TP I -

1685
NASA -Paper
Techn'ical 1685 c.1 .. - ',#

Divergence-.of Forward-Swept ,Wings

\
 TECH LIBRARY KAFB, NM

NASA Technical Paper 1685

Wind-Tunnel Experiments on
Divergence of Forward-Swept Wings

Rodney H. Ricketts and Robert V. Doggett, Jr.

Langley Research Center
Hamptou, Virginia

National Aeronautics
and Space Administration

Scientific and Technical

Information Branch

1980
 SUMMARY

An experimental study to investigate the aeroelastic behavior of forward-

swept wings was conducted in the Langley Transonic Dynamics Tunnel. Seven flat-
plate models with varying aspect ratios and wing sweep angles were at tested
low speeds in air. Three models having the same planform but different airfoil
sections (flat- late, conventional, and supercritical) were tested at transonic
speeds in FreonP12. Linear analyses were performed to provide predictions to
compare with the measured aeroelastic instabilities, which include both static
divergence and flutter. Six subcritical response testing techniques were
formulated and evaluatedat transonic speeds for accuracy in predicting static
divergence. Two "divergence stoppers" were developed and evaluated for use
in protecting the model from structural damage during tests.

INTRODUCTION

Forward-swept wing designs appear to offer selected aerodynamic performance

improvements over conventional aft-swept wings, such as higher lift-drag ratios,
lower trim drag, and better stall/spin characteristics (ref.1). In addition,
these designs may allow for improved fuselage-volume arrangements, by having the
wing box located more rearward. Two powered, forward-swept wing aircraft, both
of German design, are known to have been built to take advantage of these
improvements. These aircraft include the World War I1 vintage Junkers Ju 287
(ref. 2) and a 1960's business jet (ref. 3 ) . Until recently, serious consider-
ation has not been given to forward-swept wing designs because forward sweep
led to an unfavorable static aeroelastic characteristic, namely, static divergence
(in this paper, referredto simply as divergence). Potential gains in aerody-
namic performance were more than offset by the increase in structural mass
requi.red to provide sufficient stiffness to insure adequate divergence speed
margins. In the early 1970's, however, developments in composite structures
technology appeared to offer a solution to the problem of structural mass
increases required in the design of forward-swept wings. Analytical studies by
Krone (ref. 4) showed that divergence speeds for forward-swept wings of compos-
ite materials canbe increased substantially by optimally tailoring (arranging)
the composite lamina thicknesses and orientations without incurring significant
increases in structural mass above a so-called "strength design." As a conse-
quence of these studies, interest was aroused in applying composite materials t
forward-swept wings, particularly for fighter airplanes. The Defense Advanced
Research Projects Agency (DARPA) in cooperation with U.S. the Air Force and the
National Aeronautics and Space Administration (NASA) initiated a comprehensive
analytical and experimental program to demonstrate the feasibility of using com-
posite materials on advanced high-performance aircraft with forward-swept wings
(ref. 1).

'Freon: Registered trademark of E. I. du Pont de Nemours& Co., Inc.

L-13549
 The objectives of the present study ( 1are
) to provide some basic experi-
mental data and analytical comparisons to aid in understanding divergence
characteristics of forward-swept wings; and (2) to develop wind-tunnel exp
mental procedures applicable to studying divergence.

To accomplish the first objective, nine cantilevered wing models were

tested in the Langley Transonic Dynamics Tunnel (TDT). Seven of these models
were flat-plate wings tested to determine the effects of aspect ratio and
leading-edge wing sweep on divergence speeds in the subsonic region,
Two addi-
tional models were constructed with different airfoil shapes to determine
effort of airfoil section on the divergence boundary in the transonic regi
In this paper, the models are described, test results are presented, and c
culated results are presented for comparison with experimental results.

The second objective was accomplished by developing and evaluating sub

ical response techniques for predicting the divergence condition (dynamic pr
sure) using response measurements made below divergence. In this paper, six
different methods are described, and an application of is each
presented. Two
of these methods were recently developed by Wilmer
H. Reed 111, and their deri-
vations are presented in appendixB. In addition, two divergence "stopper"
devices were developed to prevent model damage if divergence occurs during
''
tunnel tests. These two devices, a "flow-diver ter" and a "model-constrainer,
are also described.

Use of trade namesor names of manufacturers in this report does not c

or manufacturers, either
stitute an official endorsement of such products
expressed or implied, by the National Aeronautics and Space Administration.

SYMBOLS

peak
Adynamic
amplitude, V

IR aspect
ratio,
2(s/c)

C strain-gage
proportionality
factor,
V/deg

C chord
length,
m

C1 , O lift coefficient at CI = 0

lift-curve slope
cla
Cm,ac moment coefficient about aerodynamic center
edistancebetweenelasticaxisandaerodynamiccenter,m

frequency,
f Hz

2
 torsional spring constant, N-m/deg

lift per unit span,N/m

Mach number

moment about aerodynamic center, N-m

moment about elastic axis,N-m

compression load, N

critical buckling load, N

dynamic pressure, kPa

divergence dynamic pressure, kPa

semispan length, m

angle of attack due to aerodynamic loads, deg

root angle of attack, deg

angle of attack when me = 0 (E = 01, deg

= aR - ao, deg
deflection, nnn

initial deflection, mm

divergence index parameter

mean strain-gage output, V

slope of E-versus-aR curve, V/deg

leading-edge sweep angle, deg (forward sweep is negative)

air density, kg/m3

Subscr ipts
:

exP experimental
nth n value

r value
rth
 TEST APPARATUS AND PROCEDURES

Models

Geometry.- Seven semispan flat-plate wing models were tested during t

static divergence investigation. All the wings were untapered and had a sem
span length of0 . 5 0 8 m. In planform, the models differed only in aspect ratio
and leading-edge wing sweep. The models had full-span aspect ratios of 4.0 a
8.0 and wing sweep angles ofOo, -7.5O (only = 8.01, -1S0, and -3OO. The
planform geometries for these models are shown in 1figure .

Two additional models with 10-percent-thick airfoil sections were teste

These models had a planform identical to the flat-plate models with aspec
ratios of 4.0 and wing sweep angles-15O.
of One had a conventional airfoil
(NACA 64A010); the other had
an uncambered (symmetric) supercritical airfoil
section.2 These two models, in conjunction with a flat-plate model with sim
planform, were used to determine effects of airfoil section
on the divergence
boundary in the transonic region.

Construction.- All the models were constructed of 2.29-mm-thick aluminum

alloy plate, For the flat-plate models, the leading and trailing edges were
rounded to a semicircular shape. For the models with airfoil sections, a lig
weight plastic foam was attached to the aluminum plate and then shaped t
the desired airfoil section. Transition strips (No. 46 carborundum grit) with a
width of 0.025 chord were added along the 5-percent chord on both
line the upper
and lower surfaces of all the models to assure that the boundary layer was
.
turbulent

Instrumentation.- Each model was instrumented with resistance-wire strai

gage bridges located near the wing root. The bridges were oriented to be s
tive to either bending or torsional strains.

- first
Vibration characteristics.- The first three natural frequencies
bending, second bending,andorsion - were measured for each wing model. For
comparison, mode shapes and frequencies were calculated for the flat-plate
models using theSPAR finite element structural analysis computer program
(ref. 6). Both measured and calculated frequencies for the flat-plate models
are presented in table I. The associated calculated node lines are shown in
figure 2 and were substantiated by abbreviated measurements. In figure 2, the
torsion mode is the second mode for aspect-ratio-4.0 models, and the second
bending mode is the third mode. This order is reversed for the aspect-ratio-
8.0 models.

Model Mount and Divergence Stoppers

The model wings were cantilever mounted outside the tunnel-wall bounda
layer on anI-beam support fixture attached to a remotely controlled turntabl

2This section was an early supercritical airfoil (designated NASA

SC(2)-0010) derived from the family
of cambered airfoils presented in
reference 5 .

