F

1

, -

CHORDWISE AND SPANWISE LOADINGS MEASURED AT LOW

SPEED ON LARGE TRlANGULAR WINGS a

- By Adrien E. Anderson

" _c

Pressure distributiom have been obtained *om three triangular wing modela : a -lone model having 89 aspect mtio of 2.04 and a modified h u b m e airfoil eection, 'the same wing caibined ulth a body of fineness ratio 12.5, and a mockiup of a tri" sirplane which had an aspect ratio of 2-31 and &p =A 65-006.5 sirfoil section. Pregsure data were obtained through an -1-f- attack range at zero angle of sideslip. The Reynolds nmiber a: based on the me= aer-c chord, was approximately 15 X 10 at a correspondhg Mach mer of 0 .u.

Chordxise pressure distributions, section Xf't coefficients, and section centers of preseure are pr@sent@d for several spamrlse stations and anglee of attack. Span load distributiopa are also included.

Camparison of the reaulta on the several configurations indi- cated that eimilarity among the corresponding wing characteristics (the chordKise diatribution of pressure and the nearly elliptic span- wise loading) occurred o n l y at the lower angles of attack-where errsentially inviscid potential flow existed on the wings. In the middle angl-fdttack range, the aerodynamic characteristics of the three xings differed. Lea-dge separation, which occurred on the tip sections first and then progressed inboard, appe&,red to be dependent urn. the curvature of the forwetrd portion of the airfoil section- the sharper the nose section, the earlier the separation. kadtqpedge separation resulted in a form of air flow which produced abrupt changes in the section characteristics, but only negligible changes in the integrated wing characteristics. At angles of attack near Xing stall the correspondfng w w e c t i o n characteristics for the modela were again very similar.

?.

r

I UNCLASSIFIEG

2

It uas inferred from the datu that the body carried a l i f t approximately equal t o the l i f t which would be c m i e d by the xlng area it covered.

IERIRODTECTION

Although low-aspect-ratio w-8 of triangular plan form are prim- r i l y intended for use at supersonic epemds, the designer of an airplane employing this wing design is also concerned with their lar-speed char- ter is t fcs . With regard t o the low-speed pressure distributicm on tri- angular w i n g s of aspect r a t io 2, referance 1 reported the result8 of an investigation on a triangular wing with a subsonic-type airfoil section of moderate thiclmees (NACA 0012). Due t o the interest in wings ham thfnner sections, it was considered desirable t o investigate the lo-+ speed preesure distribution on 8 wing having the same glan form and aspect r a t i o but a supersonic-type a i r fo i l section.

In addition, an appropriate strpersonio-type body was added t o the wing t o determine the effect of 8 body on the loading of a triangular plan-form wing of low &spec% ratio. Finally, the mock”up of a triangw It

lar4ng airplane of approximately the same aepect ratio, but having a thin low-drag subsonic-type a i r f o i l section (RACA 65-006.5), war3 made available t o the RACA for investigation.

e

The results of the pressure-distributian investigation on them large-scale models have been 8rmmml.ized in this report in order t o d e the much needed data available t o designers of aircraf t with w i n g 8 of triangular plan form. This report; then, in conjunction with reference 1, makes available a comparison of the loading on two triangular wings, one having a relatively thick subsoni-type ernd the other a thin super- sonio“type airfoil section. In addition, the data presented in this report indicate the effect of 8 body on the w i n g load distribution as well as prodding a qualitative comparison of the loadings on two wing- body cmbinationa, one having a thin subs0ni”type and the other a thin sugerecmic-type a i r f o i l eection.

The symbols and coefficients ueed in this report are defined as follows :

A aspect r a t io

a free-etresm angle of at tack. of w i n g chord plane, degrees

RACA RM NO. ~ 9 ~ 1 7 3 a

b wing span, feet

A C wing chord, measured parallel to air stream, feet

average wing chord (S/b) , feet - C mean aeroaynamic chord, measured parallel to sir

/ section lift coefficient

wing lift coefficient (9)

P

P t

P

- Q

S

X

X f

Y

fre-tream static pressure, pounds per square foot

local static pressure, pounds per square foot

freestream dynamic pressure, pounds per square foot

wing area, square feet

distance along chord Fram leading edge, feet

distance along Melage center line f r o m nose, feet

distance d o n g wing semispan to chord location, feet

4 RACA RM Eo. AgBl7

The three models used in this investigation were:

1. A wing model, hereafter referbed t o as the mdifieb-wedge ving, which had a triangular plan form and an aspect r a t i o of 2.04. Thle airfoil section parallel to the wing center line was derived frum 8 synrmetrical dauble-wedge secttan having a maximum thicknesa of %percent chord at 2O”percent chord. The modification conrriated o f - a 0.0025~ =Be radiue and a rounded maxfmum thicknee6 formed from an arc (0.62~ r a d i u s ) which uas tangent t o the eurface of the double wedge a t 1% and *percent chord. Both the tog and bottom of the section were rounded. In term of the resultant chord, the airfoil section had a nose radlue of 0.00254~ and a thickness of 4.83 percent a t 21.6-percent chord. This was the same wing f o r which force data were presented in reference e.

