UNCLASSIFIED

AD 2 6 8 138

ARMED SERVICES TECHNICAL INFORMATION AGENCY

ARLINGTON HALL STATION
ARLINGTON 12, VIRGINIA

UNCLAS8''I F.I.ED
NOTICE: When government or other drawings, speci-
fications or other data are used for any purpose
other than in connection with a definitely related
government procurement operation, the U. S.
Government thereby incurs no responsibility, nor any
obligation whatsoever; and the fact that the Govern-
ment may have formulated, furnished, or in any way
supplied the said drawings, specifications, or other
data is not to be regarded by implication or other-
wise as in any manner licensing the holder or any
other person or corporation, or conveying any rights
or permission to manufacture, use or sell any
patented invention that may in any way be related
thereto.
 NASA TN D-992

<~
U00
z

TECHNICAL NOTE
D-992

'-. STUDY OF FLOW OVER OSCILLATING AIRFOIL MODELS

AT A MACH NUMBER OF 7.0 IN HELIUM

By Ali Arman

Langley Research Center

Langley Air Force Base, Va.

ASTIA

CN

NATIONAL AERONAUTICS AND SPACE ADMINISTRATION

WASHINGTON December 1961
1E

NATIONAL AERONAUTICS AND SPACE ALMINISTRATION

TECHNICAL NOTE D-992

STUDY OF FLOW OVER OSCILLATING AIRFOIL MODELS

AT A MACH NUMBER OF 7.0 IN HELIUM

By Ali Arman

SUMMARY

A wind-tunnel study of unsteady flow at a Mach number of 7 in

helium has been conducted on several sting-mounted wedge, double-wedge,
and flat-plate airfoil models with three different leading-edge radii.
The data were obtained by taking high-speed schlieren motion pictures
of the decaying motion of the model as it was released from an initial
deflection.

The shock-wave position observed on the sharp-leading-edge models

during the oscillation was compared with that obtained by use of unsteady-
flow theory as well as steady-state theory. Comparison of theoretical
results indicated that no unsteady-flow effects exist over the range of
reduced frequencies k, 0.007 -5k 15 0.030, studied experimentally. The
experimental results confirmed this finding as no unsteady-flow Pffects
were detected in this reduced-frequency range. Comparison of shock-wave
positions measured for the blunt models with those calculated by steady-
state methods indicated fair agreement.

INTRODUCTION

In spite of the large quantity of steady-hypersonic-flow data which

have been acquired in recent years, almost no information is available
on unsteady- or oscillatory-hypersonic-flow phenomena. Unsteady-flow
effects are of concern with regard to flutter and dynamic stability at
hypersonic speeds. Most hypersonic unsteady-flow data have been obtained
indirectly through flutter tests (for example, refs. 1 to 4). Some
authors, such as those of references 5 to 8, have attempted to treat the
problem analytically but their works have lacked experimental verifica-
tion. In one of these papers (ref. 7) a method was developed by which
the shock-wave position of an oscillating, sharp-leading-edge airfoil
could be computed. The present study was designed to give experimental
verification of this theory and to determine the applicability of quasi-
steady flow theory. The effects of oscillations on the shock-wave posi-
tion of a blunt airfoil were also investigated and compared with those
obtained by use of the steady-state, zero-angle-of-attack method of
reference 9.
2

Schlieren motion pictures were made of a series of simple models

oscillating in helium at a Mach number of 7. The models studied were
wedge and double-wedge sharp-leading-edge airfoils as well as airfoils
with cylindrical leading edges of various radii. The reduced-frequency
k range covered by the tests was 0.007 5 k '-0.030.

