
Exrstrng'69J'otions requirc monufoc- The FAA s drrectrve esrentiolly will mole Soi"ono soid. ll cosh monufocturors qbcu{
turafs to domonsi.otc thol on qircrofi con oll rurbop'op regronol oircroh ;equol- by $250,0OO ro $5OO,O0g ro chqaer $c
operote iolery in *oter droplers hoving prohib;fing flight in fteozinq roin ond dr,1- honker for ;cing tosts.
on overogo d;ometer ol 2O5O l"nic.ans. zle, rcid Alorn Brodin. p.esrdent o[ mor. ln oddition ro rh€ ATR72-20O, qn Erh-
The diomoler of dropreh .n.ount6rod by k6ting for Aoro lnr6rnotionol R6gionql broer EMBI20 Brosilio ir the onfy otb
lhe Amerrcon Eogte ATR72 probobiy ex-
{AlR) Morkotrng lnc. AIR is o sol€s con- 6r rsgionsl oircroft thot hos flown b+
ccod.d 200 hicro4s, occordinq to tha 3orlium o[ $6 lormer ATR, Jehtraom ond h;nd rh. NKC-135A'. Th6 in-flighr tosr!
NTSE
To iaorn more obout fraezing driz:le
Avro cornponie:. Proyid. q 'criticol doto point' lor morr
Brodin roid the snlirs U-S. flcd of no{s !,[octur6r5 ond rhc FAA ihqr connot ba
ond ro;n, lh6 FAA is plonn;ng on rnrsmo lion 158 AlX4? ood ATRZ2 oircrslr how derived ftom groundbored rcs*ng, Scl"
lionol icing !ympor;um currendy scheci- been litttd with lorgol wing deicers o: wall v{lno so.c,. I
AVIATION WEE( & SPACE lEcHNotocy/fcbrvory S. 1996 37
Frqeee I
l,.
130
\.133- l:9!t*iR::3t6
: :
F
l
.ti
rl.l
(,1
Q{
it
/"4, F .-_; __;,;, _.,
"", ; ,*,_
I t;8!1rr6
o?r 6.t q h ro tt.!
3h
ta
q: ;."!lllR:tB:o :PrttiS:!33"
f,
(l\ ^rr
:::::-5a!iro{r
t! r-
ry*€
j
!
1.
rU>
t,
$l i' ":p
li I rR ti
rl\,
t
'.(
ill- i::d jo
>+
i -* _ . G.-_
'l'
.l T
Fl o
o vJ \^t
I
-l\
.rl
,!l ,iz \
l(l
;llc v j;)
i/!€ ./
+l ', V l:'.N
<l 2 .*-o\,.\
-.^
zl :
I
T---
ldu) ro.r -+t( ca'r.^ 1 31
$!
e
r)(,sqL. F,r ,ic{r !iv! 4lo . (/.re-ZZl I't/o*"
tt1A (o|f t ; NJna
rir '&r,o,rt' ^
&\.'i
A60orr + {o.J Dotrrr"rr
v.t g.
1. rtt .
(F-rr) -$i It. ri]t .
P€ a Gdr.,n4
".t<.€ C6aeaa,r6,J - +ND NA.A (o4 - 4,5 A rRFarLs
(ae< 'eer.ea reoarea)
V€Loc'rY PLor r
r !.1.. r 7- lt9a A,ifort tru.Lr a !rl. ?'
?- ?_1991
NAc{ b+ - +tt
d. +' roe r-tt*e C. , ... ia o(. ?: T., ioJr.cE (r.r.ro
N{rAnlf ?r€treaa l^, \
t,t
+.-- @ q.lo c
'L L'
'/6
.a
16 t. a. t. ao
e
'k
t
CA
-'l
; ln
f. F
132
About the Author
June 1994
Harry Riblett
415 Riblett Lane
Wilmington, Delaware 19808
n2/994-s479
ADDENDUM t'lo. 7 133
Aircraft Srabili-tv - lri.plified
is surprising hor9 lIEry builders ask this question - ,'I am using arrfoil
(XyZ) on my airplane. hhat shou.Ld I use for ttre itt CG lirrlit? " It \.Duld be nice
]l ]! yer" rtlat sinpJ-e, but it isn't, Aircraft longitudinal stability, and the aJt
CG liJrdt, depend on many factors, and the parUicul-ar airfoil
belng used is only
one factor affecting the ploblern.
9i:g !;.Flvrnq Winq Airplanes. l,et,s bregin by looking at the npst sj:rrcle case, a
llyrng wrng. As Ln any airpLane, thD conditions mlst be satisfied, tt]al is, it
b. sta!.ica]]r stabre, and
T::, reguires that the cG be located it arso must be dvnarnicaLlv stabt;: i;trc stalirity
IrErely in rine r,ritrr tne ritt vecbo!. ttnt is, at the
center of pressure of the. arrplane. so tlrat the airplane unai,sturb& ,iff tfy
leveL. Since rer cannot shift q.t" cG fore and aft in f1ight, ,r. th;;;;r;i need, for
(
riarrlt nibJ.ett
4/17/1996
136
{.8
o
g lr
I
f
E
;
a
i - rrPtJtlot ? >
' _1,.r-_. t._-lIrq,'a'P
: I
tr'
{,,. ti
d €d
i .tr
E5t fl
{a
o
,? fl
It:tl
!.1
J
-l rl
?g I
,.