2. A modified.Kedge winpbody model, which consisted of the a b o v m t i o n e d wing combined with a slender and polnted body of revolution. The radius r of‘ the body at any stst ion xf was obtained f r o m the following eqpation:

3. A mock-up of a triangula3Ywing airplane which had an aspect ra t io 2.31 and an HACA 65-006.5 airfoil section parallel t o the King center line. This model, which will be referred t o as the N M A @- series wing-body model, wa8 equipped with conatanhhord trailing- edge controls. The control had a horn balance, a nose radius, esd a smell but unsealed gap. The ducting sgetem was gpen and the power plant removed.

We-view drawings of the two K.Lng-baciy modele appear i n figure 1, while figure 2 contains photographe of the models as mounted for t e~ l t ing in the Ames 4” by 8&foot wind tunnel. Additional dime- eional data for the three models and the ordinates of the W A 6 5 ueries airfoil section will be found i n tables I and 11, r e s p e e tively.

The pressure orifices, in the case of the modFfied”redge wing, w e r e located along chord l i nes a t six spaznriee s t a t i m in the top and bottom surfaces of the right half of the wing. (See fig. 3(a) .) The body ueed in the modifiebiwedge w i n g a o d y conibination covered the pressure orifices along the chord l ine a t the 0.05 semispan

KACA RM No. AgB17 5

station. lh the HACA 65se r i e s wing-body model , the orifices were located in the left half of the wing along a chord l ine st the body juncture and four lines of oonetanfypercent local span. (See fig. 3(b).) There was a slight deviation of the orifices framthfs linear distributipn 88 8 result of interference of the orifices xith the intern81 structure of the xlng.

The preesure data were obtained through the angl+of-attack raage a t zero angle of aideelip f o r the three models. An addttiollgll run -6 made on the HACA 6$-eerieaoWing-bod.y conibinatian dth the trailing-edge f lap deflected up 10 . The aynamic pressure f o r the tes t s uas approximately 25 paunds per aqusre foot (Mach nuuiber = 0.13) with corresponding Repol- ntrmbers (based on the mean aerodyngnic chord) of 14,3 X 10' for the modifiedbredge wing and 16.4 X 10 for the NACA 65se r i e s xing"body c d i n a t i o n .

The pressure coefficiente, reduced from the .measured local s t a t i c pressure^, are believed accurate t o within percent. The chordwise preseure plots for the modifietkredge King and win&+body models were obtained directly since the pressure orifices were located along ohord lines. However, on the W A 6 w e r i e e x ingaody model, the fact that the majority of the pressure orifices were not located along chord linea llvrde it neceseary t o conetruct preltm-lnnry plots of the data (as indicated i n f ig . 4).

Val~es of section nol?lbal-force coefficient and eection center of pressure were derived from the chordwiee preaaure plots by means of mecWcal integrat ion and calculation. The value6 of aection l i f t coefficienta do not include the effects of the forcea pars l le l to the chord. Representative calculaHana of the chordwise forcea indicated a maxlmuIIl increase of only percent in the section lift coefficients.

when the KACA 6meriea -ody model data are taken into consideration, the lack of preesure orifices near the leading edge and the wing t2y makes the nature of the data somewhat questionable in these regions. On the other hand, BP integration f o r a = 8.3O of the section normal forces plus an spproximebtion of the force on the fuselage gave a value of CL which m a only 8 percent less than the force test CL. With such agreement it is reasonable t o expect that the result8 far the RACA 65serles *body model are also of sufficient accuracy t o be indicative of general trends .

The valuee of wing lift coefficient used in computing the span- loading coefficient8 were obtained from the force-test lift curves of figure 5. The w e of liftcoefficient values from the force test was necessary since the preesures on the body were not determined.

The values of the angle of attack of the wing have been corrected for a m t r e a m inclination and for wind"tunnel"wal1 effect, the latter correction being that for a wing of the same pan and having . elliptic loading, but with an unswept plan form,

RESULTS AND DISCUSSION

Separati-Vortex Air Flow Over Triangular W F n g s

In reference 2 it was reported that, on certain occasions dm+ ing the forc"best investigation of the King with doublewedge aec- tim, condensation trails appeared revealing the presence of two vortices extending downstream somewhat in the manner shown in figure 6(a). The vortices, which resulted when the eeparated flow off the sharp Leading edge coalesced, first appeared at an angle of attack correeponding to a wtng lift coefficient of approximately 0.6. HO 8

c o n m a t i o n trails were ever noticed at the lower angles of attack, or when the wing leading edge wss rounded. This lack of vapor trails did not necessarily indicate the nonexistence of the vortices; they may have been weaker.

.c

During the investigation reported herein, vapor trails did not appear on any of the models. In the case of the modifieihredge wfng model, however, their prelaence was detected by means of a survey of the flow above the wing. Their presence i s also apprent in the chordwise pressure plots for the modifiedwedge wing and -body models. There is some evidence in the pressure plot8 for the HCLCA 6 ~ e r i e s -body model to indicate that separation vortices may have existed on this model also. An understanding of the pattern of this vortex type of flow will aid in the interpretation of the resulta of this investigation.