4
A splitter plate was mounted to the support fixture
to provide a reflection
plane for the model. The turntable provided the capability of changing the
wing angle of attack during the test. A photograph showing the model mounting
arrangement is presented in figure3 .

Because of the potentially destructive nature of aeroelastic instabilities,

precautions are commonly taken to minimize the risk of model damage during w
tunnel tests. In the present study, two devices (each attached to the support
fixture) were developed for preventing excessive model deformations during
divergence. These devices are described in detail in appendix A. One device,
called a "flow-diverter" (shown infig, 4 ) , is simply a hinged plate that
deflects the airstream when itis deployed. Deflecting the airflow changes the
relative angle between the flow and the wing leading edge and in effect redu
the forward sweep, It is shown later that this situation yields a higher diver-
gence dynamic pressure. In addition, the dynamic pressureis decreased by the
shielding effect that the flow-diverter offers; that is, the is exposed to
wing
lower velocities in the wake of the device. The other device, called a "model-
constrainer" (shown infig, 5), consists of a pair of arms that are hinged to
the support fixture at one end and have wheels or rollers at the other end,
When this device is deployed, the arm swings out the span of the model, and th
wheels bear on the wing upper and lower surfaces to stiffen the model and retu
it to its undeformedshape, Both devices were demonstratedto prevent diver-
gence from occurring at dynamic pressures at least 33 percent greater than the
wing-alone divergence dynamic pressure.

Wind Tunnel

The present investigation was conducted in the Langley Transonic Dynamics

Tunnel (TDT). The TDT is a continuous-flow, single-return tunnel with a 4.88-m
square test section (with cropped corners) having slots in all four walls
(ref. 7 ) . The flow is generated by a motor-driven fan. The tunnel is equipped
to use either airor Freon 1 2 as the test medium at pressures which vary from
near vacuum to slightly above atmospheric. The range of Mach numbers is from
near zero to 1 . 2 . Both the density and the test-section Mach number are con-
tinuously controllable. The tunnel is equipped with four hydraulically acti-
vated, quick-opening bypass valves. When model instability is encountered,
these valves are actuated to rapidly reduce the dynamic pressure and Mach numb
in the test section.

Test Procedures

Low-speed tests.- Divergence tests of the flat-plate wings, evaluation of

the subcritical response divergence prediction techniques, and evaluation of
the divergence stoppers were conducted simultaneously in air at atmospheric
pressure. For these low speed tests, the determination of
a typical divergence
point proceeded in the following manner. With the angle of attack set at some
low positive value to keep the model lightly loaded a single
in direction,
the fan speed was increasedto the desired test-section dynamic pressure. This
initial dynamic pressure was chosen to be relatively far below the divergence
condition. At this dynamic pressure, data were collected to evaluate the sub-

5
critical response methods. This process involved stepping the model through a
range of angles of attack and acquiring data at each angle. The model was
returned to its original position, tunnel speed was increased to a slightly
higher dynamic pressure, and model response measurements were repeated. This
stepwise increase in dynamic pressure was continued until divergence was
reached. When divergence Occurred, damage to the model was prevented either b
or by actuating the
deploying a divergence stopper (if one had been installed),
four bypass valves in the tunnel.

Transonic tests.- Tests conducted in the transonic regime used1 2Freon

as
the test medium. The following procedure was used to vary Mach number and
dynamic pressure (shown in fig.6). With the tunnel evacuated to o
lwa stagna-
tion pressure, the fan speed was increased until the desired maximum test Ma
number was reached. The dynamic pressure at this tunnel condition was rela-
tively far below the divergence dynamic pressure. Subcritical response data
were collected at this tunnel condition in the same manner as described for
low-speed tests. Next, while the Mach number was held constant, the test-
section dynamic pressure was increased
by bleeding additional Freon
1 2 into the
tunnel through an expansion valve. When the desired dynamic pressure was
obtained, tunnel-flow conditions were held constant, and subcritical response
data were again acquired, This process was repeated until either divergence
reached or sufficient subcritical data were obtained to predict the divergen
condition at this Mach number. During this process, the flow-diverter divergenc
stopper was used to protect the model from damage.

To define the divergence condition at another Mach number, the fan spee
was decreased until the desired Mach number was obtained. With the Mach numbe
again held constant, the procedure of acquiring data and stepping the dynam
pressure was repeated in the manner just described. In this way, the divergenc
boundary was defined throughout the region of interest.

Data acquisition.- During the tests, the output signals from the model
strain-gage bridges were recorded on oscillograph strip recorders. The Spectral
Dynamics Corporation 330A Spectrascope (spectrum analyzer) was used to deter
frequencies and peak amplitudes. The tunnel data acquisition system was used t
calculate and display the parameters needed for the subcritical response predi
tion techniques.

TEST RESULTSAND DISCUSSION

Experimental results acquired during the testing of models can be divid

into two categories: ( 1 ) data from model tests at low speeds in air, and
(2) data from model tests at transonic speeds in 1Freon
2.

Low-Speed Results

All flat-plate models were tested at low speeds in air at standard atm
spheric pressure. Results of these tests are presented in figures
7 and 8 for
the aspect-ratio-4.0 and aspect-ratio-8.0 wings, respectively, as plots of
dynamic pressure versus wing sweep, Calculated flutter and divergence bound-

6
 7 for
Aspect-ratio-4.0 wings.- The calculated results presented in figure
the aspect-ratio-4.0 models are similar to the results presented by Diederich
and Budiansky (ref.10) and show that two distinct instabilities, divergence and
flutter, exist with varying wing sweep. As shown in the figure, the calculated
divergence dynamic pressure increases as the wing sweep changes to Oo.
-30° from
Conversely, the calculated flutter dynamic pressure decreases as the wing sweep
changes from-30° to Oo. The flutter mode is primarily wing first bending but
contains a small amount of coupling with torsion and second bending. The calcu-
( h = Oo) model is 18.5Hz.
lated flutter frequency for the unswept

The measured divergence and flutter points shown in 7 figure

are in good
30° forward-swept models experi-
agreement with the calculations. The 15O and
enced divergence instabilities. The unswept model experienced a flutter insta-
bility and had a measured flutter frequency of Hz.21.0

Aspect-ratio-8.0 wings.- Calculated results presented in figure

8 for the
aspect-ratio-8.0 models appear more complex than results for the aspect-ratio-
4.0 models. Three separate calculated instability boundaries are shown for
these wings: a divergence boundary and two flutter boundaries. Two of the
instability boundaries are similar to those described for the aspect-ratio-4.0
wings. One is a divergence boundary in which calculated divergence-dynamic
-30° to Oo. The other is the
pressure increases as the wing sweep changes from
upper calculated flutter boundary in which the flutter dynamic pressure
decreases as the wing sweep changes from
-30° to Oo. This flutter mode is
primarily wing first bending but contains a small amount of coupling with sec
bending and torsion. The calculated flutter frequencyfor the unswept modelis
28.9 Hz.

The other calculated flutter boundary shows up as a "hump mode" in the

analysis. Traditionally, the hump mode is characterized by the damping-versus-
velocity curve shown in sketch (a). This curve movesor up down with variations
in air density p . (At some densities, for example, p l , the hump mode lies

Sketch (a)

is not unstable.) In fig-

totally below the zero-damping line and therefore
ure 8, however, the humpmode seems to be a function of wing sweep and dis-
appears for sweep anglesgreater than Oo. The flutter modeis primarily wing
second bending but has small
a amount of coupling with torsion and first bend-
ing. Calculated flutter frequencies for the 7.5O forward-swept and the unswept
models are 48.1 and 47.9Hz, respectively.