SYMBOLS

b semichord, ft L
8
8
diameter of cylindrical-leading-edge airfoil, in.
D
9
x y nondimensional rectangular coordinates attached to center of
5D5 cylindrical leading edge of airfoil

k reduced frequency, bw/V

q dynamic pressure, lb/sq ft

T period of one cycle of oscillation, 2A/W, sec

V free-stream velocity, ft/sec

a deflection angle, deg or radians

a semiwedge angle, deg

8 angle between shock wave and free-stream velocity vector,

deg or radians

p density, slugs/cu ft

T thickness, percent chord

a) circular frequency, radians/sec

t time, sec

f model frequency, cps

Subscripts:

r theory of reference 7

8 steady-state theory
 3

APPARATUS AND PROCEDURE

Wind Tunnel

This investigation was conducted in the 8-inch-diameter test section

of the Langley hypersonic aeroelasticity tunnel which is an ambient-
temperature tunnel using helium as a test medium. This tunnel is of the
blowdown type; the test section used had a Mach number of 7 and a maxi-
mum dynamic-pressure capability of 5,000 pounds per square foot. A more
detailed description of this facility is given in reference 1.

)Schlieren System

A conventional single-pass schlieren system was used, which con-

sisted of two 12-inch-diameter, 6-foot-focal-length mirrors. The light
source was a mercury vapor lamp contained in an air-cooled case. A
high-speed camera recorded the motion at approximately 3,000 frames per
second. The adaptation of the schlieren system for this particular
study is discussed subsequently.

Models

The models used for this study had wedge, double-wedge, and flat-
plate airfoil sections, sketches of which are shown in figure 1. All
models had an insert constructed of 1/8-inch-thick 17-7 PH stainless
steel with an integral shaft for mounting to the sting support. Some
of the models were lightened by drilling holes in the steel insert. A
balsa overlay was used to produce the basic shape of the models. For
the thicker flat-plate models, different size dowels were used to give
the desired leading edge. The wedge and double-wedge airfoils had a
fiber-glass coating over the balsa. The pertinent dimensions of all
the models are listed in table I and a photograph of typical models is
presented as figure 2.

A sketch of the model sting support is shown in figure 3. This

support was constructed of stainless steel and was rigidly attached to
the tunnel wall so that the model was at zero angle of attack. The
natural frequency of the model mounted on the support could be varied
by adjusting the position on the model's shaft at which it was clamped
to the support. A photograph of the wedge model mounted on the sting
support is shown as figure 4.

A small notched plug, 1/4 inch in diameter, was threaded to the

steel core of each model as shown in figures 2, 3, and 4. A wire,
hooked over the notch in the plug, was used to give the model an
4

initial deflection. For the purpose of this study it is assumed that

the plug located on the upper surface of the model does not disturb the
flow on the lower side.

Tests

The procedure used for the tests was to quickly bring the tunnel
test-section conditions to preselected values and then release the
models and allow them to oscillate through a decaying motion at nearly
constant test-section conditions. Oscillations were introduced by
attaching a wire to the notched plug on the models and to the outside
of the tunnel. The models were deflected to a predetermined angle by
putting tension on the ire which was then cut to release the models
at the proper time during each test. The wire hookup and the release
of the models are shown in figure 5. It should be noted that the wire
is carried downstream as soon as it is cut to release the models. High-
speed schlieren motion pictures were taken of each test at approximately
3,000 frames per second.

Data Reduction

Although the schlieren system used in these tests is a conventional

single-pass system, certain modifications were made for this particular
study. A horizontal knife edge was used in order to obtain better sensi-
tivity to the nearly horizontal shock waves. The characteristics of the
schlieren system with the horizontal knife edge are such that the shock
wave on the upper surface is white and the boundary layer is black,
whereas on the bottom surface the boundary layer is white and the shock
wave is black. In conventional systems the model appears as a black
silhouette. In order to distinguish the model outline from the boundary
layer, the upper half of the model side was painted white and the lower
half was painted black (fig. 4); a spotlight was directed at the painted
side of the model nearest the camera. Typical enlargements from the
1 6 -millimeter motion pictures taken at intervals of about 0.2 of a period
for a complete cycle of oscillation are shown in figure 5 for each of the
three types of models. The effect of the front lighting can be seen by
comparing figures 5(a) and 5(b) in which the models were illuminated with
figure 5(c) in which the front lighting was not used.