$

The following description of t h e f o m t i o n of t h e separation vortices is based on t h e visual observation of vapor trails on the wedge w i n g (referenoe 2). and the survey of t h e flar above the mod3- fied-wedge-wing model. The vortex, oonaidering one-half of the wing at a time, originated firet on the lead- edge near the wing t ip , following separation of the flow Fn t h i s region. Line A of figure 6(b) represents the vortex pattern for a = 30. hcreming t h e angle of etttaok caurred the point of origin of the vortex to move

NACA RM No. AgB17 7

.

J

.

forward along the lea- edge toward the apex of the wing until the point of origin reached the apex at about a = 8 O . The angle of sweepback of the yoetex increased with angle of attack, the rate of increase betag largest at the lower angles of attack of the wing. Also, at the lower angles, that portion of the vortex lying over the outer 30 percent of the wing spen uas bent back elightly toward the free stream, m e latter condition is represented by line B of fw ure 6(b). A t an angle of attack of about 18*, the angle of sweep back of the vortex seemed fairly well established and apparent- the strength of the vortex begas to dlmfnish toward the apex (the yapor

had stalled, all evidence of the vortex had disappeared. , trails of fig. 6(a) or line C of fig. 6(b)). By the time the wing

G e n e r a l Comments

Catparison of the results of this investigation with those for the wing with M C A 0012 airfoil section ,(reference 1) indicated that some generalities exist concerning the wing characteristics which should be kept in mind throughout the discussion to follow. Similar pressure distributions occurred over the wings only at the lower angles of attack where esaentially fnviscid potential flow existed on the w i n g s . The preesure dietributione became dissimilar in the middle angl~f-attack range follaring flow separation from the leading edge. The occurrence of lealllng-edge separatfon and the intensity of the resultant vortex-type flow appeared to be dependent upon the airfoil section. The modifiemdge wing and win@~ody models showed signs of leadlng-edge separation first and a stronger' effect of the vortex-type flow. In the case of the HACA 65series -body model, separation was delayed and the effect of the vortex lese strong. It is d2fficult to conclude f r a m the pressure diagrma for the model with NAca OOl2 sections (reference I) i f the vortex- type flow was present. The nrinor bumps in the pressure diagrams make it appear likely that this model had a very weak vortex-type flow. At angles of attack near wing stall, where the Fnfluence of the vortex--type flow had became negligible, the characteristics of the three models were nearly similar. In general, the characteristics of the NACA 6Fseries w i m o d y model more closely reedled those for the model with W A 0012 eection than those for the two models with modif ie-dge sections.

Chordwise Pressure Distribution

The chordwise pressure diagrams for the modifiei+wedge-wing model are presented in figure 7. Whereas data were obtained at very

8 NACA RM No. &IS17

small increments of angle of attack on the modifie-dge wtng and King-body models, results are presented for o n l y those angles for which the pressure diagrams are indicative of the general changes in loading on the wing. The section loading increased toward the wing tip at the very low angles of attack, that is, up to 4.2', when essentially inviscid potential flow ejListed on the wing. High tip loading at law angles of attack was 8 characteristic of the NACA W E King (reference 1) . Howeyer, the tip sections of the modifieb wedge wing stalled at much lower angles of sttack, which was quite likely due to the relatively small nose radius and the abrupt change of curvature xhich results at the juncture of the m B e radius and the sides of the wedge, in contrast to the smooth transition on the NACA 0012 section.

The first signs of leading-edge separation are found in the chordwfse pressure distribution f o r the 0.p-Semispan station at an angle of attack of 4.20 (fig. 7(c)) where the value of the negative pressure at the leading edge has decreaered slightly from that for the previous angle of attack and there is a nearly unfform negative pressure area extending over the forward half of the chord length. As the angle of attack wae increased, the seperated flow area moved progreeslvely inward a8 indicated by the chardwise extent of the nearly uniform negative pressure area on the inboard sections.

The separation of the flow did not result in a subeequent loas in load on the section. Any loss in load as 8 result of a loss In peak nose pressure was counteracted by an increase due to the 1- pressure bumps in the chordwise pressure diagrams, which resulted when the separated flow gradually formed the previously described vortices. While these bump6 first appear in the pressure diagrams for the outboard stations at the low mgles of attack, they are more apparent over the major portion of the wing at an angle of.attack of 14-5O. (See fig. 7(e).) When the diagram of this figure are r e b u n with the abeciesa proportional to the local chord length, as has been done in figure 8, the trace of the vortices over the wing becomes even more apparent. In this figure, positive preeeure coefficient values are dtted.

