\0 .--' H

I" o ~ o

:z;

I ..

RM No . A7L16

NACA

RESEARCH MEMORANDUM A WIND-TUNNEL INVESTIGATION AT LOW SPEED

OF VARIOUS LATERAL CONTROLS ON

A 450 SWEPr-BACK WING

By Edward J. Hopki ns

Ames Aeronautical Laboratory Moffett Field, Calif.

NATIONAL ADVISORY COMMITTEE FOR AERONAUTICS

WASHINGTON April 9, 1948

https://ntrs.nasa.gov/search.jsp?R=19930085888 2020-04-16T03:37:01+00:00Z

NACA RM No. A7L16

NATIONAL ADVISORY COMMITTEE FOR AERONAUTICS

RESEARCH MEMORANDUM

A WIND-TUNNEL INVESTIGATION AT LOW SPEED

OF VARIOUS LATERAL CONTROLS ON

A 450 SWEPT- BACK WING

By Edward J . Hopkins

SUMMARY

A wind- tunnel investiRation was conducted at low speed of chord-extension controls , conventional a ilerons, and spoilers on a 450 swept-back wing of aspect ratio 4.5 and of taper ratio 0.5. Measurements were made of the lift , drag, pi tching moments, rolling moments, and the rates of roll produced by the various controls. The effect on the pitching-moment characteristics of "fences" on the upper surface of the a irfoil parallel t o the air stream was also determined.

The results indicate that the conventional ailerons were more effective in producing rolling moments than either the chord­extension controls or the spoilers . Maximum effectiveness of the spoilers was obtained with the spoilers perpendicular to the air stream.

The fences parallel to the air stream extended the linear variation of pitching-moment coefficient with lift coefficient from a lif t coefficient of 0.45 to 0.80, but did not affect the longi­tudinal instability at higher lift coefficients.

INTRODUCTION

One of the major problems involved in the use of swept wings i s the provi sion of adequate lateral control, especially at high lift coefficients. The experimental data of reference 1 indicate that the effectiveness of conventional ailerons in producing rolling moments is considerably reduced by incorporating sweepback in the wing plan form. Therefore, the effectiveness of a different type of lateral-control device, consisting of the rearward extensi on of the wing chord, was investigated on a 450 swept­back semispan wing. Spoilers were also tested on the same wing t o

2 NACA RM No. A 7L16

de termine the effect of spoiler location relative to t he a ir stream on their effectiveness for providing rolli ng moments .

As simple sweep theory indicates tha t the damping in roll is reduced by sweep and, since it wa s believed that the chord-extension controls would increase this damping, comparative mea surements were made of the rates of roll produced by t hese controls and by conven­tional ailerons on a full-apan model of the same pl an form as the semispah wing. In order to obta in a more comprehensi ve comparison of the effectiveness of the controls, the r olling moments were also measured at various angl e s of yaw.

I n an attempt to control the out board spanwi se flow i n the boundary layer and thereby delay separation a t the wing tip, the effect of fences alined in the free-£tream direction on the upper surface of the semispan model was determined.

SYMBOLS, COEFFICIENTS, AND CORRECTIONS

The data are presented in the form of standard NACA coeffi­cients and symbols. All forces and moments are presented about the stability axes with their origin on the root chord at t he same fore and aft location as a point at 25 percent of t he mean aerodynamic chord of the plain wing.

( twice lift OfqSSemiSpan model) CL lift coefficient \

CD (twice drag 0qfSSemiSpan mOdel) drag coefficient \

~Cn increment of drag coefficient caused by the extension of the controls

C2 rolling-moment coefficient ( rOlli~ moment)

pb

2V

q

Pit~ing-moment coefficient . twice pitching moment of

\. q~

helix angle of roll, radians

semispan model)

angle of attack of the wing chor d line, degrees

angle of yaw, degrees

free-stream dynamic pressure (~p~), pounds per s quare foot

•

NACA RM No. A7Ll6

S

b

p

V

p

c

A

Subscripts

L

R

u

full-span wing area, square feet

full wing span, feet

rate of roll, radians per second

airspeed, feet per second

air density, slugs per oubic foot

wing chord, feet

wing mean aerodynamic chord, feet

aspect ratio (b;)

eff(ctive Reynolds numb)er VC x(turbulence factor)

kinematic viscosity where the following turbulence factors, as determined from sphere tests, were used:

(a) 2.34 - turbulence net inl

(b) 1.0 - turbulenoe net out

conventional aileron deflection measured in a plane perpendicular to the hinge line, degrees

left aileron

right aileron

uncorrected values

The data obtained from tests of the semispan model were corrected for the effects of the tunnel walls by the method of reference 2, which does not · consider corrections for a swept-back wing. In order to facilitate the reduction of the data, the corrections were assumed to be identical to those for a model of unswept plan form of the same aspect ratiO, span, and taper ratio. The corrections applied to the data obtained from tests of the semispan model are as follows:

lIn order to increase the effective Reynolds number for the full­span model, a turbulence net was installed in the wind tunnel.

3

4 NACA RM No. A 7L16

fu,l (jet-boundary correction) = 0 . 652 CLu

~ (streamline-curvature correction) = 0.0646 CLu

DeD = 0.0133 CLu2

UCm = 0.00188 CLu

DeL = -0.004 CLu

A previous check of the corrections for a similar model of a swept­back wing indicated sweepback to have a negligible effect on the magnitude of the corrections.

The drag coefficients presented for the semispan model are not the absolute values as the drag of the reflection turntable is included. However, the incremental drag coefficients caused by the controls are believed to be approximately correct.

The rolling moments produced by the chord-extension controls on the semispan model were not corTected for reflected load effects, as it was desired to obtain only the com.parative effectiveness of the various controls.

No corrections have ~een applied to the data obtained from tests of t he full-apan model because of the small size of the model rela­tive to the size of the test section of the wind tunnel.

MODElS AND APPARA.TUS

A semispan model and a full-apan model were used for the inves­tigation in the Ames 7- by 10-foot wind tunnel. The wing panels of the full-span model were three-elghts of the scale of the semispan model. Both models had NACA 64A210 (a:0.8) airfoil sections2 parallel to the plane of symmetry, the 25-percent chord line swept back 450

,

an aspect ratio of 4.5, and a taper ratio of 0.5 . A summary of the geometric charac t eristics of the models is presented in table I.

The semispan model was mounted on a turntable that was flush \fith the tunnel floor which simulated the plane of symmetry (fig. I). The forces and moments a cting on the model were measured by the normal six-component vrind-tunnel balance system.

The full-span model was mounted on a sting support as shown in figure 2. Rolling-moments of the full-apan model were measured,

2The symbol A represents an airfoil sec tion with straight sides near the trailing edge.

NACA RM No. A7Ll6 5

exclusive of the forces on the support, by means of a cantilever electrical strain gage. No allowance was made for interference effects of the sting on the model. The model was allowed to rotate unrestrained about an axis parallel to the air stream. In this manner rates of roll produced by the various controls were deter­mined by timing a given number of complete revolutions of the model. In order t o obtain steady rates of roll, the model was statically balanced about the axis of rotation at each angle of attack by the addition of lead weights at the nose of the model. The angle of attack of the model was changed by rotation about a lateral axis located at 19.8 percent of the mean aerodynamic chord.

The dimensional data for the chord-extension controls tested on the semispan model are presented in table II and figure 3. These controls (made of O.09l-inch sheet steel) were attached to the upper surface of the wing which was recessed to provide a smooth contour. The controls grojected along the airfoil mean camber line, giving an angle of 2 55' between the wing chord plane and the controls. Controls A and B were assumed t o be extended by rotation about a point on the wing trailing edge (fig. 3) . No consideration was given to the fact that control A could not be retracted within the wing plan form as it was desired to determine the maximum effective­ness obtainable with such a control. Control B was similar to control A except that control B could be retracted within the wing plan form. Control C was merely a constant-chord extension and control D was similar t o control A except that control D covered the entire wing span.