Jl
ir -r
Ir^ er
ci t- t
! ila
-
'.'1 q
a, {lr
ra ?"*
a! ti
j-.
-t lo f :
r:.l t
.:
lo
It
t !,*+l h o
I
.t
.<t;l
;r
{q
ol
's t
nt 3 E, f
{ I J t
.,J
!: 1
I
ADDNiDUM NTI4BER 8 137
Vortex C,enerator Kits
This paper discusses the highly effective vortex generator kits that are avaj-l-
able to lolver stall speeds and inprove lcrvr-speed controllability, especialLy on
prop-driven twins with one engine out. the subject was discussed at scrne length in
my papers "Design Notes for Tapered [iings", and "Tanr.ing the Sharp Statl of t]e NACA
2301-2 Airfoil". !'Je also elaborate on an artj.cle concerning lAcro-Dynandcs Inc VG
kits printed in the January 1996 issue of "Arr Progress', magazine, and we strongly
reccnrrend that anyone interested in the subject should read that exceLlent article.
The vG kj-ts basically contain tr4o sets of vorEex generators. one set j-s install-
ed on the top surface of the ping. across the entire wlngspan, at the fOg chord pos-
ition. the second array is instarled as crose as possj-b]e to the l-eading edge of the
rudder, both sides. In ssrE cases, this second set is instaued near th6 triiung
edge of Lhe fin. The prirE y effect. of the vc's is to i,ncrease the trrickness of
the bounciary rayer doenstream of the vc,s. thus, they increase the effective thick-
ness of the wing, just as if "gon6.' r*ere appl_ied to the top surface. And since the
added thrckness is al-.l on the top surface oi the wing, this causes an uF.rard shj-ft
of the airfoil mean line, that is, an increase of wing carnber. Both the increased
thickness and the increased carnber result in a higher rnaximurn lift coefficient for
the wing' l"bre urportantry, hcr.€ver. these vc's, when placed at the ,rOc position
on the I.IACA 2301x airfoils that are used on npst GA liaht twins, el-iminate or at
.least soften the needless and troublesare discontinuity in the rnean li-ne tlrat these
airfoil-s have at the .15C Position. This el-minates the fl-c'vi separation that occurs
at that point at high angles of attack. l,Ji-thout the frorr, selErati.on, the stalr be-
cdr'es soft and gentre. and this is the key to good aileron i"=ponse. even with one
engine out. Ttrus, for the 2301x airfoiLs, the vc's give rnproved controJ-J-abirity
( eh-mination of \nrc
) as \nelf as a higher C]nax, ,direreas on the Cube and Charnps, \.!trj.ch
a]ready have soft-stalr airfoils, the prinary effect is sinply an increase in clnax.
The rudder VG's reduce fls^, separation on the 1o,l pressure side of the defl-ect-
ed rudder. without vc's, the si:rp1e hinged rudders used on nlfst of today,s light
twi'ns llave terribl-e flovr separati.on at the hinge rine, tl.us high drag, ihen one
engine quits ard hard rudder opposite the dead engine is applied to iold headrng.
In this condj-tion, the rudder beccnes a large, effective drag bra-te, just what ire
don't need.
r'-nus the vc's a.re very effective band-ards that address the poor wing and tail
designs that ' are used
on today's Iight tvrins (and ccnrnuter twlns ai well). Of course,
the better approach is to design and build the planes properly in tbe first p1ace,
Three things are required, at1 withi-n current tectnorogy: (l) soft stalr airioil-s,
such as "GA" airfoi-rs (2) t'ting tips ar least l2t thick ior turbulent airfoirs, and
I5t thick for lani,nar flol airfoils, and (3) Low drag. effective rudder designs,
such as the articulated rudder used on the USAF C-I7.
Hhy does Vlichi-ta continue to build this trash, htlen \i€ knov, better? It,s the
fauLt of the FAA t1,'pe cerLificate systern, for the re-cerEification costs associated
vrith najor design changes effectj.vel-y freezes t].e design. Thus rne are forever v€dded
to 40 and 50 year-ord technology. to the detrirent of aafety, Ttris must be changed.
Ttre first thing that FAA shoul-d do is to rescind ( after a suitable grace period of
perhaps trno years ) the tl,?e certificates of at1 prop--driven twins that cannot show
a unc lo,rer thEn the normal stal1 speed of tl'e a-irplane. Ttren the arended type
certificate for the rnproved "no \.trc" nrcdel shoul-d be issued follo,ring a sinpie
fl-j-ght denpnstration. wj.thout the prohibitiveLy expensj-ve "normFl-" c"itificallon
procedure. Imported airplanes t"ouJ-d have to me€t the sanE perfornEnce reguirenents.
VJj-th that, 'ne',^ould at last tEve true twin-engine reliability, whrch we do not
have with today's obsolete, garbage airplanes.
FRIDAY F
A(r.
The kit tor lhe Baion 5581C/!/E contains 106 vodex generalors mounted on lhe
wirgs and on both sides of the rudd€r, plustuo strakes on tl* outside ol theengine
nac€lles, ,t redlces the Baron's Vmca by 10 knots and sells tor S2450,