C.P. No. I171
2
MINISTRY OF DEFENCE (AVIATION
SUPPLY)
AERONAUTICAL RESEARCH COUNCIL
CURRENT
PAPERS
An Hypothesis for the Prediction
of Flight Penetration
of Wing
Buffeting from Dynamic Tests
on Wind Tunnel Models
by
D. G. Mobey
Aerodynamics Dept., RAE., Bedford
LONDON:
HER MAJESTY’S STATIONERY OFFICE
1971
PRICE 45p NET
i
UDC 533.6.013.43
: 533.693
CP No.1171*
October 1970
AN HYPOTHESISFOR THE PREDICTION OF FLIGHT PENETRATIONOF WING BUFFETING
FROMDYNAMIC TESTS ON WIND TUNNEL MODELS
D. G. Mabey
SUMMARY
Buffeting
wing buffeting
coefficients
appropriate to the maximum flight
penetration
of
for both transport and fighter
type aircraft
are deduced from
the comparison of flight
observations and measurements of unsteady wing-root
coefficients
thus deduced
strain on stiff
wind tunnel models. The buffeting
are appropriate for predictions
of buffet penetration
on future aircraft.
These predictions
are likely
to be particularly
useful for comparative tests on
project models with alternative
wing designs.
necessary buffeting
coefficients
are derived rapidly from the unsteady
wing-root strain measurements.
The tunnel unsteadiness (which must be known) is
used as a given level of aerodynamic excitation
to calibrate
the model response
The
at the wing fundamental frequency; a detailed knowledge of the structural
characteristics
of the model is thus not required.
*
Replaces RAE Technical
Report 70189 - ARC 32684.
a
3
INTRODUCTION
4
THE RELATION BETWEENTUNNEL UNSTEADINESSAND THE MODELRESPONSE
6
DETAILS OF ANALYSIS
COMPARISONOF MODELBUFFETING CONTOURSAND FLIGHT PENETRATION
4
7
BOUNDARIES
8
5
SIGNIFICANCE OF BUFFETING COEFFICIENTS
9
6
DISCUSSION
10
7
CONCLUSIONS
11
Table 1
Assessment of buffet penetration
criteria
12
References
Figures 1-14
Illustrations
1
2
3
Detachable
abstract
cards
3
1
INTRODUCTION
Dynamic tests on a stiff
wind tunnel model can give useful predictions
if scale effects on the boundary
for buffet onset boundaries on aircraft'
layer development are small, and if the tunnel unsteadiness does not exceed
In addition to buffet onset boundaries,
the criteria
specified
in Ref.2.
maximum flight
penetration
also of current interest.
boundaries for transport and strike aircraft
are
These penetration
boundaries should be related in
The severity of buffeting
loads
some way with the severity of buffeting
loads.
3,495
in flight
can also be predicted from dynamic tests on stiff
models* if the
scale effects on the excitation
spectra are small and if the buffet excitation
spectra are uncorrelated with the tunnel unsteadiness.
For this classic method 3,495 of predicting
buffet
stiffness
and damping on both the model and the aircraft
fied but this information
may not be available
for the
project stages.
In the new method described here, also
of wing-root strain on models, the mass, stiffness
and
severity the mass,
must be fully speciaircraft
during early
based on measurements
damping need not be
specified
for either
the model or the aircraft.
The hypothesis is that
the tunnel unsteadiness (which must be known), can be used as a given
level of aerodynamic excitation
to calibrate
the model response at the wing
coefficients
from the
fundamental frequency, and hence to derive buffeting
coefficients
are a measure of the
buffeting
measurements. These buffeting
generalised force in the wing fundamental mode due to any distribution
of
It has been concluded from past experience
pressure fluctuations
on the wing.
with 9 aircraft
models that levels of buffeting
coefficient
obtained in this
way can be identified
appropriate to the maximum flight
penetration
of buffet
for both transport and fighter
type aircraft.
The buffeting
appropriate to the heavy buffeting
limit appears consistent
6&7
of normal force fluctuations
.