The bending back of the outer portion of the vorticee is indi- cated by the rapid rearward shift of the brmtr, on the preserure dia- grams f o r the 0.75-semispan station through the angl-f-ttack range between 4.20 and 14.50. The effect of the decrease in vortex strength is found in the preserure diagrams for the inboud statione at the higher eagles of attack where the reduction in the efze of the bumps is more pronounced. For example, notice the formation and the s u b sequent reduction of the larpressure 'frump at the 0.4f5-~emispm

I

L

NACA RM NO. ~ 9 ~ 1 7 9

station. A t Q. = 8.30 (fig. ?(a)) separation has taken place and the vortex has formed. In figure 7( e) the bung has increased both pres- surexise and chordwise. Hear a =.20,80 (fig. 7(f)) the bump has be- g u n t o diminish in s fze and’is no longer apparent at =.= 26-90 (fig. 7 ( g ) ) as a result of the vortex dhinishing.

The pattern of the chordwise loading at mayimum lift (fig. 7(h) ) is very similar t o that fo r the wing with aAca 0012 section ( r e f e r ence 1). The preesure diagrams f o r the stations outboard of 0.60 semispan indicate comglete sepration;with the negative pressure values increasing toward the root. The diagrams fo r the inboard stap t ione give indication of various degrees of separated flow, with the most inboard station having the- least si- of separated flow.

The preasure-dietribution curv-es f o r the modifM+wedge xinep body model are preeented in f igure 9. The prlrmarg effect of adding the body w a ~ t o shift the origin of the separation vortices outboard. The outward shift of the vortices is mre clearly seen by comgaring the pressure curves f o r the -ody conibinstion (fig. 10) with the Wing-alone pressure curves a t t he same angle of attack. (See fig. 8. ) Although the bumps on the inboard stations had a more negative pres- sure peak, they extended over a shorter chord distance, the net result being a reduction i n lift on the m o a r d stations as compared t o the wingglone model.

The chordwise p r e s s m s t r i b u t i o n curves f o r the RACA 6 M e r i e s “body model are presented in figure ll. In general, the pressure diagrams for this model closely resable those for the W A 0012 wing (reference I). The chordwise distribution of pressure Etnd t€ie value of the w i n g lift coefficient at xhich separation occurs on a given section are slaibr. A t the 0.60-semispas station, f o r -le, the pressure diagram agree closely a t t h e law angles of attack and probably would show closer agreement in the middle rmge of d e of attack had it been p0ssib;le t o record pressures at the very leading edge of the RACA 6j-series King. At maximum lift, the corresponding pressure diagrams for the two winga are again very similar.

The data outboard of the 0 ,hemiapas s ta tdon on the IUCA 65- series -body model are open to qwation bemuse of the previously mentioned scarcity of Treasure orifices. For example, the presaure diagram for the O.g€Lsemispan station give the hrpressian that separated flaw existed even a t the vei-y lowest angles, which i a p r o h ably not the cme. Following complete separation, however, t h e p r e e sure diagrams f o r the O.~emispsm s t a t i o n r e e d l e thoae f o r the m o d i f i e & w e d g a model and the value8 of the eection UFte c-e favorably. It seem8 logical t o surmise then that, a t angles of attack

10 NACA RM NO. ~ 9 ~ 1 7

greater than 12O, the 0.gQ-samispan"station diagrams are indicative of the actual chord loading.

The discontinuities which appear +e pressure diagram a f t of the 0 .khord s ta t ion a r e the preseures at the control-mrface slot and not the previauely mentioned bumps due to the separation vortex. There is some evidence in the pressure diagrams which indicates the presence of a vortex. The pressure diagrams fo r the 0.45, 0.60-, and 0.7!3-eemispan stations in the angl-f-attack range between 8 . 3 O and 20.8' (figs. 11( c ) through ( e ) ) have negative pressure peaks which move aft and broaden out with angle of attack. These peaks do not appear t o be as extensive or as high ae those on the modif ied-wedg- wing model. It is believed, therefore, that a vortex-type flow did exiet, but it m a one which was samewhat less intense than on the models vith the modifiekwedge a i r f o i l section.

The close reseniblance of the IWCA 6weries Xing and the NACA 0012 w l n g data seems to indicate that the rate of change of curva- ture of the section near the leading edge, as well as the nose radiue and thickness of the section, has a large influence on leading- edge separation. Consider the values of nose radius and tbichnees f o r the three sections under discussion:

Muximum thicknelss Rose radius Section percent chord percent chord

1.58 25

.2a

It seems unlikely that the small differences In thicknes8 and nose radius could account for the large differences between the pressure dtagrame for the modified wedge and the NACA 6Weries sections. The more probable factor is the rate of change of curvature of the forward portion of the airfoil section.

Cosrparison of the preS8ure-diStribtltiO~ diagrams for the NACA 6Weriee wing-body co.uibhation, having the trafling-edge controls deflected up 10' ( f ig . 12), with those for controls neutral ( f ig . 11) a t t he 8me angle of attack indicates the normal type of sh i f t of the loading on the aft portion of the section chord.