7
 Measured divergence and flutter points shown in figure 8 are in good agree-
men witht calculated results. The15O and 30° forward-swept models experienced
divergence instabilities. The 7 . 5 O forward-swept model experienced a flutter
instability at4 6 . 4 Hz that had the appearance of a second wing bending mode.
( A node line existed near the wing tip.) This instability agrees with the cal
culated hump mode. The unswept model experienced a flutter instability with
flutter frequency of30.0 Hz. A region of significant response was observed,
however, fram q = 4 . 1 to 6.2 kPa in which the primary model response fre-
quency was about4 5 . 0 Hz. This is probably a region of low damping for the
hump mode.

Transonic-Speed Results

Three models were tested in the transonic-speed range in1 2Freon up to

M = 0.9. The purposes of these tests were ( 1 ) to acquire transonic data and
( 2 ) to determine the effect of airfoil shape on divergence.A l l these models
had an aspect ratio 4of . 0 and a wing sweep -15O.
of The models had three
different airfoil shapes- flat plate, conventional, and supercritical. In
figure 9, the measured divergence dynamic pressure for each model is presented
as a ratio to the dynamic pressure at= M 0.6 for different values of Mach
number. As shown in the figure, the region of minimum divergence dynamic
pressure, or the so-called "transonic dip," occurs at an appreciably lower
number for the conventional airfoil than for the supercritical airfoil. Also,
the width of transonic dip appears to be narrower for the conventional air
than for the supercritical airfoil. The flat-plate results show a decrease in
dynamic pressure in the transonic range, but a minimum was not observed. Ana
ysis using linear aerodynamic theory shows a transonic boundary in good ag
ment with measured flat-plate results. Linear theory is therefore useful for
the analysis of thin wings. For accurate analysis of thick wings in the tran-
sonic region, however, a more sophisticated theory is needed.

SUBCRITICAL RESPONSE - DESCRIPTION AND EVALUATION

TECHNIQUES

Subcritical response testing techniques are frequently used in flutter

testing to predict a flutter instability before it occurs (ref.
1 1 ) . Because
flutter is often a destructive phenomenon, such predictive methods allow an
risk of damaging the model. In many
approach to the instability with a minimum
cases, confidence in these methods is high enough to define a flutter ins
ity boundary without actually experiencing flutter.

A similar procedure was desired €or use in divergence testing. Therefore,

several methods were investigated to predict static divergence based on the
response of the model at dynamic pressures below the stability boundary. These
methods can be classified as either static or dynamic in nature. The static
methods include inverse mean strain, Southwell, divergence index, and constant
load. The dynamic methods include inverse peak amplitude and frequency. For
all these methods, data were acquired and analyzed in the transonic region
constant Mach number with varying dynamic pressure. this
In manner, shifts in
center of pressure due to Mach number changes were eliminated from the data

-
 The s i x methods were e v a l u a t e d o n several models a t v a r i o u s Mach numbers
w i t h similar r e s u l t s . For i l l u s t r a t i v ep u r p o s e s , however, o n l y t h e Mach 0.8
results for the aspect-ratio-4.0 flat-plate modelwith a wingsweepof -15O a r e
p r e s e n t e d . A t t h i s Mach number, t h e model e x p e r i m e n t a l l yd i v e r g e d a t a dynamic
p r e s s u r e of 2.52kPa.Discussionandevaluationofeach subcritical method
f o l l o w s . Two new methods,divergenceindexandconstantload, are d e r i v e d i n
appendix B.

S t a t i c Methods

The basic d a t a t h a t were used i n t h e s t a t i c p r e d i c t i o n m e t h o d s are pre-

s e n t e di nf i g u r e1 0 . Data f o r a rangeofmodelangles of a t t a c k were a c q u i r e d
i n t h e manner p r e v i o u s l y d e s c r i b e d i n t h e s e c t i o n "Test Procedures." A first-
o r d e r least-squares f i t was used t o c a l c u l a t e t h e slope X a n dt h ea n g l e a0
o ft h ee q u a t i o n E = X (aR - ao) f o r t h e d a t a a t eachdynamicpressure.

I n v e r s e mean s t r a i n method.- One of t h e s i m p l e s t p r e d i c t i o n m e t h o d s is t h e

i n v e r s e mean s t r a i n t e c h n i q u e , which e v a l u a t e s t h e c h a n g e i n t h e r e c i p r o c a l of
t h e mean s t r a i n w i t h t h e c h a n g e i n dynamic p r e s s u r e .T h i s method t a k e s advan-
t a g e of t h e f a c t t h a t t h e wing t i p d e f l e c t i o n , t h u s , t h e root bending moment and
s t r a i n ,t e n dt o w a r di n f i n i t y as t h ed i v e r g e n c ec o n d i t i o n i s approached. Con-
v e r s e l y ,t h ei n v e r s e so ft h e parameters tendtowardzero a t divergence.There-
fore, thedivergencedynamicpressure is t h e p o i n t a t w h i c h t h e i n v e r s e mean
s t r a i n is z e r o a n d i s p r e d i c t e d by e x t r a p o l a t i n g a second-order least-squares
f i t of t h e s u b c r i t i c a l d a t a . B e c a u s e t h e d a t a most o f t e n d i s p l a y e d a "concave
up" s h a p e ( p o s i t i v e s e c o n d d e r i v a t i v e ) , t h e f i t was r e q u i r e d t o be "concave up"
or l i n e a r i n t h e l i m i t i n g case.

Infigure11,the r e s u l t s o fa p p l y i n gt h i s method a t model a n g l e so f attack

of 0.09Oand 0.17O are p r e s e n t e d .F o rt h e s ec a s e s ,t h ep r e d i c t e dd y n a m i cp r e s -
sure i s w i t h i n 4 p e r c e n t of t h e measuredvalue.

Southwellmethod.-Thismethod was developedby R . V . S o u t h w e l l i n 1 9 3 2 to

p r e d i c t t h e c r i t i c a l buckling load of a c o l u m nl o a d e da x i a l l yi nc o m p r e s s i o n
( r e f .1 2 ) .I n1 9 4 5 , it was s u g g e s t e d by A l e x a n d e rF l a x( r e f . 13) t h a t t h i s
methodcould be used i n s t u d y i n g a e r o e l a s t i c problems l i k e d i v e r g e n c e . The
S o u t h w e l le q u a t i o n is

where 6, is t h e i n i t i a l d e f l e c t i o n m e a s u r e d l a t e r a l l y a t t h em i d d l eo ft h e
column, 6 is t h ed e f l e c t i o nm e a s u r e d from 6, for e a c ha x i allo a d P, and
PC, is t h e c r i t i c a l b u c k l i n gl o a d .

Equation (1) is similar i n form t o t h e e q u a t i o n t h a t c h a r a c t e r i z e s t h e

e l a s t i c d e f l e c t i o no fa n a e r o e l a s t i c system as d i v e r g e n c e is approached.(See,
 for example, eq. (B1Oa) i na p p e n d i x B.) F o rt h e aeroelastic s y s t e mt h e corre-
sponding equation becomes

where < is t h e i n i t i a l wing root a n g l eo f a t t a c k , a, is t h ea n g l e of a t t a c k

due t o aerodynamic load, q i s thedynamicpressure,and q D i s t h ed i v e r g e n c e
dynamic p r e s s u r e . When a e is p r o p o r t i o n a l t o t h e model s t r a i n measurement &
t h r o u g ht h ee q u a t i o n a, = CE, where C is t h ep r o p o r t i o n a l i t y f a c t o r , equa-
t i o n ( 2 ) can be p u t i n t h e f o r m

which is l i n e a ri n & and €/q w i t h slope q D . I nf i g u r e 1 2 , d a t a are pre-

s e n t e di nt h i s form f o r t w o a n g l e so f attack. A f i r s t - o r d e r least-squares f i t
of t h e d a t a was used to p r e d i c t t h e d i v e r g e n c e d y n a m i c p r e s s u r e ( s l o p e o f t h e
l i n e )i nt h e s e cases. The p r e d i c t e dv a l u e s are w i t h i n 4 percentofthemeasured
divergencedynamicpressure.