It
obtained
Reynolds
cies and
that the
is interesting
that almost the same buffeting
coefficients
have been
on 2 similar
research models at different
scale tested at the same
number but with different
structural
damping, different
wing frequendifferent
levels of reference tunnel unsteadiness.
This suggests
basic hypothesis with respect to the use of the tunnel unsteadiness
as a scale
*
coefficient
with measurements
for the buffet
excitation
is valid.
Ordmarywmd tunnelmodelsmadewith sohdwmgsofsteelor hghtalloy areusedfor buffetmgtests1,3,4,5
For thesemodelstheh& ratloof(modeldenslty)/(freestreamdennty)ensules
that thestructuraldampIng
coefficmntpredommates
overtheaerodyndnuc
dampmgc.oelliaent,sothatthetotdl dampux$
of thewmgfundaISlmpltclt m equatmn(1)
;ene$l mode1sIndependent
ofwmd velocityanddensity Thmunportantobservatlon
4
2
THB RELATION BETWEENTUNNELUNSTEADINESSAND THE MODELRESPONSE
The basic hypothesis is that the response of the model wing to the
unsteadiness in the air stream before the onset of significant
flow separations
on the model can be linearly
related to the tunnel unsteadiness (linear systems
are generally assumed for response calculations)
and that the tunnel unsteadiness does not interfere
with the development of the flow separations.
At any angle of incidence above buffet onset the wing responds to both
the tunnel unsteadiness and the buffet pressure fluctuations.
If we assume
that the same linear relationship
between the wing response and the tunnel
unsteadiness applies between the wing response and the buffet pressures, model
response is then a direct measure of the buffet pressures and may be calibrated
by the known tunnel unsteadiness.
If this hypothesis can be substantiated,
curves of unsteady wing-root strain (model response) against angle of incidence
can be transformed into curves showing the variation
of equivalent excitation
below
or buffeting
coefficients
on the model. The corresponding excitation
buffet onset is the tunnel unsteadiness function,
at the wing fundaGnF,
mental bending frequency
The tunnel
pressure
fl.
unsteadiness
fluctuation
is defined
JnFo
coefficient
is given by
so that
the total
rms
m
72
PI4
where
n
w
v
=
=
=
and
F(n) = contribution
to
and
F/q2
in a frequency
at the wing fundamental
JnFo
p
nF(n) d (log n)
i
-co
flW/V
tunnel width
free stream velocity
The tunnel unsteadiness
analysis is
where
=
band
frequency
Af.
required
for this
= P/P(E)~
= pressure fluctuation
in a frequency
pressure
q = ip V2 kinetic
E = analyser bandwidth ratio
Af/f.
band
f
at frequency
fl
5
The tunnel
the
side
wall
working
unsteadiness
of
section
of
ments
approximated
which
would
relate
the
the
tunnel
ally
in
is
also
craft
model
from
the
range
where
of
local
Perhaps
fluctuations
tunnel,
rather
spectra
however
for
example,
working
section
unlikely
mode.
to
assume
to
establish
the
some fixed
Models
quency
and because
frequency
value
This
tunnel,
is
gener-
and
this
and military
to
air-
avoid
and this
means
tunnel
model
that
unsteadiness
As mentioned
wing.
wing
Table
is
above,
free
unsteadiness
of
incidence
incidence
on the
These
waves
planar
in
coming
the
any
peaks
the
datum
level
different
tunnels
for
the
this
the
response
buffet
have
from
centre
of
significant
the
working
not
clear.
line
of
the
may not
fan
or
section8,
tunnel
compressor
because
at
working
tunnel
the
section
the
prediction
the
at
is
is
approximately
low
reduced
predominantly
at
the
wing
to
flow
incidence.
and for
correct
equal
necessary
tunnel
i.e.
c (aircraft)
wing
is
the
hypothesis
approximately
pres-
should
it
of
model
planar
However,
may be compared,
of
remote
spectrum
unsteadiness.
unsteadiness
these
be related
unsteadiness
the
the
although
response
from
of
is
spectra
significant
in
in
be measured,
working
remote
aircraft,
1).