Section L i f t Characteristics

The variation of section lift coefficient vith w l n g angle of

. c

NACA R?4 No. AgB17 ll

t

attack is shown in figure 13. As might be anticipated f r o m previous work, the lift curves are nonlfnear and the lift-cme slope *rases toward the tip. In the case of the modified-xedge wing, the formation of the separation vortex resulted in a a appreciable increase in lift on the section, particularly on the outboard atationa where the infl- ence of the vortex flow was the strongest. This effect is indicated by the increase in slope of the section lift curves of the outboard stations above a n angle of attack of 3O or 4'. It should be remembered that the section lift values are plotted against wing angle of attack. Actually some change in the l o c a l induced angle of attack may occur simultaneously with the change in section l i f t such that, if cz were plotted against the l o c a l angle of attack, there would be no apparent increase in the rate of change of section lift with angle of attack.

Adding the body to the modlfiedrwedge wing resulted in a slight reduction of lift on the M o a r d etations at a given angle of attack. The cause of this effect is the fact that adding the bow shifted the origin of the separation vortices outboard and hence the vortex was weaker for a specific station at the 8ame angle of attack.

The section lift curve for the 0.2-emispan station of the ISACA 6~series wing-body model (fig. l3(c)) is nearly coincident with the l i f t curves for the same station of the modifie&wedge -body m o d e l . At the 0.43, 0.60, and 0.75 stations, bowever, the c m e s for the W A 65series wing-body model are less steep fnitially, but do have a rapid rise, the latter occurring at higher angles of attack. Here is evidence pointing not o n l y to the existence of a separation vortex but to its delayc-f;o higher angles of attack by use of a subsonic-type airfoil section. The section lift curve for this model at the 0.- semispan station bear8 no reseniblance to the curve for the modifieb wedge wing-body model over the first 12' angle of attack, as would be expected from the previous discussion of the pressure diagrams for this station. At the higher angles, where it is re860n8ble that complete separation has occurred, the tX0 curves for the 0.9Cbsemispan station agree more closely.

The section l i f t curves, particularly those for the modifie& wedgewing models, indicate a primary cz- followed by a l o s s in lift and e subsequent recoveryto a secondary cz-. In several cases, section lift coefficients for intermediate angles of attack were computed to verify the fairing of the lift curves in the region of the primary CZ-. The form of the lift curve in the region of the primary cz- appears to be directly related to changes in the vortex pattern. Cansequently, the section lift curves fall i n t o three classifications as regards the loss in lift following the

12 W A RM Bo. AgB17

primary cZ-, The curve for the 0.90-semispan atation indicates an abrupt lose in lift a8 a consequence of the rapid ehift of the vortex Fmrard from the tip. In contrsat, the curve for the O.’lf)-semiepan etation ha6 a nearly flat top ae a reault of the eteady re-d shift of the vortex while maintaining its intensity. Qn the Inboard statione the vortex did not ahift appreciably but decreased In intensity slowly, resulting in 8 lift break which, while extending over a short angl-f-attack range, ie not as great as for the out- board stations. Primary cz- point8 are found in the lift curves for the U C A 6 ~ e r i e a *body model also.

Were the separation vortex not present or its influence negli- gible, 86 on the most inboard statio-, no primary CZ,, would exist and we would find only the eecondary CZ-, or normal aectlon etall. The values of the secandary cz- on theee models varied from approximately 0.5 near the tip to as high as 1.6 at the root section Fndicating a strong three-dbmmional effect, for the value of c usually associated with thin wing sections is 0.8.

%lax

The integrated effect of the eection lift waa f o U to be ger+ erally smooth with angle of attack as indicated by the cumes of lift versus angle of attack for the four modela. (See fig. 5 , )

!he slope through zero lift of the section Uft curves of figure 13 has been plotted along the wing span in figure 14. The curve through the points was drawn asaumiag elliptic loadfng and an average value of wing lifticurve elope of 0.040 from the force-test data. The close agreement of the section lift values to the t h e e retical dletribution points to the exietence of appro-tely elliptic loading.

Center of Preesure

The chordwise centewf-pressure variation with angle of attack for the modela ie preeented in figure 15. The separation vorticee of the rnodifieb-wedgdng model are again noted to have had a etrong influence on the section characteristics. The center of preesure on the 0.Omenispan station remained practically constant at about 0.33 x / c On the next three statione outboard, there m e a forward movement of the cent& of pressure as the vortex flow bui l t up and then a gradual rearward ehift a s the influence of the vortex decreased. At the O.>semispan station, however, the resnrard shift of the center of preasure was quite rapid between a = bo and loo, 3

aB would be expected from the rapid rearward shift of the vortex along the section chord. The shift of the center of preseure wae

NACA RM No. AgBl7 13

then forward as the influence of the vortex decreased (a = 10' to 18O). The tip station followed 'a somewhat similar pattern.

The addition of the body to the modified-wedge wing shifted the origin of the vortex outboard and hence chmged the centers of pressure at a given angle of attack. These changes in the centers of pressure were most pronounced at angles of attack above 6O, with the 0.75 and 0.90 ,stations being least affected.