From the 16 -millimeter schlieren motion pictures 8-inch by 10-inch

prints were made of the pertinent frames. The slope of the shock wave
was obtained as shown in the following sketch; the pertinent angles
were -'. ;ured directly from the photographs:
 V

L As indicated in the sketch, the finite leading-edge radius and boundary-

3 layer effects produce curvature of the shock at the leading edge. The
5shock angle measured was the angle between the downstream asymptote of
the shock wave and the free-stream velocity vector.

RESULTS AND DISCUSSION

The test conditions including test-section density, velocity,

dynamic pressure, model frequency, and the reduced frequency for each
test are presented in table II.

The method of reference 7 gives the perturbation of the shock wave

for small angles of attack about some mean angle. The theory was devel-
oped for simple harmonic motion and it was assumed that the reduced fre-
quency was much less than 1. The method of reference 7 predicts a curved
shock wave for the oscillating wedge, whereas steady-state theory indi-
cates a straight line for the shock wave. However, at the frequencies of
the present test the curvature terms in the theory of reference 7 are
negligible. The theory also predicts the phase-angle lead or lag between
the model oscillation and shock wave. The theoretical phase angles were
very small for these tests and the examination of the experimental data
has substantiated this result.

The actual response of the models was that of a decaying harmonic

motion, rather than pure harmonic, with damping on the order of 1 per-
cent critical. The small amount of damping allows the harmonic-motion
theory to be applied over short periods of time.

A comparison of shock-wave angles calculated by steady-state theory

with those calculated by the method of reference 7 is shown in figure 6
for the wedge model at M = 7.0 for flow deflection angles ranging from
3.50 to 11.50 and k = 0.028. The steady-state equations used in this
comparison were obtained from reference 10. As can be seen, the two
results are in close agreement.

The effect of frequency on the shock-wave angle is shown in fig-

ure 7. Both theory and experiment are shown for a constant amplitude
6

of oscillation, a = 20, at various reduced frequencies of the wedge

model. No effect of frequency is indicated by the unsteady-flow theory
nor detected experimentally over the range of reduced frequencies covered
by the study. In general, the measured change in the shock-wave angles
is slightly less than the calculated change. It should be noted that,
when the lines indicating the results of the unsteady-flow theory are
extended to zero reduced frequency, they are tangent to the result
obtained from steady-state theory.

Figure 8 illustrates the effect of oscillation amplitude on the

shock-wave angle 8, with k = 0.011 to 0.028. The two sets of data L
define the envelope of the shock angle on the lower surface of a typi- 8
cal model during a cycle of oscillation. Theory and experiment appear 3
to be in good agreement for amplitudes up to at least 0.035 radian. 9
No unsteady-flow treatment is available for calculating the shock-
wave position of blunt airfoils. However, the schlieren motion pictures
indicated that the shock-wave shape was not only independent of frequency
over the test range but oscillated in translation with no apparent
pitching motion. The shock-wave shape was calculated by the steady-
state, zero-angle-of-attack method of reference 9 by using a shock
detachment angle of 360 which was obtained from some unpublished data
obtained at the Langley Research Center. The shock shapes for the
blunt-leading-edge models can be seen in figure 9. Although this theo-
retical method for determining shock shapes is more valid near the nose,
fairly good agreement is obtained as far back as 4 diameters. No detect-
able phase angles between the model motion and the shock-wave motion were
observed over the test range for the blunt-leading-edge models.

CONCLUDING REMARKS

A wind-tunnel investigation of flow over oscillating airfoils has

been conducted on several rectangular-planform models at a Mach number
of 7. Schlieren motion pictures were taken during decaying oscilla-
tions of the models in pitch. Comparison of results obtained by use
of unsteady-flow theory with those obtained by use of steady-state
theory indicated that no unsteady-flow effects exist over the range of
reduced frequencies k, 0.007 k - 0.030, studied experimentally. The
experimental results confirmed this finding as no unsteady-flow effects
were detected in this reduced-frequency range. The shock-detachment
distance and shock-wave position on the blunt-leading-edge models could
be calculated reasonably well by established steady-flow procedures
within the frequency range of the tests.