angles
some way.
they
buffeting
tunnel
in
to
for
the
of
of
because
c (model)/fl
(see
where
the
the
should
relationship
full-scale
the
the
tunnel
a Mach number
advisable
with
wind
to
wind
is
measure-
convenient
incidence
fluctuations,
sharp
made in
method3’4*5
fl
angle
generate
the
can be used
as the
for
is
the
measured.
it
of
these
of
on civil
incidence
low
or compressors
tests
that
at
a transonic
response
of
at
are
Hence
fans
that
line
on
slotted
that
It
buffeting
model
pressure
to
with
be used
centre
precisely
angles
between
the
if,
associated
classic
in
related
are
some evidence
However
angles
response
probably
fundamental
bears
model
are
waves
order
low
pressure
sure
of
and bottom
analysis.
in
Mach number
than
the
the
relationship
and the
from
to 0.85
low
top
was measured
excitation.
section
in
at
is
and most
at
the
unsteadiness
importance.
model
or
this
tunnel
severity
critical
that
The precise
not
the
to relate
the
essential
the
see Ref.2,
on the
for
M = 0.75
on the
be below
There
to
of most
selected
speed
tunnel.
is highest
range
the
section
fluctuations
response
the
@nP(n),
working
be preferred
generally
sources
wind
obviously
shocks
is
the
pressure
transonic
it
closed
unsteadiness
is
should
the
function
1
fundamental
the
fre-
.
6
3
DETAILS OF ANALYSIS
Fig.1
shows a typical
curve of unsteady wing-root strain signal at the
fl, plotted against angle of incidence(taken
from
wing fundamental frequency
Ref.1). If these signals are divided by the appropriate kinetic
pressure
to changes in Reynolds number,
we have, if the flow is insensitive
q = 1P v2,
wing-root
strain
signal/q
= CB(M,cO
(1)
where CB(M,a) is a dimensional function of Mach number M which is indepenif the total damping of the
dent of q at a given M and angle of incidence,
Before the onset of flow separations on
wing fundamental mode is constant'.
the model, most of the curves in Ref.1 and numerous tests in other wind tunnels 10
show that CB(M,o) is constant equal to CB(M,o = 0). This is the portion of
at the appropriate
the model response caused by the tunnel unsteadiness
J-Z2
If we assume
Mach number and the same frequency
fl.
CB(M,cr = 0')
=
KJnF(n)
then
(2)
C;(M,a = 0')
where
CA(M,c( = 0')
= ; CB(M,a = 0')
is dimensionless
and
l/K
= G(n)
is a scaling
This scaling factor is different
for every model.
It
of
and stiffness
distribution
of the model, the sensitivity
and the total damping in the fundamental mode. Unfortunately
are often not quoted in many buffeting
experiments.
If the
factor
l/K is applied to the coefficient
CB(M,a = 0) for
numbers the dimensionless
model-response
C;](M,a = 0)
,
I
factor.
depends on the mass
the strain gauges
these factors
same* scaling
all other Mach
can be directly
to the tunnel unsteadiness JnF(n).
If the scaling factor
to the coefficients
CB(M,o) above buffet-onset,
curves of
compared
l/K
is also applied
Ci(M,o) are obtained.
Fig.1 shows a typical
example. The level
C;(M,cr = 0) represents the tunnel
unsteadiness and the model response to that unsteadiness.
The subsequent
increase in Ci(M,a)
as the angle of incidence increases gives a measure of
the integrated pressure fluctuations
arising from the wing buffet pressures and
*
It is not necessary to assume the same scaling factor
l/K
for all Mach
numbers. However the assumption of a constant scaling factor provides a severe
test of the hypothesis and does effectively
allow the best overall match between
the tunnel unsteadiness and the model response over a range of Mach numbers.
7
of the model response to this excitation.
Having used the tunnel unsteadiness
to establish
a datum buffeting
scale, this signal must now be subJnF(n)
tracted to give the true buffeting
level in the absence of tunnel unsteadiness.