The rearward shif't of the centers of pressure on the HACA 65- series wing body was less abrupt and extended over a shorter chord- wise distance. The resultant variation points to the influence of a weaker vortex-type flow than on the modifieb-wedga models. Deflecting the controls up loo resulted in a forward shift of the centers of pressure, a8 would be anticipated.

The integrated effect of the eection center-of-pressure Wri- tion was found to be generally more gradual with angle of attack, as indicated by the curves of lift coefficient versus pitching-moment coefficient for the four models. (See fig. 5 . )

Span Load Distribution

The s-loaMstribution curves (fig. 16) were approximately elliptic in shape at l o w angles of attack for two modifie+wedge wing models. Once the tip section stalled (a = 4.50), there was a progressive shift of the load Inboard. As each section stalled, the next section inboard carried a greater load, until ultimately the loadlng might be described as almost triangular in shape. The addi- tion of the body, as might be expected f r o m the discussfon of the section lift curves, shif'ted the loading outboard slightly at the higher angles of attack. The relatively close agreement of the wing- body span load distributions with those for the wing alone and the close agreement as to the total lift f r o m force test (fig. 5 ) indi- cates that the body carried a lift approximately equal to the lift on the wing area which it covers.. The distribution of this l6ad along the f'uselage, of course, remains to be &%.ermined.

It would be anticipated that the loading on the NACA 65eeries -body model would be nearly elliptic at the lower angles of attack froln the fact that the two modifi&wedg~wing models and the NACA 0012 wing model had elliptic loading at small angles of attack. However, the span loading curves (fig. 16(c)) do not appear to be elliptic at any angle of attack. This failure t o show elliptic loa& ing may be due to the chordwiee s lo t at 8Fpercent semispan produced

14 NACA m NO. ~ 9 ~ 1 7

by the horn balance on the control surface or to the deduced values of c2 near the tip being in error due to the previously discussed arrangement of pressure orifices. Some assurance of elliptic loading is found in the spanwise dfstribution.of sectioelif%-curve slopes (fig. 14) where it will be noted that, except for the 0.90 statim, the points representing the section-lifkurve slope for the three models agree quite well.

Comparison of the results of this investigation with those for the wing with NACA 0012 airfoil section (RACA TN No. 1650, 1948) indicated that aome generalities exist concerning the loading characteristics of triangular plan-form w i n g s . Similarity among the corresponding Xing characteristics (the chordwise diatribution of pressure and the nearly elliptic spamdse h a d i n g ) occurred only at the lower angles of attack where essentially inviscid potential f l o w existed on the wings. In the middle angle-of-attack range, the aero- dynamic characteristics of the three wings differed. Leading-edge separation, which occurred on t h e t ip sections first and then progressed inboard, resulted in a vortex-type flow. The intensity of the separation vortex and the angle of attack at which it first occurred on the wing appeared to be dependent upon the curvature of the forward portion of the a i r f o i l section. It occurred earliest and most intensely on the wings Kith the relatively sharp leading edges. The effect of the separation vortex was to produce abrupt changes in the section characteristics, but the effect on the inte- grated wing characteristics was generally negligible. At the angles of attack near the wing stclll, where the effect of the separation vortex -8 nil, the corresponding section characteristics for the three wings were agaln very similar.

It w 8 ~ inferred from the d a t a that the body carried a l i f t approximately equal t o the lift Which would be carried by the wing erea it covered.

Ames Aeronautical Iaboratory, National Advisory Committee fo r Aeronautics,

Moffett Field, Calif.

I

4

HACA RM No. A9Bl7

1. Wick, Bradford H.: Chardxise and Spanwise Loadings Measured at Lar Speed on a Triangular W i n g Having an Aspect Ratio of Two and an RACA O O E Airfoil Section. N M A TN No. 1650, 1948.

2. Anderson, Adrien E.: An Investigation at Lox Speed of a La;L.ge Scale Triangular Wing of Aapect Ratio %.- II. The Effect of Airfoil Section Modifications and the Determination of the Wake Downwash. RACA RM No. A7H28, 1947.

16

4''

I I t e m =de'

w i n g Span, feet

Area exposed outaide of - 307 Area, square feet

25 .oo

fwlage, square feet 307 Mean aerodynamic chord,

feet 16.37 D i h e d r a l , d s m e ~ 0 Angle of incidence, degrees

2 .Oh Aspect r a t i o - "

IRngth, feet "- Maximum diaaaster, feet "- Finernee ratio Maximum diameter wing-

"- span ra t io "-

Fuselage

Trafling-dge control Area ( t o t a l aft of hinge

A r e a ( t o t a l w f t h horn ) , SqUere f 00% "-

balance), square feet "- 1 I

Station

0

1.25 .!X

. 2.5 5 -0 7.5 10 00 15.0 20 .O 25.0 30 .O 40.0 50 .O 60.0 70 .O 80.0 go .o 95 90 100.0

Ordinates

0 +- .5l2 +- .771 f 1.032 z 1.415 k 1.719 f 1:gn f 2.379 f 2.685 * 2.920 * 3.087 f 3.244 f 3.133 f 2.719 f 2.095 f 1.327 * '548 f .210 0

L. E. radius: 0.2s percent chord

c

r

!