Langley Research Center,

National Aeronautics and Space Administration,
Langley Air Force Base, Va., September 26, 1961.
 7

REFERENCES

1. Morgan, Homer G., and Miller, Robert W.: Flutter Tests of Some Simpl
Models at a Mach Number of 7.2 in Helium Flow. NASA MO 4-8-59L,
1959.
2. Miller, Robert W., and Hannah, Margery E.: Flutter Investigation of
600 to 800 Delta-Planform Surfaces at a Mach Number of 7.0. NASA
TM X-325, 1960.
L
8 3. Gibson, Frederick W., and Mixson, John S.: Flutter Investigation at
3 a Mach Number of 7.2 of Models of the Horizontal- and Vertical-Tail
9 Surfaces of the X-15 Airplane. NASA MEMO 4-14-59L, 1959.

4. Lauten, William T., Jr., Levey, Gilbert M., and Armstrong, William 0.
Investigation of an All-Movable Control Surface at a Mach Number
of 6.86 for Possible Flutter. NACA RM L58B27, 1958.

5. Hayes, Wallace D., and Probstein, Ronald F.: Hypersonic Flow Theory.
Academic Press, Inc., 1959, pp. 130-139.

6. Morgan, Homer G., Runyan, Harry L., and Huckel, Vera: Theoretical
Considerations of Flutter at High Mach Numbers. Jour. Aero. Sci.,
vol. 25, no. 6, June 1958, pp. 371-381.

7. Sewell, Geoffrey L.: A Theory of Uniform Supersonic Flow Past a Thir

Oscillating Aerofoil at Appreciable Incidence to the Main Stream.
Aero. Quarterly, vol. V, pt. 3, Sept. 1954, pp. 185-194.

8. Sewell, Geoffrey L.: Theory of an Oscillating Supersonic Aerofoil.

Aero. Quarterly, vol. II, pt. I, May 1950, pp. 34-38.

9. Love, Eugene S.: A Reexamination of the Use of Simple Concepts for

Predicting the Shape and Location of Detached Shock Waves. NACA
TN 4170, 1957.

10. Mueller, James N.: Equations, Tables, and Figures for Use in the
Analysis of Helium Flow at Supersonic and Hypersonic Speeds. NACA
TN 4063, 1957.

9
8

TABLE I,.- TEST MODEL DATA

L
8
ArolWih, Sa, Chord, TYL.E.9
Aeioin Wegt Spnn percent radius,
seciog n.chord in.

Wedge 87.6o 4 2 20 0.002

Wedge 115.06 4 2 20 .002

Double 240.36 4 4 lo .002
wedge

Double 142.9 4 4 10 .002

wedge

Blunt 71.21 4 2 3.125 .0625

leading
edge

Blunt 78.86 4 2 6.2.5 .125

leading
edge

Blunt 86.51 4 2 10 .190

leading
edge ______ __ _ _ _ _ _ _ _ _ ___ _ _ _ _
 9

TABLE II.- TEST CONDITIONS

L Shaft
8 Airfoil length, P V, q, f) b,/V
section in. slugs/cu ft ft/sec lb/sq ft cps
9
Wedge 0.55 0.000112 5,754 1,851 311 0.028
1.0 .000114 5,768 1,888 222 .020
2.0 .000102 5,698 1,668 118 .011

Wedge .65 .000111 5,761 i,844 268 .024

1.0 .000106 5,768 1,760 201 .018
2.0 .000120 5,747 1,973 110 .010

Double .55 .000122 5,677 1,969 94 .017

wedge 1.0 .000106 5,684 1,796 70 .013
2.0 .000112 5,698 1,941 38 .007

Double .55 .000106 5,698 1,736 112 .021

wedge 1.0 .000106 5,684 1,717 86 .016
2.0 .000117 5,677 1,881 46 .008

Blunt .55 .000112 5,719 1,834 329 .030

leading 1.0 .000102 5,754 1,685 229 .021
edge 2.0 .000111 5,768 1,824 131 .012

Blunt .55 .000117 5,754 1,933 298 .027

leading 1.0 .000097 5,845 1,656 217 .019
edge 2.0 .000112 5,705 1,814 123 .011

Blunt .55 .000111 5,670 1,775 286 .026

leading 1.0 .000121 5,747 1,992 206 .019
edge 2.0 .000119 5,768 1,942 117 .011
10