If the tunnel unsteadiness does not exceed the criteria
in Ref.2 there should be
no correlation
between the tunnel unsteadiness and the wing buffeting
and so
we can calculate a corrected buffeting
coefficient
C;;(M,a)
=
C;(M,d2-
C;@i,a = 002)
-
(3)
The angle of incidence at which Ci(M,a.) first
differs
from zero is buffet
onset.
Contours of buffetingcoefficients
arethen readily obtained as a function of Mach number and angle of incidence or lift
coefficient.
Fig.2 shows a fair correlation
between CA(M,a = 0) and JnF(n)
for most of the models discussed, implying that a simple relationship
between
the model response and the tunnel unsteadiness is justified,and
that the
damping of the wing fundamental mode does not change much with Mach number.
The wind tunnels used include 3 with slotted working sections and 2 with
perforated working sections.
The poor correlation
between Cb(M,u = 0) and
and implies
JnF(n)
f or model D in the RAE 3ft x 2.2ft tunnel is exceptional
that the sidewall pressure fluctuations
in this case were grossly misleaditig
as regards fluctuations
in angle of incidence on the centre line.
When these
tests were repeated recently in the larger ARA tunnel, where the tunnel unstsadiness on the centre line
CA(M,(r = 0) and JnFo
is known precisely,
much better correlation
of
was obtained; these curves are also shown in Fig.2.
Indirect verification
of the hypothesis is provided by some recent
measurements on a research configuration
with a high aspect-ratio
wing of
leading-edge sweep A = 27'.
Two models of similar
scale were tested at the
same Reynolds number with a slightly
different
transition
fix, having different degrees of structural
damping, different
wing frequencies (90 Hz
and 140 Hz) and thus different
values of the reference unsteadiness,
J=G.
Both models gave almost identical
buffeting
contours (Fig.3).
4
COMPARISONOF MODELBUFFETING CONTOURSAND FLIGHT PENETRATIONBOUNDARIES
Figs.4 to 12 show contours of buffeting
coefficient
Ci versus lift
coefficient
CL for 9 aircraft
models from the flight
buffet onset to maximum
penetration
boundaries.
Table 1 derived from these figures shows Cg appropriate to maximum penetration
and comments on the influence of Reynolds number
from the buffet onset and the development of flow separations.
For the transport aircraft
models (A, B) the buffeting
limit corresponds with
Ci = 0.006.
I G,
For the fighter
aircraft
models (C to J) the heavy buffeting
than for transport aircraft
and corresponds with
C;; = 0.012 to 0.016
limit
is higher
.
.
For the fighter aircraft
there is considerable scatter from the flight
Hence for fighter
aircraft
onset boundary to the Ci = 0.004 contour.
following
buffeting
criteria
are suggested:
Buffet onset
Light buffeting
Moderate buffeting
Heavy buffeting
5
C;;
C;;
Cg
C;;
=
=
=
=
buffet
the
0
0.004
0.008
0.016.
SIGNIFICANCE OF BUFFETING COEFFICIENTS
The order of magnitude of the fluctuating
normal force coefficient,
CN
ms, on a model with a high aspect ratio unswept wing due to flow separations
ought to be given by the corresponding fluctuating
pressure-coefficient,
which
should in turn be of the same order as Ci, i.e.
CN rms
= O(C;;)
= O(O.016)
.
(4)
The total rms normal force contours for 7 NACA aerofoils
under widely differing
6
buffet conditions
are given by Polentz
and appear to satisfy equation (4) as
the following
table shows*.
Polentz buffet
intensity
Onset
Light
Moderate
Heavy
*
Polentz
took buffet
Present buffeting
criteria
% rms
C;;
0.005
0.010
0.020
0.040
0
0.004
0.008
0.016
onset as
CN rms = 0.005,
nel unsteadiness signal and the maximum discrimination
tunnel unsieadiness level for these tests apparently
specified
in Bef.2.
the same level
as the tun-
of his measurements.
satisfied
the criteria
The
9
Similarly
Fig.13 shows that the maximum fluctuating
normal force,
CN rms, at a low frequency parameter,on a series of delta wings at vortex
= 0.008 to 0.010 for
h = 45' to
breakdown conditions'
varies from
= 0.014 to 0.019 for
A = 7o". Hence equation (4) appears to be
GG)
valid for these two extreme classes of wings with different
separated flows;
the heavy buffeting
limit can thus be given a general physical significance.