I 1 - 0

- - u

fa) Mod;fied-wedge

Figure L- General arrangemenf of fhe wing-body mode/s investigafed

+

20

1

f3.03'

I 4 I ! \

" t-

"37 (b) NACA 65-series wing-body mode/.

Figure 1- Conchded

~ C A BM NO. ~9.~17 21

(a) Modifi-wedge Kfng m o d e l .

t

- . .. - . . . . . . . . ... .

4

....

..

. . . . . . . . . . . .

* . . . . . . . . . .

4 4

.

. . . . . .

i .I

. . . . . . . . . . . . . . . . . . . . . . . . . ....

D I 1 I

(b)NACA *se&s whg-bot@ model.

Figure 3.- Approximate pressure &ifice distribution.

. . . . . . . . .

I

3

. . . . . .

I

I 1

Figure 4- Method of deriving chwdwjse pressure diagrums on NACA 65-serjes wing-body moobl. 3 .-

- . . .. . . . . . . . . . . . . . . . . .

. - . . ... . . . . . ..

I I

c

.. . , . .

4

Figwe 5.- Force test lift and pitching-mnt w a s .

ro \D

I . . .. . . . . . . . . . . . . .. .

I I

... . . . . . . .

I

. , . . . ". . . .

' . '

. .. .. .

NACA RM NO. ~ 9 ~ 1 7 33

I

I (b) Variution of vortex patfern with angle of affack,

modified-wedge wing.

34 HACA m NO. ~ 9 ~ 1 7

Flagged symbds .und dashed curves indicate / o m surfoce p r e m s .

Chord sfafion, x/c

figure Z- Pressure distribution alon chord for varfous angles of oftack of modified-we B qe wing model.

mm BM NO. ~ 9 ~ 1 7 c

35

Chord station x/c

(b) a, 2.i0 .

-6.0-

-5.0 -

-4.0 - 0 . *? $ -3.0 -

z -2.0 -

-1.0-

Chord station x/c

(b) a, 2.i0 .

HACA BM No. A 9 1 7

- 6.0

-5.0

k

%' C

-4.0 " ,

a?

-3.0 P, 0 Q

-2.0 5 $ -1.0 u, u,

0

1. oo 1 I I .2 I

.4 .6 .8 -To' Chord station, x/c

(cJ a, 4.2".

Figure Z- Continued

37

NACA RM NO. ~ 9 ~ 1 7

-6.0

-5.0

-4.0 " C

*? $ -3.0 p1 0 CI

-2.0 s 8 -f.O

rr, rr,

0

39

r t

"3 nl I

I. oo 1 I I .2 .4 .6 .8 Lo -557 1 I . . J

Cbord stofion, n/c

(0 a, 20.8O.

Figure Z - Continued

c

NACA RM NO. ~ 9 ~ 1 7

-6.01 f 1 I

4

LT\ -*"-*"4---4-- ' ' ' " -"/ ""

90

HACA BM HO. ~ 9 ~ 1 7

-6.0 -zol . r f

\

-5.0

-3.0 2 vl 5 9 -2.0

-/. 0

/h) a, 33.0".

41

- ... - . . . . . . ". .

I

. . . . . . . . . . . . . . . . . . . . . . . . - - . . . . . . . . . . .

+

Figuw 8.- Pressures on top surface of the mcdified-xedge w i n g model; a, 14.5O. f

. . . . .

W .!=

. . .

,

45

I

F/agged sym&o/s and dashed curves indicaie /ow# swfuce pressures.

-5.0

-4.0 Q

e \ s -3.0 Q

8 -2.0

t 8 -1.0 d

'4 0

0

/.Oo 1 I I .2 2 ' 6 .8 I

Chord dation, x/c

(U) Q, doo

Figure 9.- Pressure distrbufion along cbord for various angles of affack of modified-wedge wing-body model.

-5.0

Q -4.0

%' 0

t, .? -3.0

5 0 Q -2.0

i i .90 c .90

.

Figure 9- Continued.

47

-5.0

-4.0

%* C -? -3.0 % 9, 0 u -2.0

t 3 -/.o a

h 0

1. Oo t L .2 .4 I .6 I .8 I i o = q z g 7 Chord sfoofion, x/c

Figure 9.- Confhued.

48 ~ C A RM NO. ~ 9 ~ 1 7

0 , .60

h 5 c - -f.O %

0

LOo 1 I 1 1 .P A .6 .8

Chord sfation, x/c

1

Figure 9- Confinued

mcx RM NO. ~ 9 ~ 1 7

-5.01 r t t t d

LOo L .2 .4 .6 .8 I I I I

i o =qgzp- Chord sfafhn, x/c

(e) a, 14.50,

Figure 9.- Confinued.