An a d v a n t a g e i n u s i n g t h i s method is t h a t t h e b a s i c d a t a ( f i g , 10) only

need to be a c q u i r e d a t a s i n g l e a n g l e of a t t a c k . However, care m u s t b et a k e ni n
c h o o s i n gt h i sa n g l e so t h a t t h e a l l a w a b l e s t r e n g t h l o a d s o f t h e model are n o t
e x c e e d e db e f o r et h ed i v e r g e n c ec o n d i t i o n is i d e n t i f i e d .

The p r e d i c t i o n methodcan be improved i f a l l t h e basic d a t a ( f i g , 1 0 ) a r e

u s e di nt h em e t h o d .I nt h i s case, t h e method i s modified so t h a t t h e s l o p e
X (where X = € / 6 ) o ft h e basic d a t a is u s e di n s t e a d of a v a l u eo fs t r a i n at a
s i n g l ea n g l eo f a t t a c k . Equation ( 3 ) t h e n becomes

I n f i g u r e 13, a r e s u l t o f a p p l y i n g t h i s method is p r e s e n t e d a s a p l o t of X
v e r s u s X/q. Again, a l i n e a rl e a s t - s q u a r e sf i t was used t o p r e d i c tt h ed i v e r -
gencedynamic pressure (slope), which i s less t h a n 2 p e r c e n t lower t h a n t h e
measured value.

Divergenceindexmethod.- The divergence index method is d e r i v e d i n

appendix B. T h i s method is basedonthe same e q u a t i o n as theSouthwellmethod;
however, t h e t e s t i n g p r o c e d u r e a n d g r a p h i c a l p r e s e n t a t i o n of t h e t e s t data are
d i f f e r e n t . The t e s t procedure c o n s i s t s ofmeasuringmodel s t r a i n as t h ea n g l e

10
 of attack is v a r i e d f o r a series o fc o n s t a n td y n a m i cp r e s s u r e s (shown i n
f i g . 1 0 ) . A t eachdynamic pressure qn, t h e slope An is measuredandthe
divergenceindex parameter An is computedfrom t h ee q u a t i o n

1 w h e r et h es u b s c r i p t r d e n o t e s a r e f e r e n c ec o n d i t i o nw
, hich is u s u a l l y associ-
a t e dw i t ht h e lowest v a l u eo fd y n a m i cp r e s s u r e . The number of An v a l u e st h a t
can be c a l c u l a t e d i s 1 less t h a n t h e t o t a l number of d i f f e r e n t dynamicpressures
foranygivenreferencecondition.

As shown i na p p e n d i x B, A is r e l a t e d t o q by t h ef o l l o w i n ge q u a t i o n :

A = 1 - (i)
T h i s is a s t r a i g h t l i n e which passes t h r o u g hu n i t y at q = 0 a n di n t e r s e c t st h e
q-axis a tt h ep r e d i c t e dd i v e r g e n c ed y n a m i cp r e s s u r e qD.

R e s u l t s of a p p l y i n g t h i s method are p r e s e n t e d i n f i g u r e 14. From t h e basic

d a t a( f i g . 1 01, f o u rv a l u e so f A were c a l c u l a t e du s i n gf i v ev a l u e so fd y n a m i c
p r e s s u r e . The d i v e r g e n c ed y n a m i cp r e s s u r e was determinedfromequation ( 6 ) by
a p p l y i n g a f i r s t - o r d e r least-squares f i t t o t h e A v e r s u s q d a t aa n dc a l c u -
l a t i n gt h eq - i n t e r c e p t . The l e a s t - s q u a r e s f i t was f o r c e dt h r o u g hu n i t ya l o n gt h e
A-axis.The p r e d i c t e dd i v e r g e n c ed y n a m i cp r e s s u r e i s w i t h i n 1 p e r c e n to ft h e
measured value.

T h i s method a p p e a r s t o g i v e accurate r e s u l t s e v e n f o r v a l u e s o f d y n a m i c
p r e s s u r ef a r removedfrom t h ed i v e r g e n c ec o n d i t i o n .T h i s is a t t r i b u t e d t o t h e
inclusionoftheLinterceptinthedata. Much a c c u r a c y is g a i n e d i n t h e
l e a s t - s q u a r e s f i t by h a v i n g t h i s u n i t y p o i n t so f a r removed from t h e r e s t of t h e
data.

Constant-loadmethod.- The f i n a l s t a t i c method t o be d e s c r i b e d i s t h e

constant-loadmethod, also d e r i v e di na p p e n d i x B. I n t h i s method,theaerody- -
namicloadmeasured by t h e s t r a i n g a g e s o n t h e m o d e l is h e l d c o n s t a n t as t h e
dynamic p r e s s u r e is i n c r e a s e dt o w a r d a d i v e r g e n c ec o n d i t i o n .T om a i n t a i nt h i s
c o n s t a n tl o a d ,t h ea n g l eo f attack i s v a r i e d .

T h i s method is b a s e d o n t h e same e q u a t i o n as t h a t u s e d i n t h e d e r i v a t i o n o f
thedivergenceindexmethod. The e q u a t i o n f o r t h i s method i s o b t a i n e d by rear-
ranging equation ( 3 ) :

11
 By d e f i n i n g E t o be c o n s t a n t t, h ee q u a t i o n is l i n e a ri n @ and q. The
divergence.dynamic pressure qD occurs when q?i is e q u a l to z e r o or, i no t h e r
words, when crosses tqh-ea x i s .

I na p p l y i n gt h ec o n s t a n t - l o a dm e t h o d to t h e basic d a t a i n f i g u r e 1 0 , t h e
valuesof E are determinedby f i r s t e x t r a p o l a t i n g t h e d a t a a t eachdynamic
pressure
- to the no-load (E = 0 ) c o n d i t i o na n dt h e nu s i n gt h er e l a t i o n s h i p
a = C~R- Qo. The results o f a p p l y i n g t h e method are p r e s e n t e d i n f i g u r e 1 5 .
A l i n e a rl e a s t - s q u a r e sf i t was used t o e x t r a p o l a t e t o @ = 0 t o p r e d i c t t h e
d i v e r g e n c ed y n a m i cp r e s s u r e . The p r e d i c t e dd i v e r g e n c ed y n a m i cp r e s s u r e is
w i t h i n 1 p e r c e n to ft h em e a s u r e dv a l u e .

Dynamic Methods

I n a d d i t i o n t o t h e s t a t i c m e t h o d sp r e v i o u s l ye x p l a i n e d , two methodsof
analyzingthedynamicsignalfromthestrain-gagemeasurements were i n v e s t i g a t e d
t o d e t e r m i n et h e i ra c c u r a c yi np r e d i c t i n gd i v e r g e n c e .B o t hm e t h o d su t i l i z e dt h e
spectrum a n a l y z e r to o b t a i n t h e d a t a . Themodel was t e s t e d a t a no-loadcondi-
t i o n and was randomly excited by t h e airstream.

I n v e r s e p e a k amplitudemethod.-Thismethod is basedontheassumptionthat
thedynamicamplitude of thedivergencemodeltends toward i n f i n i t y a s t h e
d i v e r g e n c ec o n d i t i o n is approached.Theinverse of t h ea m p l i t u d e ,t h e r e f o r e ,
w i l l a p p r o a c hz e r o , A similar approach was f i r s t used by Sandford e t a l .
( r e f , 1 4 ) to predict t h e f l u t t e r i n s t a b i l i t i e s ofan aeroelastic model.