It is interesting
to note that the light buffeting
limit suggested for fighter
aircraft,
Ci = 0.004, is of the same magnitude as the maximum pressure fluctuations associated with an attached turbulent boundary layer (O.OOZ< nF(n)< 0.003).
JDISCUSSION
6
The correlations
established between buffeting
contours and maximum flight
penetration
in Figs.4 to 12 are surprising
because it might reasonably be
expected that the severity of buffeting
in flight
would be based on the dimensional level of vibration
(either estimated by the pilot or measured by an
rather than a dimensionless buffeting
coefficient.
accelerometer),
There are
two alternative
explanations
for the correlations
established.
Either
(1) the severity of wing buffeting
is not really the limiting
factor
so that the pilots of fighter
or strike aircraft
tend to fly right up to a
handling boundary, such as pitch/up
(as on aircraft
E and F) or stalling
(as
on aircraft
C). This handiing boundary might coincide with the heavy buffeting contour.
Or,
(2)
'q'
factor,
the pilot
may instinctively
include
as he tends to do in the application
in his assessment buffeting
a
of steady loads to the aircraft.
If he does introduce a 'q' factor, pilot-defined
boundaries for light,
moderate
and heavy buffeting
at constant altitude
would be uniformly spaced above the
buffet onset boundary where Mach number effects are small, and correspond with
constant values of pressure-fluctuation
coefficients
measured in the tunnel
and hence of buffeting
coefficients,
Ci (cf. Fig.14 for the Venom aircraft
with the sharp-leading-edge,
Ref.11).
The pilots
of transport
aircraft
generally
sit
further
from the nodal
points of the wing fundamental mode than pilots of fighter
or strike aircraft
and would not wish to approach a handling boundary, even if sufficient
thrust
was available.
Thus for transport aircraft
the maximum penetration
coefficient
C;; = 0.006 seems more reasonable than the value of 0.016 for fighter
aircraft.
This limit for maximum buffet penetration
for transport aircraft
of C; = 0.006
is based on measurements on only two models and may need to be revised as
additional
tunnel/flight
comparisons become available
for this class of aircraft.
10
Although the present hypothesis is believed to offer a quick estimate of
buffet penetration
limits
for transport and fighter
type aircraft
from tunnel
measurements of unsteady wing-root strain,
the method of Refs.3, 4 and 5 is
Both these methods
still
needed to compare tunnel and flight
buffet loads.
must assume the absence of significant
Reynolds number effects from model to
full scale, although there is no doubt that the complex mixed flows which
generally appear during buffet penetration
are likely
to be somewhat scale
.
.
sensitrve,
particularly
at transonic speeds. In addition both methods assume
that in flight
the buffet manoeuvre is stabilised
(or steady in a statistical
Se*Se).
In fact this can be rarely true for the buffet penetration
of transport aircraft
and is quite difficult
to achieve even on fighter
aircraft.
7
CONCLUSIONS
Buffeting
coefficients
appropriate to the maximum flight
penetration
of
wrng buffeting
for both transport and fighter
type aircraft
have been deduced
from the comparison between flight
observations and tunnel measurements of
The buffeting
coefficients
thus
unsteady wing-root strain using models.
deduced may now be used for predictions
of buffet penetration
limits
for
These predictions
will be particularly
useful during comparfuture aircraft.
tive tests for projects with alternative
wing designs.
The necessary buffeting
coefficients
are derived rapidly from the
unsteady wing-root strain measurements. The tunnel unsteadiness (which must
be known) is used as a given level of aerodynamic excitation
to calibrate
the
model response at the wing fundamental frequency; a detailed knowledge of the
structural
characteristics
of the model is thus not required.