Spoilers were tested on the upper surface of the semispan model in the positions shown in figure 4. The spoilers (made of O.05l-inch aluminum alloy) were mounted perpendicular to the wing chord plane and extended 1 inch above the wing surface. Four fences (1 inch high) alined parallel to the plane of symmetry were also tested on the upper surface of the semispan model (also shown in fig. 4).

A chord-extension control ~ontrol Al ) similar to control A was tested on the full-span model. Instead of extending the control along the airfoil mean camber line as on the semispan model, it was extended tangent to the airfoil upper surface. This resulted in an angle of ~181 between the control and the wing chord plane. A plain, unsealed aileron of 20-percent chord and 50-percent span was also investigated on the full-epan model (fig. 2).

6 NACA RM No. A7Ll6

RESULTS AND DISCUSSION

Chord-Extension Controls on the Semispan Model

The results obtained from the tests of the chord-extension controls on the semispan model are presented in figures 5 and 6. As shown by these data, the chord-extension controls are characterized at low angles of attack by low rolling effectiveness which is con­s1derably improved as the angle of attack is increased . The maximum lift increments and the lift-curve slope increase caused by these controls are approximately proportional to the increase in wing area as illustrated by the following table which compares the percentage increase in maximum lift coefficient and the percentage increase in lift-curve slope with the percentage increase in area:

IncreasA in Increase in Increase in Control maximum 11ft lift-curve area (%) coefficient (%) slope (%)

A (160) 12-3 12.7 9

B 10.4 10.4 7 c 7.7 8.2 7 D 24.5 30.9 20

As would be expected, the chord-extension controls caused an increase in the longitudinal stability of the model.

A Chord-Extension Control and a Conventional Aileron on the Full-Span Model

In order to obtain a comparison of the performance of the chord­extension control with that of a conventional aileron, tests were conducted upon the full-epan model with a conventional aileron of 20-percen~ chord and 50-percent span and with a chord-extension control of 50-percent semispan similar to control A extended 160

•

Measurements were made of the rolling moments at various angles of yaw throughout the angle-of-attack range and the rates of roll with the model rotating unrestrained. It was found that at small angles of attack control A, as tested on the full-span model, was incapable of producing steady rates of roll with the model unre­strained. Therefore, the control was extended tangent to the upper surface of the airfoil (referred to as control Al ), thereby increasing the control deflection relative to the wing chord plane

-- ---- -- -- --- -- --. -- -- ~--- - ~------- ----

NACA RM No. A7Ll6 7

from 20 55' to ~lS'. All the results presented for the chord­extension control on the full-epan model are for a deflect ion of if>lS'.

The rolling-moment coefficients and tee wing-tip helix angles measured with the chord-extension control A1 and with the conven­tional ailerons on the full-epan model are presente d in figure 7. As shown by these data, the ailerons were considerably more effec­tive than the chord-extension controls. Also the effectivene s s of the chord-extension control was seriously reduced as the angle of yaw was increased. The wing-tip helix angles which the conventional ailerons are capable of producing were also estimated from the measured rolling-moment coefficients, using the damping in roll of reference 3 reduced by the cosine of the sweepback angle. The results, shown in figure 7, indicate that simple sweep theory gives a good first approximation of the damping in roll at small angles of attack, but at higher angles of attack where the wing tip was stalled the conse~uent reduction in the damping in roll should be considered.

By means of a turbulence net and by varying the dynamic pre ssure from 5 to 50 pounds per s~uare foot, the Reynolds number of the full-epan model was varied from 0.27 X 106 to 2.08 X 10 6

•

The effect of this variation of Reynolds number on the wing-tip helix angles produced by control A1 is shown in figure S.

Spoilers on the Semispan Model

The data from the tests of the spoilers of 50-percent span on the semispan model are presented in figure 9. As shown by these data, the largest rolling-moment coefficients were measured for the spoiler perpendicular to the air stream. However, at high angles of attack, there was either a complete reversal in spoiler effectiv­ness or the effectiveness was seriously reduced, depending on the spoiler location. The spoilers were considerably more effective at low angles of attack, but were less effective at high angles of attack than the chord-extension controls. The conventional ailerons tested on the full-span model were more effective than any of the other controls tested.