Thismethodofdivergenceprediction was a p p l i e d i n t h e f o l l o w i n g m a n n e r .
The d a t a f o r t h e dynamicmethods were o b t a i n e d a t t h e same time as t h o s e f o r t h e
s t a t i c method. A t eachdynamicpressure, a spectrum ofthedynamicresponseof
t h e model i n a n o - l o a dc o n d i t i o n was recorded. Thewing f i r s t bending mode was
i d e n t i f i e d fromthe spectrum, and i t s p e a k r e s p o n s e was m e a s u r e d .I nf i g u r e 16,
r e s u l t s of a p p l y i n g t h i s method are p r e s e n t e d . The i n v e r s e so ft h e p e a k mea-
s u r e m e n t sa r e plotted a g a i n s t dynamicpressure. The d a t a were e x t r a p o l a t e d
u s i n g a second-order least-squares f i t t o p r e d i c tt h ed i v e r g e n c ed y n a m i cp r e s -
sure, which is t h ev a l u eo ft h ep o i n tw h e r et h ei n v e r s ee q u a l sz e r o , The pre-
dicteddivergencepressure is w i t h i n 2 p e r c e n t o f t h e m e a s u r e d v a l u e .

Frequencymethod.-Anothermethodwhichusesthe spectrum of t h e dynamic

response is b a s e d o n t h e f a c t t h a t t h e f r e q u e n c y of t h e d i v e r g e n c e mode i s z e r o
a t d i v e r g e n c e .( F o rt h i sr e a s o n ,t h ei n s t a b i l i t y i s called s t a t i c d i v e r g e n c e , )
Frequenciesfrom spectrum d a t a are u s e d t o " t r a c k " t h e wing f i r s t - b e n d i n g mode
frequencyfromthe no-wind v a l u e t o z e r o .

I n f i g u r e 1 7 , t h e r e s u l t of a p p l y i n g t h i s method i s p r e s e n t e d . A second-
o r d e rl e a s t - s q u a r e s f i t was used t o p r e d i c t t h e d i v e r g e n c e dynamic p r e s s u r e
whichoccurs when t h ef r e q u e n c y equals z e r o . The predicteddivergencedynamic
p r e s s u r e i s w i t h i n 1 p e r c e n to ft h e measured v a l u e .

12
 Remarks on Prediction Methods

In general, the static methods seemed to consistently give better quality

data than the dynamic methods. The subcritical data from static methods were
more repeatable and showed less scatter throughout the dynamic pressure range.
It was particularly difficult to acquire good quality frequencies and dynamic
amplitudes for data below1 . 0 Hz. The methods which use a linear fit of the
subcritical data for predictions were m r e accurate than those which use a
second-order fit. The linear fit methods converged to accurate predictions more
rapidly (farther from qD) than did the nonlinear methods. The Southwell,
divergence index, and constant-load methods consistently yielded similar pre-
dictions. This situation appears to be true because the three methods were
derived from the same basic equation (eq. ( 2 ) ) , Of these three static methods,
the Southwell data contained the least amount of scatter. Although the example
given previously in the paper shows that the inverse mean strain method is
accurate, other examples showed it to be conservative by underpredicting the
divergence dynamic pressure. This result may be due to the choice of the
second-order curve fit for extrapolation.A hyperbolic curve fit, for example,
might be more appropriate.

CONCLUSIONS

An experimental study of the static aeroelastic divergence of forward-swept

wings has been described. Nine plate mdels, two of which had airfoil sections
attached, were used in the study. The models were wind-tunnel tested at low
speeds in air, and at transonic speeds in Freon1 2 to determine divergence
characteristics. Subcritical response testing techniques for predicting diver-
gence were formulated and evaluated. Two divergence stoppers were developed and
tested to determine their effectiveness in protecting a model at divergence.

The important results follow:

or the model-
1. A divergence-stopper device, such as the flow-diverter
constrainer, can be effectively used during divergence testing to help protect
the model from destruction.

2. Linear theory accurately predicts the aeroelastic behavior, including

divergence and flutter, of thin forward-swept wings. For accurate predictions
of divergence characteristics of thick wings in the transonic region, a more
sophisticated theoryis needed.

3. The aeroelastic divergence boundary can be accurately defined using sub

critical response testing techniques.

Langley Research Center

National Aeronautics and Space Administration
Hampton, VA 23665
June 16, 1980

13
 APPENDIX A

DIVERGENCE STOPPERS

Wind-tunnel t e s t i n g of aeroelastic models o f t e n j e o p a r d i z e s t h e model

becauseinstabilitiessuch as f l u t t e r a n d d i v e r g e n c e p r o d u c e l a r g e , r a p i d l y
i n c r e a s i n g model d e f o r m a t i o n st h a tc a nl e a d to s t r u c t u r a l failure. I t is com-
mon practice, therefore, t o take p r e c a u t i o n s d u r i n g t e s t i n g to m i n i m i z et h e r i s k
of model damagewhen an aeroelastic i n s t a b i l i t y is encountered. As mentioned i n
t h e bodyof thepaper,theapplication of s u b c r i t i c a l r e s p o n s e t e s t t e c h n i q u e s
is one way to minimize r i s k s . However, even i na p p l y i n gs u b c r i t i c a lm e t h o d s , it
is u s u a l l y n e c e s s a r y to d e t e r m i n e a t l e a s t o n e i n s t a b i l i t y p o i n t to v a l i d a t e t h e
met hod.

Another means of r e d u c i n g t h e r i s k of model damage is to conduct aeroelas-

t i c model s t u d i e s i n wind t u n n e l s t h a t h a v e a means of r a p i d l y r e d u c i n g flow
dynamic p r e s s u r e .T h e s e methods i n c l u d e spoilers t h a t are d e p l o y e di n t h e tun-
n e l d i f f u s e r t o p r o v i d e a c h o k i n g e f f e c t and valve-piping arrangements that are
used to s h o r t c i r c u i t t h e flow between the low-speed l e g of t h e t u n n e l a n d t h e
t e s t - s e c t i o n plenumchamber.The s p o i l e r is u s u a l l yv e r ye f f e c t i v e ,b u tb e c a u s e
largeloadscan develop o n t h e d e v i c e when it is d e p l o y e d , t h e spoiler is
u s u a l l yu s e do n l yi n small t u n n e l s . The valve-pipingsystem is more applicable
to l a r g e t u n n e l s (i.e., theLangleyTransonicDynamicsTunnel)and does n o t h a v e
as much e f f e c t as t h e spoiler. However, i n b o t h cases, t h e wind t u n n e lm u s t be
equippedwith t h e d e v i c e b e f o r e it c a n be u s e d .B o t hr e q u i r ee x t e n s i v em o d i f i -
c a t i o n s to t h e t u n n e l .

Othermethods of m i n i m i z i n g t h e r i s k of model damage i n c l u d e p h y s i c a l l y

r e s t r a i n i n g t h e model w i t h cables which are normally slack b u t become t a u t when
t h e model d e f l e c t i o n r e a c h e s a preset v a l u e .A l t h o u g hs u c hr e s t r a i n tm e t h o d s
are u s u a l l y effective i n m i n i m i z i n g model damage, t h e p r e s e n c e of cables d i s -
torts t h e f l o w o v e r t h e model and may change its aeroelastic characteristics.