49

1. oo 1 I I I .2 4 .6 .8 I

Chord station, x/c

( f ) a, 20.73

Figure 9.- Continued

-6.0

-5.0

-4.0

e -2.0 .60 a cr, cr, g -1.0 .45

-4.0

C 9 .$ -3.0 k 8

e -2.0 cr, a cr, g -1.0

i \

EACA RM HO. ~ 9 ~ 1 7

>

-5.01 r f

-P"-4"""'

Chord stotion, x/c

(h) a, 3/.0°.

c

Figure 9. - Condude d.

I

. . .- . . . . . . . . . . . . . . . . .

f I I

. . . .

I I

P F

Figure l0. - Presaures on top surface of mcdlfledvedge wing body model.; aI 14.5°. 8

HACA RM NO. ~ 9 ~ 1 7 55

Flagged symbo/s ond dashed curves indicate lower swfoce pressures.

. /.oo t .$ R L .6 I .8 I -I

1.0 -qg7 Chord station, x/c

ka.) a, OJ0.

Figure 1L- Pressure tfkfr1bution dong chord for vurhus angles of attrrck of NACA 65-series wing-body mode/ controls neutrd

RACA BH NO. ~91317

Chord station, x&

(b) a, 4 . 2 O .

57

t

-50 c f

-2.0

o$""Ix,',-iil"

1. Oo .2 .4 .6 .8 i o Chord sfaflon, x/c

r

NACA RM NO. ~91317

r f

(d) a, 1 2 . 5 O .

Figure 11.- Conthed.

t

59

60

.

NACA RM NO. ~ 9 ~ 1 7

-5.0 I

/. 0: I .2 A' .8 LO T g g 7 L 1.. . .-I

.6 Chord station, x0c

( f ) a, 20.8'.

Figure /I.- Continued.

Figure /L - Conthued.

62 mACA RM No. A917

Chord sfation, x/c

(h) 4, 290°.

Figure / L - Continued.

mm RM NO. ~ 9 ~ 1 7

.

1. oo 1 I I I .2 .4 .6 -8 $0 I

Chord slution, x/c

a, 33.0" -

Figure / L - Concluded

64 NACA RM No. A 9 1 7

Flagged symbo/s and dashed curves ihdicuie lower wface pressures.

LOo t '.I .P .4 .6 .8

I I I

65

.

""I c ,

r t /

/.oo I L I I L 1

.2 .4 .6 .8 1.0

(&) a, 16.5s

Figure E.- Continued

,

66 HACA FM NO. ~ 9 ~ 1 7

Chord staf ion, x/c

(e) a, 248*.

figure 12.- Continued.

Chord station, x/c

(d) a, 32.9..

Figure 12.- .Concluded.

68

1.6

l.4

1.2

c,- /. 0

0 *? 2 2.8 0 0 c,

.6 3 C 0 2 .4 8 .2

0

-. 2

HACA RM No. A 9 1 7 .

Ang/e of attack, a, deg

(01 Modified- wedge wing model.

Figure 13.- Secfion lift curwes for several spanwise wing sfotiom.

c

HACA RM NO. ~ 9 ~ 1 7

Angle of utfaack, a, deg

(b) Modifled-wedge wing-body model

~ C A EM NO. ~ 9 ~ 1 7

Angle of attack, (I, deg

(c) NACA 65-series wing-hdy model, controls neufrd.

Figure 13.- Continued.

.

.

.

,

Ang/e of offuck, a, deg

(dl NACA 65-series wing-body mode4 frdfing-edge contrzds up P.

Ffgure 13.- Conc/uded

c

. .. .. . ". . . . . . - . . . . . . . . . . . . . . . .

. . . . . . . . . . ,

I m

. . . .

I

. . . . . . . . . . . . . . . . . . . .

’’ I b 8

‘ b

(a) Modified-wedge wing model.

Semiwan station

+“”d

6”- 6 0 ”“ .75 GL-“” .go

, Angfe of attack, a, deg

Figure 15.- Center of pressure for severai spanwise wing stations. w?v= (b) bfodified-wedge wing-body model.

. . . . . .

. . . . . . .

. . . . . . . . . . . . . . . . . . . .

Angle of attack, Q, deg (c). NACA 65-serios . wbg-booj mode/, controls nmtraf.;

Sed- datbn

""-4

--,SO " - .75 "".90

Angle of attack, a, deg (dl NACA 65-series wk?pbody mod& fraiffng-edge c0n:rofs up H)p

F ~ U ~ V d- Concluded.

. . . . . . . . . . . . . . . .

a . I *

. . . . . . . . . . .

HACA RM No. -17

Percenf semiwan

(a) Modified-wedge wing mode/. -

Figure E- Span lood d/sfr/bufion for several angles of atfack.

75

UCA RM NO. ~ 9 ~ 1 7

L

k e n # serm'spon

(b/ Modified- wedge wing-body model =i@%7

Figure 16- Confhued

NACA RM No. AB17 77

-, Y

0 IO 20 30 40 50 60 70 80 90 100 . .

(c) NACA 65-series wing-body mode/, confrok neufrd.

figure 16- Continued

e .8

\ B .4

Percent semlspan

Percent semispn

(d) NACA 65-series wing-body mode/, froihnq-edge confro/. -537

up /Of

Figure 16.- Conchded.