Fences on the Semispan Model

During the course of the investigation fences were tested on the upper surface of the semispan model in an effort to extend the linearity of the pitching-moment characteristics to higher lift coefficients. As shown by the data in figure 10, the fences did

-~ ----

8 NACA RM No. A7L16

extend the linear variation of pitching-moment coefficient with lift coefficient from a lift coefficient of 0.45 to O.So. However, the longitudinal s tability at higher lift coefficients was not improved.

CONCLUSIONS

The results of the wind-tunnel investigation of several lateral­control devices and fences on a 450 swept-back wing of aspect ratio 4.5 and of taper ratio 0.5 indicate:

1. The conventional ailerons were more effective in producing rolling moments than either the chord-extension controls or the spoilers throughout the useful angle-of-attack range.

2. The maximum eff ectiveness of the spoilers in producing r olling moments was obtained with the spoilers perpendicular to the air stream.

3. Simple sweep theory provided a good first approximation of the damping in r oll at small angles of attack, but was unsatisfactory f or predicting the damping in roll at higher angles of attack when the wing was partiall y stalled.

4. The fences parallel t o the air stream increased the maximum lift coefficient for a linear variation of pitching-moment coefficient with lift coefficient fr om 0.45 to 0.80, but caused no improvement in the longitudinal stability at higher lift coefficients.

Ames Aeronautical Laboratory, National Advisory Committee f or Aeronautics,

Moffett Field, Calif.

REFERENCES

1. Letko, William, and Goodman, Alex: Preliminary Wind-Tunnel Investigation at Low Speed of Stability and Control Character­istics of Swept-Back Wings. NACA TN No. 1046, 1946.

2. Swanson, Robert S., and Toll, Thomas A.: Jet-Boundary Correc­t ions for Reflection-Plane Models in Rectangular Wind Tunnels. NACA ARB No. 3E22, 1943.

3 . Swanson, Robert S., and Priddy, E. LaVerne: Lifting-Surface­Theory Value s of the Damping in Roll and the Parameter Used in Estimating Aileron Stick Forces. NACA ARB No. L5F23 , 1945.

- - - -- - - - -- - - - ------

NACA RM No. A7L16 9

TABLE 1.- GEOMETRY OF MODELS

Dimension Semispan wing Full-span wing

Aspect ratio 4.5 4.5

Taper ratio .5 .5

Sweepback of 0.25-chord 45 45 line, degrees

Airfoil section NACA NACA 64A210 (a=0.8) 64A210 (a=0. 8 )

Span, feet 4 3

Area, sCluare feet 7.097 1.996

Mean aerodynamic chord, 1.844 .6g2 feet

Root chord, feet 2·371 . 889

Tip chord, feet I 1.185 .444

10 NACA RM No. A7L16

TABLE 11.- DIMENSIONAL DATA FaR CH<JRD..-EXTENSION CONTROlS ON THE SEMISPAN MODEL

Angle between wing

Control Control span trailing edge and Area control trailing edge (sq ft) Wing semispan in plan view (deg)

A 0.50 16 0.87

A .50 11 .593

A .50 6 .326

B .44 16 .74

C .50 0 (constant-chord .55 control)

D 1.00 8 1.74

- - - - -- - - --- - _._- - - --- '

NACA RM No. A7L16 11

(a) Plain wing

A -lOSll

~) Control A extended Figure 1.- Semispan model mounted in the Ames 7- by 10-foot wind tunnel.

I-I

- I

I

I

I

I

L-

(a) Conventional ailerons deflected ~) Control A~ extended

Figure 2.- Full-span model mounted on the sting in the Ames 7- by 10-foot wind tunnel.

s; (")

:x>

~ ~ o

:x> ~ f-' 0\

f-' UJ

-- --- --- --- --- -- -- --- --- ---- -- -- ----- ---- -- _._- -----

I I I

I . I

- I

I

- I

I . I

I

I

I

I I I

- I

I I I

-- -- ------ ---- -- - - - --- -~j

[- - ----~--=----~---~--

NAeA RM No . A 7L16 15

---,--______ ~_~/4.222 Retracted position A ~

Control A Control B

Section A-A I no scale

Control C Control 0

All dimensions in inches ~

A'qlJre 3.- Chord-extension controls tested on semispan model.