Inthepresentstudy two d e v i c e s , or i n s t a b i l i t y stoppers, t h a t can be used

to p r e v e n t damage to aeroelastic wind-tunnel models d u r i n g t e s t i n g were devel-
opedanddemonstratedsuccessfully.Bothdevices are m e c h a n i c a l l y simple, capa-
ble o f r a p i d a c t u a t i o n a t any test c o n d i t i o n , a d a p t a b l e f o r u s e i n a n y wind
tunnel,noninterferingwith t h e f l a w f i e l d around t h e model,and noninterfering
w i t h t h e dynamic characteristics of t h e model. Forpurposes of d i s c u s s i o n ,o n e
d e v i c e is referred to as t h e f l o w - d i v e r t e r , a n d t h e other is r e f e r r e d to as t h e
model-constrainer. The purpose of t h i s appendix is to describe each of t h e s e
dev i ces .
Flow-Diver ter Device

The f l a w - d i v e r t e r is i l l u s t r a t e d s c h e m a t i c a l l y i n f i g u r e s 18 and 19. This

d e v i c e was d e v e l o p e df o ru s ei nd i v e r g e n c es t u d i e s of forward-sweptwings.The
simplest form of t h i s d e v i c e is a h i n g e d p l a t e t h a t is mountedupstream of t h e

14
I

APPENDIX A

model. I n thestoredposition(fig. 1 8 ) , theplate is recessed i n the wind-

tunnelwall, or s p l i t t e r p l a t e , so that there is no aerodynamic interference
produced by the plate. For thin plates, a f l u s h mounting w i t h the wind-tunnel
wall is satisfactory. When t h e device is actuated(fig. 191, theplate is
deployed into the airstream by a quick acting, remotely controlled, pneumatic or
hydraulicactuator. When the plate is deployed, the flow is diverted over t h e
outboard portion of the model, which has the e f f e c t of reducing the sweep angle
of the model. T h i s reduction i n sweep angleincreasesthedivergence speed.
Furthermore, the dynamic pressure is decreased by the shielding effect that the
flow-diverter offers; the wing is exposed t o lower velocities i n the wakeof the
device.

For applications i n which the model is mounted off the wind-tunnel wall on
a s p l i t t e r p l a t e , two plates may
be used (shown i n fig. 2 0 ) . The inner plate
diverts the flow behind the s p l i t t e r p l a t e so that the air is channeled over the
inboard portion of the model through a hole i n the splitter plate. The purpose
of t h i s inner plate is to relieve the suction pressure behind the outer plate,
which functions as the previously described single-plate device, and to allow it
to bemore effective i n turningthe flow. I f a more gradualturning of the flow
is required, a multisegment outerplate can be used. A two-segment device is
shown i n figure 21 .
Both thesingle-plate ( f i g . 1 8 ) and the two-plate ( f i g . 20) flow diverters
wereused i n thepresent study. Both applications proved effective i n rapidly
returningthe model to an undiverged condition. Although the s t a t i c deformation
was reduced, the models d i d experience sane randandynamic response that was
apparently produced by turbulent flow off the edges of theoutboard plate. The
randm response was less for the two-plate case, andwas notconsidered t o be
excessive i n eithercase.

Model-Constrainer Device

The model-constrainer, which is shown schematically i n figure 22, is appli-

cableto both f l u t t e r and divergence testing. As i l l u s t r a t e d i n thefigure,the
device consists ofanarm that is hinged a t oneend to the splitter plate (or
tunnelwall). A pair of s o f t wheels, or rollers, is attachedtotheother end
of the arm. When thedevice is actuated,the arm rotates awayfrom thewall,
and the wheels r o l l along the upperandlower surfaces of the model, t h u s
returning the model t o i t s undeformed shape and preventing either static or
dynamicmodel deflection. The device is operated by a remotely controlled pneu-
matic,orhydraulic,actuator. To minimizeaerodynamic interference, thedevice
can be recessed i n the wind-tunnel wall, s p l i t t e r p l a t e , orfuselage half-body,
depending on theapplication. Although the illustrations are for application to
a forward-swept wing, t h e device is equally applicable to aft-swept wings for
use as a f l u t t e r stopper. For aft-swept wings, thedevice would bemounted
downstreamof t h e model.

The model-constrainer wasshown t o be very effective i n restraining the

model
when divergenceoccurred. An advantage of t h i s device over the flow-
diverter was that the models d i d notexperience random excitation.

15
 APPENDIX B

DERIVATION OF Two METHODSFORPREDICTINGSTATIC DIVERGENCE OF

WINGS FROM SUBCRITICALTEST DATA

Wilmer H . Reed I11

Langley Research Center

I d e a l i z e d Aeroelastic System

I nt h i sa p p e n d i x , a d e r i v a t i o n is given for t h e t e s t i n g t e c h n i q u e s r e f e r r e d
t o i n t h e bodyof t h e p a p e r as t h e " d i v e r g e n c e i n d e x " m e t h o d a n d t h e " c o n s t a n t -
load" method.Thesemethods are developedonthe basis of a s i m p l i f i e d aero-
e l a s t i c systemwhich is assumed t o r e p r e s e n t a " t y p i c a l s e c t i o n " o f a flexible
wing. As shown i n f i g u r e 23, t h e typical s e c t i o n c o n s i s t s of a r i g i d a i r f o i l
mountedon a t o r s i o n a l s p r i n g l o c a t e d a t t h e e f f e c t i v e e l a s t i c a x i s of t h e s y s -
tem, The base o ft h es p r i n gc a n be i n c l i n e d a t a na n g l e CIR r e l a t i v e t o t h e
flow d i r e c t i o n as a means of l i f tc o n t r o l . The a n g l er e p r e s e n t st h e twist
of t h es p r i n gd u e t o aerodynamic loads on t h e a i r f o i l . Thus,theaerodynamic
a n g l eo f a t t a c k of t h e a i r f o i l is t h e r i g i d - b o d y a n g l e p l u s a n i n c r e m e n t due
to a e r o e l a s t i c deformation

The l i f t f o r c e per u n i ts p a n 2 a c t i n g a t t h ea i r f o i lc e n t e r of p r e s s u r ea n d
t h e manent a b o u t t h ea e r o d y n a m i cc e n t e r mac are, r e s p e c t i v e l y ,

where by d e f i n i t i o n , is independent of a n g l e of a t t a c k .

Thus, theaerodynamicmanent about t h e elastic a x i s is

I+ = e2 + mac

where e is t h ed i s t a n c eb e t w e e nt h ea e r o d y n a m i cc e n t e ra n dt h e elastic a x i s
(e is p o s i t i v e when theaerodynamiccenter i s forwardofthe elastic a x i s ) .

For a f l a t - p l a t e a i r f o i l i n t w o - d i m e n s i o n a l ,i n c a n p r e s s i b l e flow, cll0

and %, ac are zero,andequation (B4) becomes simply

16
r

To retain the simplicityof this form of the moment equation for the
more general case in whgch C Z , ~and Cm,aC are nonzero, it is convenient
to introduce an angle shown in figure 23 and defined as

-
a = aR - .a
where .a is the rigid-body angleof attack for which the aerodynamic moment
about the elastic axis is zero. The valueof .a required to satisfy the con-
dition me = 0 is, from equations (B2), (B3), and (B4), found to be

(B5) now becomes

The general form of equation

balanced by the
This is the aerodynamic moment about the elastic axisiswhich
torsion spring so that

Substituting equation (B8) into equation (B9) and solving for ae gives

(B1Oa)

where

(B1Ob)

Because the denominator of equation (BlOa) vanishes qas qD, causing the
+

twist of the spring to become infinitely large, qDis the dynamic pressure at
divergence. Furthermore, because dynamic pressureis always a positive real

17
 APPENDIX B

> 0,e
quantity, equation (Blob) indicates that divergence can occur only when
i.e., when the elastic axis is behind the aerodynamic center.

From equation (BlOa) the divergence dynamic pressurebemay expressed in

terms of the experimentally determined quantities q, or, and ae.

To relate divergence predictionsfor an actual wing structure to those

derived for the idealized two-dimensional section treated here, the quantity
ae in equation (B11) is assumed to be proportional to the output
of a strain
gage on the wing structure which senses elastic deformation of the wing du
load as in the equation

A typical setof data taken during the test

of a wing structureis illustrated
.
in sketch (b) The data are linear and have a slope 1 such that

P
x
E

0 a
aO R
Sketch (b)

Divergence Index Method

The test procedurefor the divergence index method consists

of measuring a
of attack is varied at constant dynamic pres-
strain-gage output as the angle

18
 APPENDIX B

sure. A series of such measurements are taken at successively higher dynamic

pressures, Si, 92, . . ., qn, holding Mach number constant if cmpress-
ibility effects are significant. Letting be the slope measured at qn
and using equations (B12) and (B13), it is easily shown that

Letting x, and qr be values for a reference condition thatis well below

theexpecteddivergencecondition,theratioisformedtoeliminate C
such that

where qr < qn < qD. Solving equation (B14) for qD yields

Thus, for each set of €-versus-aR measurements taken at different values

q, of
a new prediction of qD is obtained from equation (B16).