Rowever, as in many other types of test using models significant
scale
effects can occur between model and full-scale.
The use of buffeting
coefficients
as a criterion
also implies a degree of similarity
between the type
of structure
used on the aircraft
involved in the analysis and on the particular
aircraft
under review.
In addition the hypothesis itself
involves many assumptions which are not readily verifiable,
so some care is needed in its application.
.
Table 1
ASSESSMENTOF BUFFET PENETRATIONCRITERIA
(Hz)
-I-Wing frequency
Aircraft
type
Model
(fl
(fl
Transport
Fighter/Strike
F
G
H
I
(UnpuhJlished)
f,
c) m
c) a
Reynolds number effects
C"B for maximum
flight
penetration
Buffet
onset
287
360
1.26
1.48
0.004-0.006
0.006
Yes - small
Yes - small
580
260
526
1.42
0.84
1.50
0.014
0.014-0.008
0.012
NO
NO
NO
665
204
283
323
155
1.15
1.58
0.75
1.50
1.22
0.006
0.012
0.010
No heavy buffeting
0.016
Yes
Yes
Yes
Yes
Yes
-
small
large
large
large
large
Separation
development
Large
Unknown
Unknown
Unknown
Large effect inferred
from pitch-up
differences
Very large
Unknown
Unknown
Significant
Unknown
12
SYMBOLS
c
cB' c;,
c;
lift
CL
rms
l/K
transformation
P
i
q/ = JP v2
R
V
w
(2) and (3)
rms normal force coefficient
wing fundamental
F(n)
G(n)
M
n
(l),
coefficient
fl
JnFcn,
equations
coefficient
normal force
cN
CN
average wing chord
buffeting
coefficients
frequency
factor
Hz
equation
(2)
unsteadiness at frequency
fl = p/q(E) 1
contribution
to -52
p /q in frequency band Af
contribution
to CN rms in frequency band Af
Mach number
flw/V frequency parameter
pressure fluctuation
in tunnel at frequency
fl
rms pressure
kinetic pressure
Reynolds number on c
velocity
tunnel width
analyser bandwidth ratio
Af/f
tunnel
angle of incidence
density
13
REFERENCES
-No.
Author(s)
Title,
etc
1
D. G. Mabey
Comparison of seven wing buffet boundaries
measured in wind tunnels and in flight.
ARC CP 840
(1964)
2
D. G. Mabey
Flow unsteadiness
and model vibration
in wind
tunnels at subsonic and transonic speeds.
RAE Technical Report 70184 (1970)
3
W. B. Huston
A study of the correlation
between flight
wind tunnel buffet loads.
AGARDreport 111 (ARC 20704) April 1957
4
D. D. Davis
W. B. Huston
The use of wind tunnels to predict flight
buffet
loads.
NACA RM L57-O-25, NAC TIL 5578, June 1957
5
D. D. Davis
Buffet tests of an attack-airplane
model with
emphasis on analysis of data from wind tunnel tests.
NACA RM 57-H-13, NACA TIL 6772, February 1958
6
P. P. Polentz
W. A. Page
L. L. Levy
The unsteady normal force characteristics
of selected
NACA Profiles
at high subsonic Mach numbers.
NACA RM A55-C-02, NACA TIB 4683, May 1955
7
P. B. Earnshaw
Low speed wind tunnel experiments
sharp-edged delta wings.
J. A. Lawford
and
on a series
of
ARC R & M 3424
August 1964
a
G. Schulz
On measuring noise of bodies in a flow in subsonic
wind tunnels - Part I.
DLR Report 68-43 (1968)
9
D. G. Mabey
Measurements of wing buffeting
ARC CP 954 (1965)
E. J. Ray
Buffet
R. T. Taylor
a systematic series of wings determined
sonic wind-tunnel study.
10
and static
on a Scimitar
aerodynamic characteristics
NASA TND 5805 June 1970
model.
of
from sub-
14
REFERENCES(Contd)
-No.