-----~~ - --~~. - - - - ---~-

-- - - -- -- -- ---- -- -- -- -- - -

Spoiler A ---~_ 8 __

C --.J

C) C)

~ t\j

t

---1

Fences

Note(I)AI/ dimensions in inches

C) C) ~ C,\J

(2) Fences and spoilers extended I inch

above the wing surface,

~

A'glJre 4.- Spoilers and fences tested on the semispan model.

~ 0'\

~ (")

~

~ !2: o . ~

~ 0'\

NACA RM No . A7Ll 6 l7

12

(.:)" /.0

..... ~ .8 c::

. ~ ~

~ .6 ...... Cb ~ ~ .4 ~ ..... ......, .2

o

-.2 ~~ ·~~.08 ..... c:: ::::Cb ~ ~.04 ~ '-

~

~.~ 0 '-

r-

~ ~

I~ ~

~ -4 0

n. ..... "'l:dl /

L ~/ ('---<. >-~ '>-< r

~ 0 0 y--r-': .-' ;.--

:rt: ::Q: tn. D: rH= ~ ~ -K. [{{

~ ~ tg. 0-v

B:: ~ I.:J

~ fY ~ ~ . ~

16 ....

A 11""" 6'~ ' 0

,;,;; W o' ~ <> t>.

~~ 0

'r ~

~

It .... ..-( }--i 0 y ~>- y<-

~

',..--1 ~ :{"

4 8 12 16 20 24 28

Angle of attock, a., deg.

.Q]

~ . ~ L'o.

~ ~ """ '-' ~ '\ <{ I

t) ~ ~)'i r £. (!

c: I~ ( ~

ff1 ~~ ,~

C ~ J, G

./ 0 -./ -:2

Pitching- momen coefficien t, em

Figure 5.- Effect of chord-extension control A on aerodynamic characteristics of the

semispan model . Ref{ = /.881110".

t

18 NACA RM No . A7Ll6

:r -D- -n- -D-N:

cY A;J Ll

/ lD , J',

P" / ~ " It--~ ~

n"'" P' /-; ~ r~ I ~

o ,~ Y. . ---" .x:: ~ ~ .

...{ 1::: I""" .,E::;;l f"""

..g- Ig, H- ..,.., f.o- r-' ~ -=--

1.4

1;:( A J:~ [;J I"{ ~ ~ p= ~ ~ )Q. ~

12

J /-~ P'"' 1.0-fo-' ~ I~ \i r\ 1: 1\ ~

I)-' ~ ~ rv '-' ~ UI

1.0 ~ ...

~ ~ lP' ~ ~ / kI P

't '" Y

..... ' .8 t::

Ib "

I~ P' ~ J '7'

[Jd ~ ~ ~

(.) "

.6 ..... ..... Ib

~Ff

###u ( f-'

A l' --Av ' ....., ' , ,

c ' 16° ' 16° ,/ /:,' 80

iP If (> ~ ~ 0 0

!I' Control A Control 8 Control C Control 0 Plain wing

[1'

() (.) .4

.::: " ...J .2

0

l:&' Ie

..... ~ ./ 0 -./ -:2 V fl

Pitchinq- moment ...I . ~ iJ--1;r coefficient, em /

if

l( 5:d d ·H H )--;-. ---1.

lG ./ f~ ~. ;y- , . ,

-.2

~ (,,)~/2 Ib "::J

"

.~ ~'.08 "'Ib ~E: ~ ~.04

(.)

g- .~ 0 '-

~ -4 0 4 8 12 16 20 24 28

Angle of ottack, a., deg.

Figure 6.- Effect of chord-extension controls on aerodynamic characteristics of the semis pan

model. Reff" /.88 x /0 6.