l00

- 0.40,,

St eel insect B~ala a Dwel

Balsa--- Piber glss 040s

Steel insert

(a) Wedge and blunt-leading-edge models.

Figure 1.- Sketches of models.

 ON 0.751

Steel insert Bla015 :.CY

Fiber Lisa'

(b) Dlouble-wedge model.

Figure 1.- Concluded.

12

co

rJ2
2E

13

Vp

3c

00
C 4J

V 4-)

-41Jq
Figure 4, Wedge-airf oil model mounted on support. L-60-3830
 tO0 t -0.2 T

t 0.4 T t -0.6 T

t-0.8 T t T

(a) Wedge model. L-61-5105

Figure 5.- Schieren photographs of airfoil models.

16

t 0.4 T t - 0.6 T

t =0.8 T t T

(b) Double-wedge model. L-6l-51o6

Figure 5.- Continued.
 17

A A N-

100

Id

:0b. 0~L

bf)
-p -"p r
 20___

Calculated re!sults
18 0

18

Q5 , deg

14

Line of agreement

12

Qr deg

Figure 6.- Comparison of shock-wave angles calculated by steady-state

theory and by the method of reference 7 for various flow deflection
angles. k = 0.028.
 '9

.3

.2

0, radians

0 .01 .02 .03

-v
Figure 7.- Effect of reduced frequency on the shock-wave angle of the
wedge models.
 20

9, radians
o Experiment
- Theory

.1B

0 .01 .02 .03 .04

a, radians

Figure .- Effect of' oscillation amplitude on shock-wave angle e

3E
21

*5.

Theory, reference 9
- -0.125 - in.-diam. leading edge-
0.25 - in.-diam. leading edge
C?4.0 --- 0.38 - in.-diam. leading edge

3.0 _ _

2.0 -_ __

1.0 -_ __

0 1.i7i
-1.0 0 1.0
77
2.0 3.0
I 4.0
X/D

Figure 9.- Shock shape of oscillating blunt-leading-edge bodies.

, ASA-Lati~iplv, 1961 L-839

 Im c.n

zz CO

4).

0 0

0 0 W W

02 X: p 04)l 0 V. 0 ;.4 0

0~4 (M ' C o 0:1.0 - 0 D1m..)4

Q) w) 0 a (L)

-0 ' o 2- Vm': 1O
4~c 4 )

0 14 V). Z .d 0( M M 4

0 0 V 4 Q) I Cl 4 - (

z.r4 w - onCl.0C24 (1

>00
z > 0 c"

Q0) n o ua)o V
>) I0)CLL Q 1 )
4)0 C 0 E
I~~ ) P4E00
.4
z < z44 0 >C 4
C))Z MO Ho0c lcsm L Zlz

0
0. 0C
.' 0 >M CU > 0

.C 4)l
r~ =.* < NV-00=Ec -
z ri'4 HaUB fl ) )9C H<4 4z)44).n

4
0z . I>4 )z. ' V >A
4)
CI oi$l4)0V al

<W)I 0s < 44 1

Vz ') .1 z

0 4
4) (n
o d0u 5
W)O .04) 4)0 w bf )
~
(V ' V

: 14 H.4 Q, :3~ -WH

', ~zb 0>'.~
Z~ 0r
1 4 0 .
fl
,,.4Z a
a 14 a)~
a)0~
0.ci

HA SOm 0 a .0 0 4).4
c: Z1>. lr I 0 ,4 '

v- ~4 C -. o- ri sV

edO to .4o.4 m~H >4 0w-o )

UE)~4
CL~.fI