To provide a convenient graphical display which further aids in the esti

mation of the divergence dynamic pressure from subcritical data, a so-called
"divergence index" parameter is defined as follows:

Values of A are computed from the following equation:

An application of the equations to a set of subcritical is illustrated

data in
sketch (c). When plotted against q, the parameter A decreases linearly with
negative slope (-9D-I and crosses the q-axis at divergence.

19
 APPENDIX B

1.0

A = 41

"
U
4

Sketch (c)

When the measured data contain scatter, the accuracy of divergence pre
tion may be enhanced by applying the method of least squares. In this case, the
unknown qD would be determined from a set of n equations (n > 2 ) which
relate 91, 92, ... , qn to the observed slopes x,, X2, ,, .. x,.
Constant-Load Method

This method - involves

varying the dynamic pressure while controlling the
rigid-body angle CL so as to maintain constant strain measurement € . A prac-
tical advantage of this method over some othersis that it minimizes therisk of
overloading the model as the divergence conditionis approached during tests,
Because the aerodynamic load is held constant(ae = Constant), from equa-
tion (BlO) the derivative of c$i with respect to q, becomes

An illustration of this methodis shown in sketch (a), which illustrates that

@ versus q plots as a straight line with a negative slope ofae.Diver-
gence is indicated when @ = 0, that is, when = 0 or, in other words, when
the load is developed entirely by the angle of attack associated with the aer
elastic deformation ae.

20
I

APPENDIX B

E = Constant

+
ae
\ = 9 [

0
4
\
'\ 1
Sketch (d)

21
 REFERENCES

1. Forward-Swept Wing Potential Studied. Aviat. Week b Space Technol.,

vol. 110, no. 5, Jan. 29, 1979, pp. 126-127.

2. Holzbaur, Siegfried: Swept-Forward Wings. Interavia, vol. V, no. 7, 1950,

pp. 380-382.

3. Taylor, John W. R., ed.: Jane's All the World's Aircraft, 1970-71. McGraw-
Hill Book Co. , c. 1970.

4. Krone, Norris J., Jr.: Divergence Elimination With Advanced Composites.

AIAA Paper No. 75-1 009 , Aug. 1975.

5. Whitcomb, Richard T.: Review of NASA Supercritical Airfoils. ICAS Paper

NO. 74-10, Aug. 1974.

6. Whetstone, W. D.: SPAR Structural Analysis System Reference Manual - System

Level 1 1. Volume 1 : Program Execution. NASA CR-145098-1 , 1977.

7. Yates, E. Carson, Jr.; Land, Norman S.; and Foughner, Jerome T.: Measured
and Calculated Subsonicand Transonic FlutterCharacteristics of a 45O
Sweptback Wing Planform in Air and inFreon-12 in the Langley Transonic
Dynamics Tunnel. NASA TN D-1616 , 1963.

8. Watkins, Charles E.; Woolston, Donald S.; and Cunningham, Herbert J.: A Sys-
tematic Kernel Function Procedure for Determining Aerodynamic Forces on
Oscillating or Steady Finite Wings at Subsonic Speeds. NASA TR R-48, 1959.

9. Desmarais, Robert N.; and Bennett, Robert M.: User's Guide for a Modular
Flutter Analysis Software System (FAST Version 1.0). NASA !I'M-78720, 1978.

10. Diederich, Franklin W.; and Budiansky, Bernard: Divergence of Swept Wings.
NACA TN 1680, 1948.

11. Flutter Testing Techniques. NASA SP-415, 1976.

12. Southwell, R. V.: On the Analysis of Experimental Observations in Problems

of Elastic Stability. Proc. Roy. Soc. (London), ser. A, vol. 135, Apr. 1,
1932, pp. 601-616.

13. Flax, Alexander H.: The Influence of Structural Deformation onAirplane

Characteristics. J. Aeronaut. Sci., vol. 12, no. 1, Jan. 1945, pp. 94-102.

1 4 . Sandford, Maynard C.; Abel, Irving; and Gray, David L.:

Development and
Demonstration of a Flutter-Suppression System Using Active Controls. NASA
TR R-450, 1975.

22
I I Meas u r ed I Calculated 1
A' rn

deg fl ' f2' f 3' fl ' f 2' f 3'

Hz HZ Hz Hz Hz Hz

Aspect ratio 4.0

0 7.1 31 - 6 44.2 32.2 7.546.5

-

-15 6.8 31.2 43.6 31.3 7.145.3

a7.2 35.5 48.0
b6 .7 30.5 42.8

-3 0 5.7 27.8 39.8 41.4 28.7 6.0

I

Aspect r a t i o 8 . 0
-
0 7.0 59.8 43.8 46.3 7.459.8

-7.5 6.9 57.2 7.3

42.8 45.5 60.0

-1 5 41.2 6.656.4 43.2 7.060.4

-30 5.5 57.4 33.45.8 35.3 60.7

aModel w i t h 64A010 a i r f o i l c o n t o u r .
bModel w i t h s u p e r c r i t i c a l a i r f o i l c o n t o u r .

23
 c = 0.254 m

1-1

0 0
A = Oo A = -15 A = -30

(a) A s p e c t r a t i o 4.0.

c = 0.127 m
l-4

-
Flow

0 0 0
.A = O A = -7.5 A = -15 A = -30 0

(b) A s p e c tr a t i o 8.0.

Figure 1.- Planformgeometry of the experimental models.

24
 I
Second mode
I
I
""
t ""
T h i r d mode
I
I
I
I
I
I
I
I
I
I
'////////////////A
0 0 0'
A = o A = -15 A = -30

(a) Aspect-ratio-4.0 models.

I
I

"t-"
.i
iI

0 0
A = Oo A = -7.5 A = -15 A = -30'

(b) Aspect-ratio-8.0 models.

Figure 2.- Calculated node lines for aspect-ratio-4.0 and aspect-ratio-8.0

models. (Node lines for first mode are along the cantilever root.)

25
 L-78-1969
Figure 3 . - Typical modelmounted onsupportwithsplitterplateinstalled.
Figure 4.- Flow-diverterdeviceinextendedpositionforstoppingdivergence.
 678-4291
Figure 5 . - Model-constrainer device in extended position for stopping divergence.
 20 .o

10 .o

8.0

6 .O
Divergence boundary

4 3

kPa
4 .O

q=?@
v’ 4
2 .o
f
/
/
1.o /
.8

.6
0 .2 .4 .6 .8 1 .o 1.2
M

Figure 6.- Test procedureforoperationinFreon 12 i n

Langley Transonic Dynamics Tunnel.

29
 EXPERIMENT
0 Flutter

lo r Calculated
a Divergence

8 .-

6 - -

g9
kPa

4 -- Calculated
divergence

Stabl e

2 -

0 I I I I I
-40 -30 -20 - 10 0 10

Figure 7.- Experimental instabilities (M < 0.3) for aspect-ratio-4.0

wings compared with calculated boundaries.

30
 EXPERIMENT
0 Flutter
10 A Divergence

Calculated
8
boundaries

4,
kPa

2 Stabl e

0
-40 -30 -20 - 10 0 10

Figure 8.- Experimental instabilities (M < 0.3) for aspect-ratio-8.0

wings caupared with calculated boundaries.