Author(s)
11
R. Rose
0, P. Nicholas
Title,
Flight
etc
and tunnel measurements of pressure
fluctua-
tions on the upper surface of thewing of a Venom
aircraft
with a sharpened leading-edge.
ARC CP 1032 (1967)
100
Model C
M=0*60
R=l.3 x 106
f, = 580 Hz
Wing - root
strain signal
P”
.1-1
Onset
I
2
d
0
I
4
Original
I
6
I
8
d
IO’
measurements
cl*025
Cg
(M, c$=
Wmg-root
cl8
(M, c() =
‘/K
stroln
C’B
Where
jn=
0.020
‘/K
C8 (M, ac>
C,
(M,
signal
/ 7
0
1
d=O)
O*OlS
o-01 0
3.
0.005
I c I
0
2
b
Fig.laeb
Transformed
Definition
I
4
6
I
8
measurements
of buffeting
coefficients
d
0
0 002
HS
287
C’8,,JnFo
2ftx2ft
Hz
tunnel
08
Model
0 002
%,
-HS 2 5ft
f=360Hr
7 JnFo
0
x2ft
IO
A
w,,$F@
tunnel
I
M
f
,
/I
I
0.2
06
04
Model
08
M
IO
M
IO
B
0 008
RAE 3ft
f =500
6,
x3ft
Hz
tunnel
,W
0 006
0004
I-
0 002
o-
I
I
02
04
-
_
Model
1
I
06
08
C
Fig. 2 Variation
of Clgo and
Mach number for different
I/%@
with
models
ARA
Bft
f=262
0.002
X9ft
Hz
-
G’Bo
e.x
tunnel
n
02
0.4
n
0.6
Model
00
M
0.8
M
I.0
D
0 008
RAE
3ft
f =260
C’B,,JS
0004
a
0.00’
4
X 2.2 fC
Hz
tunnel
o-00: 2
02
04
06
Model
Fig. 2 cant
Variation
Mach number
xof
for
D
C’g,
and J
different
models
with
0 004
RAE
3ftX2.2ft
f =665
Hz
tunnel
C’Bo 1J-m
0 002
-
02
04
06
Model
0.3
M
IO
I
08
M
IO
08
M
IO
F
0.006,
ONERA
f=204
C’B, ,W3
6ft
tunnel
HZ
I
0.2
0’
I
04
I
0%
Model
G
0 006
RAE
6,)
JnF(n>
0 004
3ftX2
f =323klr
2ft
tunnel
-
02
04
06
Model
Fig. 2 concld
Mach
Variation
number
I
of Clg,
and m
for different
models
with
0
Ons&
--+-23 5 In
Model
span
33 In
Frequency
140 Hz
1st
mod6 90 Hz
wing
Wing sweep
(LE)
27’
6.5
Taper
ratlo
I.8 I
Test
Reynolds
15x10=
number
Fig. 3
Contours
of buffeting
of differing
for two
scale
similar
models
CL
t
=N
+-
Flight
onset
‘+
--
-+
-----
hoXrth.
Flight
Rc =40X10=
.~
to
5oX\o=
t‘
^-c-c
+-
I
05
I
04
Fig. 4
I
07
I
06
Aircraft
\
A:
contours
I
08
of
buffeting
I
03
M
I
IO
0 Buffat
l
onset
Maximum
penfztrotlon
0
I
Y
n
I
04
Fig.5
I
05
Aircraft
I
0.6
B : contours
I
07
of
I
08
buffeting
M
I
OS
IO1
I
Moxlmum
flight
penetration
0
V
A
I
04
Fig.6
Flight
RE=30X10e
to
40~10~
\
\
I
05
I
06
Aircraft
C : contours
I
O-8
I
07
of
buffeting
I
OS
M
IO
’ 0
Flight
CN
35000
onset
Llmlting
to
+
x
4OOOOCt
oa0 014
0
V
A
I
04
Fig.7
I
05
Aircraft
I
06
D : contours
I
0.7
I
00
of
buffeting
I
09
M
IO
Flight
onset
=-z
-FT
Tunnd
0 RZ = I 2XlcP
0 RE =0.9x10”
Fig.8
_.