-- ---- ~.-- - . ~-------

NACA RM No. AfL16

~

" c:::

-:05

-:03

Cb -:02 e: <;:)

e: ~ -:01 c::: ..... ..... ..... ~ 0

. 3

["'---- .le-...-/ l7--:: t--... ./ [ ~ f.---tiJ ~.{'

1---+---+-+--+----+-+----+. -)

<

Q.------,....,

v v ./"() V

----- Computed from rolling-moment coef­ficients produced by -I--

o Control A, ailerons .

d AllBrOnS 8R= 19. 6_° --+-+---+--..)J--+----i

q =-19.6 i ~ k----+-V---+--+_,

~ - ~-- ==- ~--=-::_::::t---+-/---:;.O.o-+----i -v- 1/ /~-

o 4 8 12 16 20 24 A ngle of attock, a, deg.

Figure 7 - Comparison of rolling-moment coefficients

and wing- tip helix angles produced by chord­

extension control AI and conventional ailerons

on the full-span model. Reff == 1.88 K 10.-

19

--------- ---- ------- _._- - ----

.20 r.& I 0 0° ""

.16 ~I~ ~~

Cb' .12 ........ ~

~

~ A 4° 1-<1--~ {' .:1 0 8°

[l:. ~ ~' rt;> v ¢ 12° I .; a;; 1 i

\I .... V -- - - -

16' I' ~7 r> v

""" 20° I'v ~ if : 25° -.;j

'{ ./ .L ..d I

1\1 v

t:: tJ

.~ .OB

~

0 FlaggtJd symbols: TurbultJnc8 net in

1'0 I'-.. Cb

..(;::

.~ .04

...... I

0 r-o.. ~ t-o K L /. ./ J -0 -v- -0 ~./ V 1f rv: v .-<" '""

;.., ~

r;}., t:--:o L.-O- ~

A L:J

I ~ A A ,;' L A /., ./ -0-~ ef ·= :::::i'= =:F.- LO ~ L \

"-\. ' -:1. ~ v l:..J I ~

~ .t:: 0 ~ ~~

.2 .4 .6 .B lO l2 1.4 1.6 I.B 2.0 2.2

Effective Reynolds Number x 10- 6

Figure 8.- Effect of Reynolds number on values of wing-tip helix angle

produced by chord-extension control AI on the full-span model.

------------~~--

f\) o

~ ~ ~ !2\ o . !l:> ~ f-' 0\

It

•

NACA RM No. A7L16 21

12

~" 1.0

",,'

.8 c:: .~ ~ ;;:: .6 ..... Cb <::) ~ .4

.::: ..... -...J .2

~v J

0

-.2 ""~08 c:: ~ . . ~

~ .1 0 -.1 t--+-+--t--t---+---+--+--+-+--+-----i-t--+-+-+ . ~ Pit ching - mom e n t

~ """ ;;:: c:: 04 ..... <b. Cb e: <::) <b

coefficient, em

~ \", 0 ~

g-.~ \",

-.04 ~ -4 0 8 12 16 20 24 28 4

Angle of ottock, ~, deg.

Figure 9 .- Effect of spoilers on the aerodynamic characteristi cs of the semispan

model. Reff =/.88 x IO~

22

12

~ .. 1.0

~' c::: .8 Cb

' .... ~

~ 6 ..... . Cb (;)

~ .4 ~ ' .... ..... . 2

o

~ .. 1.0

~' .8 c:::

.~ ~ .....

.6 ..... ..... Cb (;) ~ .4 ~ ..... ' .... ..... .2

0

-.2

-4

NACA RM No. A7L16

I?

1

Angle of attack, IX, deg.

8 12 o 4 16 20 24 28 .08 .04 0 ;04 ~08 Pitching- moment

coeffic ient C. , 'm ,

..(). -£ -u

fo- -to: ~ - i.J..l'" ~

-----

....[1J

r ~p

o - Ploin WilJ~ ~

t~ c- FtJnces~ /

s-J 9l // ~

(' // '/ '/ '//

~ T J J J J

o .04 .08.12 .16 .20 .24 .28 . 32 .36 .40 .44 .48

Drag coefficient, Co

•

•

FiglJre 10.- Effect of IJPper-slJrface fences on the aerodynamic char(1cteristics of the

semispan model . Reff = 1.88 x 10~

_. _. _______ ., __ J