31

%...
 EXPERIMENT
0 Flat-p
a il raft oe i l
" 0- Conventional airfoil
- - -A- - Supercritical airfoil
1.2

1.o

4.8
'M=O. 6

Stable
.6

.4 .5 .6 .7 .8 .9 1.o

Figure 9.- Canparison of divergence boundaries for flat-plate, conventional,

and supercritical airfoil models. R = 4.0; A = -15O.

32
I'

g = 2.39 kPa
3.2

2.8

2.4

2.0

. 9 3 kPa
E,V 1.6
.125

1.2

.a

.4

deg
R,

Figure 10.- Basic data for subcritical response divergence

prediction techniques. R? = 4.0; h = -1 So; M = 0.8.

33
 -
2.4 - ct = 0.09
0

= Intercept = 2 . 5 8 kPa
qD I

2.0

1.6

.8 =
-
ct = 0.17'
Intercept = 2 . 6 1 kPa -4\ .52 kPa

.4

0 .4 .8 1.6 1.2 2 .o2 . 8 2.4

49W a

Figure 1 1 . - S t a t i c i n v e r s e mean s t r a i n method for p r e d i c t i n g d i v e r g e n c e .

34
 5

qD, exp = 2 . 5 2 kPa

- 0
a = 0.09
qD = S l o p e = 2 . 6 1 kPa
3

E,V

0 .2 .4 .6 .8 1.0

E/q, V k P a

Figure 1 2 . - S t a t i c S o u t h w e l l method for p r e d i c t i n g d i v e r g e n c e .

35
 1.0
‘D, exp = 2 . 5 2 kPa

.8

.6

.4
= S l o p e = 2.49 kPa
qD

.2

0 .1 .2 .3 -4 .5

x / q , (V/deg) /kPa
Figure 13.- Improved static Southwell method f o r p r e d i c t i n g d i v e r g e n c e .

36
 1.2

1 .o

.8
q,, = Intercept = 2 . 5 1 kPa

A .6

52 kPa
.4

.2

(4,W a

Figure 1 4 . - S t a t i c d i v e r g e n c e i n d e x method for predicting divergence.

37
 .7

q D = Intercept = 2.50 kPa

.5

.3

.2

Figure 15.- S t a t i c c o n s t a n t - l o a d method for p r e d i c t i o n of divergence.

38
 0 .4 .8 1.2 1.6 2.0 2.4 2.8

4, kPa

Figure 1 6 . - Dynamic i n v e r s e peak amplitude method for p r e d i c t i n g d i v e r g e n c e .

 7

kPa
2

0
1.6 .4
1.2 .8 2.4 2.0 2.8

4, kPa

Figure 17.- Dynamic frequency method f o r p r e d i c t i n g d i v e r g e n c e .

40
 Rem0 t e l y
controlled

Pneumaticactuator
(hydraulicorelectro-
mechanical actuators
can b e used also)

Wind-tunnel wall

Diverter plate

Figure 18.- Flow-diverter device with diverter plate in retracted position.

 Remotely
controlled
Pivot / / / /
valve

Pneumatic actuator
Air

Wind-tunnel wall Shaft

I\\\\\

Diverter
plate

G?
\ Diverted
airflow

Figure 19.- Flow-diverter device with diverter plate in extended

p o s i t i o n to stop divergence.
 P i n n e ds h a f t( o r
push rod with
returnspring)
Wind- t:unne 1

Inner diverter

Splitter plate

'U
Outer diverter
Model p l a te

Diverted
airflow
U
F i g u r e 20.- Flow-diverter device w i t h a d d i t i o n a l p l a t e to d i v e r t
boundary-layer a i r .
Ip
W
 Pinned shaft (or
push rod with spring)

Gear or linkage
Wind-tunnel wall

diverter plate

Figure 21 .- Flow-diverter device with segmented diverter plate.

I
 Wind-tunnel w a l l

"

Sliding sleeve
Soft wheel
(multiple wheels or
t r a i n i npgr e f e r a b l e rollers arm
i n some applications)

\ U
I \Tension spring

Figure 22.- Model-constrainer device for stopping divergencein

retracted and extended positions.
 1
f

Wind

I
I

Figure 23.- I d e a l i z e d a e r o e l a s t i c system.

 I 1. Report No.
NASA W-1685 I 2. GovernmentAccession No.
I 3. Recipient's Catalog No.
. ~

4. Title and Subtitle 5. Report Date

WIND-TUNNEL EXPERIMENTS ON DIVERGENCE OF

FORWARD- - Ausust 1980
6. Performing OrganizationCode
SWEPT WINGS

7. Author(s1 8. Performing Organization Report No.

Rodney H. R i c k e t t s and Robert V. Doggett, Jr. L-13549
-
10. Work UnitNo.
9. Performing Organization Name and Address 505-33-53-01
-
NASA Langley Research Center 11, Contract or Grant No.
Hampton, VA 23665
13.Type of Report andPeriodCovered
12. SponsoringAgencyNameandAddress Technical Paper
NationalAeronauticsandSpaceAdministration -
Washington, M: 20546 14. SponsoringAgencyCode

15. Supplementary Notes

Appendix B by Wilmer H. Reed 111, Langley Research Center.

i
~~~

16. Abstract

An e x p e r i m e n t a l s t u d y to i n v e s t i g a t e t h e a e r o e l a s t i c b e h a v i o r o f f o r w a r d - s w e p t
1 wings was c o n d u c t e di nt h eL a n g l e yT r a n s o n i c DynamicsTunnel.Seven flat-plate
models with varying aspect ratios andwing sweep a n g l e s were t e s t e d a t low s p e e d s
i na i r .T h r e em o d e l sh a v i n gt h e same p l a n f o r m b u t d i f f e r e n t a i r f o i l s e c t i o n s
(i.e., f l a t - p l a t e , c o n v e n t i o n a l a n d s u p e r c r i t i c a l ) were t e s t e d a t t r a n s o n i cs p e e d s
i nF r e o n @ l 2 .L i n e a ra n a l y s e s were performed to p r o v i d e p r e d i c t i o n s to compare w i t h
t h e m e a s u r e d aeroelastic i n s t a b i l i t i e s which i n c l u d e b o t h s t a t i c d ~ i v e r g e n c e a n d
f l u t t e r .S i xs u b c r i t i c a lr e s p o n s et e s t i n gt e c h n i q u e s were f o r m u l a t e da n de v a l u a t e d
a tt r a n s o n i cs p e e d s for a c c u r a c yi np r e d i c t i n g s t a t i c d i v e r g e n c e . Two " d i v e r g e n c e
stoppers" were d e v e l o p e d a n d e v a l u a t e d f o r u s e i n p r o t e c t i n g t h e modelfrom struc-
t u r a l damage d u r i n g tests.

~ ~~ ~

17. Key Words(SuggestedbyAuthor(s1 1 18. Distribution Statement

Aeroelasticity Unclassified - Unlimited
Divergence
Divergence stoppers
Forward-sweptwings Subject Category 05

1
Subcriticalresponsetechniques
-
19. SecurityClassif. (of this report)
Unclassified Unclassified
20. SecurityClassif. (of this page) 21. Ni60f Pages
I 22. Price
A03

For sale by the National Technical Information Service, Springfield, Vlrglnla 22161
NASA-Langley, 1980
 . .
National Aeronautics and ' ' , ~ T H I R D - C L A S S BULK R A T E ,Postage and Fees Paid
National Aeronautics and
Space Administration , I '
Space Administration
_ . ~ .'
, ,
NASA451
- Washington; D.C. > .

20546
Official Business
.Penalty' for Private Use, $300 - ..
. .

1 1 IU,A, 0 7 0 8 8 0 S00903DS
1 ' ,- DEPT OF THE AIR FORCE
AF WEAPONS LABORhTORY
.,
BTTH: TECHNICAL L I B B A B Y (SUL) ( ' I _ '

I , . \ '
KIBTLAND BFB NH 87117
"