Aircraft
E : contours
-
of
buffeting
I c
Fl\ght
RE = 24X10”
to
56~10~
CL,Cl
0.1
0.1
CT---%-
Cg
-\
-h
-
I
0.006
0 004
0
RC=l
0X10’
0
I
05
I
0.4
Fig.9
Aircraft
I
06
I
08
I
07
F: contours
of
buffeting
I
09
M
I
IO
Flight
Manoeuvre
llmlt
RE=12x10e
to
35x10”
\
\
\
\
\
Flrlng
Ilmlt
\
\
1,
\
\
\
--
-
~0006
Tunnel
RE=2-5xIO=
,/?
I
03
I
04
Fig.10
o-5
Aircraft
06
G zcontours
,
,
07
0.0
,
09
M
of buffeting
.
”
Flight
RE = 25x IO6 to
Disagreeable
buffeting
38x IO6
/
/
/
0.6
onset
--
0.4
0.2
Tunnel
RC = 0.9 x IO6 to
01
V
A
1
0.4
Fig.11
I.2 x IO6
I
I
I
O-5
0.6
o-7
Aircraft
H : contours
of
I
I
0.0
0.9
buffeting
M
I.0
0.0 e
-
C
CL ,tcN
RE =2.4 x IO=
Q.2 -
Fig.12
Aircraft
I : contours
of
buffeting
0 OIB-
Gq-
;i
0 016F\&
upper
Frequency
parameter
n=f.t/V=O
0 014-
surface
05
I
Earnshaw
and
LowFord
0 012-
Uef
/-
7
-\,76
:
o
\
:
I
0.01 o:
I
0 oc)8
-
0 oc )6
-
0 oc )4
-
0.oc
II
*I I:
)2 -
5’5’
I
O-
normal
6’Oe
I
^^
cv
$5’
Fig.13 Variation
force coefficient
I
-- j -4 C
-f f vortex
7b0
I
7h”
breakdown
I
_^
4”
#
of fluctuating
with angle of incidence
c O0
0.018
Jnc(n>
0.016
Convex upper surface
o-014
Ref.7
O-010
h=45O
,f '\
!
I
0.000
I
I
/
O-006
O-004
0.002
Fig. 13 ’
\
:
'
; i
\
I
I
I
I
I-0.27
Rose, Ref. II
CL,cN
O-8
Flight
Pilot
R = IOx106 to 20x to6
rating
0.6
0.4
o-2
0
I
I
I
I
I
04
o-2
03
o-4
0*5
Fig.14
Venom aircraft
with
I
O-6
sharp
I
leading-edge
I
AN HYPOTHESIS FOR THE PREDICTION OF FLIGHT
PENETRATION
OF WING BUFFETING FROM DYNAMIC
TESTS ON WIND TUNNEL MODELS
3IWNAa
.LHOl7d
aw j nvusav
.t!
. .
sI3aoyY xuuuu
ClNlMNOSJSLl
A0X.d tX?I.LmU
3NM JO NOLLVXUNBd
30 NOLI.XI?Bd
Z3H.L X03 SISBlUOdAB
NV
mvmvl3a
‘I
.
1
,
_-I--
,
-_._
-
.__A_
- ----
,-
---
^-;_,_-.-
- _-_
- _+..,_ _
..
----,-
I
I
,
I
_._ _ _ ___
_-------
--
_ _-_ _.-___-.- ._._ --
--
^ - -..
-
C.P. No. I 171
HER
MAJESTY’S
Pubhshed
by
STATlONERY
OFFICE
To be purchased
from
49 High Holborn,
lxindon
WC1 v 6HB
13a Castle street,
Edl”b”rgh
EHZ 3AR
LO9 St Mary Strcot, Cardiff
WI
11W
Brazennow
Street, Manchester
M60 8AS
50 I-aafax
street,
Bristol BSI 3DE
258 Broad Street, Bnungham
B1 2HE
80 Chxhester
Street. Belfast BT1 4JY
C.P. No. I I71
SBN 11 470439 2