ntrs.nasa.govsharp leading-edge extension rese~rc~ model and leading-edge devices 60-deg delta wing...
TRANSCRIPT
'.,)
"
NASA Contractor Report 165923
;(/1j5N -sf< -> / &5; 9~:3
NASA-CR-165923 19830011443
Subsonic Balance and Pressure Investigation of a 60-Degree Delta Wing With Leading-Edge Devices
Stephen A. Tingas Dhanvada M. Rao
MAY 1982 NCCI-46
NJ\SJ\ National Aeronautics and Space Adminisiration
Langley Research Cooter Hamptor., Virginia 23665
!! "-."
Ii "\" n rt:{ f\ i1'""'lI'1\ijP ,f"I; A1' """''' n E tl "1 i'..,I, ,.I J 11' ,Lv \Vi 'I U ,: '" 11\\/1' 'I ,~ \I"" ~. \\ i ,;=..~, n r. 1t i( lw .J ~ t! tbUWO ~b'=~li U U ~Vg n ;fc I' !i 11 : J " . ~ N H ~ 4 ~ d h a.; ~0f;r:!; 11 ,h l rtJJ ,J'~:: ~ ~ " D a L' 0 ~ ,; j.~ n i! NASA LANGLEV !1 II "j;".Crr:"'~'N~~-0't\il n ~MPJ~~\'/J )f . ~:';j~~r~~~,;~:W~:t-t~+:~;" =~'
~--- .
1111111111111 1111 11111 /11/1111111111111111111 NF01927
https://ntrs.nasa.gov/search.jsp?R=19830011443 2020-03-23T19:43:26+00:00Z
-,
111111111111111111@~IIJrlrlllrl~i rj~lli~1111111111111111 3 1176015038186
LIST OF TABLES
LI ST OF FIGURES
LI ST OF SYMBOLS
INTRODUCTION
BAC~GROUND .
Chordwise Slots
Fences .. . .
TABLE OF CONTENTS
Pylon Vortex Generators
Leadi~g-Edge Vortex Flaps .
Sharp Leading-Edge Extension
RESE~RC~ MODEL AND LEADING-EDGE DEVICES
60-Deg Delta Wing Model
Slot Contours
Fences
Pylon'Vortex Generators.
Leading-Edge Vortex Flaps
Sharp Leading-Edge Extension
WIND-TUNNEL FACILITY
EXPERIMENTAL PROGRAM
DATA REDUCTION . . .
PRESENTATION OF DATA
Page
iii
iv
. . . . x
1
5 ~";,
5
6
8
9
11
13
13
20
. .~.:~. 22
22
22
29 ~ i
31
32
35
40
~f3-!97/(:#·
RESULTS AND DISCUSSION
60-0eg D~lta Wing
Chordwise Slots .
Fences
Pylon Vortex Gener~tors
Sharp Leading-Edg~ Extension
Leading-Edge Vortex Flaps
CONCLUS IONS
REFERENCES
'':
< .:
ii
Page
42
42
60 ~ ..
76
91
121
144
196
200
;,'
,..".,.\
"
I.
II.
III.
LIST OF TABLES
Pressure orifice locations
Test summary
Vortex generator zero~a drag characteristics
iii
Page
17
33
96
1.
2.
3.
4.
5.
6.
7.
8.
9. - - ---
10.
1l.
12.
13.
14.
15.
16.
17.
18.
~ ;
~~ ~; -;:
.~: . '-::- i..:;
:"'1ST OF FIGURES
Photographs of -research model and l~~.di ng-edge devi ces
Drawing of 60-deg delta wing research model
Drawing of slot contours tested
Drawing Of' fences tested . . .
Drawi ng of pylon vortexgenerato,.rs tested
Drawing of pylon vortex generator with missile tested
Drawing of leading-edge vortex flaps tested
Drawing of sharp leading-edge extension tested
Basic wing force and moment characteristics - - .-~ -- ---.- -
Basic wing leading-edge static pressure variations and CPLE derived separation boundary
Static pressure distributions around the basic wing leading edge ........ -.. .
Basic wing spanwjse leading-edge thrust distributions
Basic wing leading-edge thrust characteristics
Basic wing induced drag characteristics
Comparison of;b~~ic.wing experimental data with VLM-SA theory . . . . . . . . . . . . .
Basic wing leading-edge static pressur~-thrust relationship ... '.' .. ; ...... .
Force and moment characteristics of single and multiple chordwi se slot confi gura ti on s . . . . .
Local static pressure effects of the chordwise slot at a = 160 ........ .
19. Local ~uction effects of the chordwise slot
.. :. .
~
iv'
'P~a:ge=-
14
16
21
23
24
25
26
30
43
46
47
51
53
55
57
58
61
63
65
20. Spanwise leading~edge static pressure distributions for single and multiple chordwise slot configurations
21.
22.
23.
Effects of single and multiple chordwise slots on leading-edge thrust ....... .
Effect of multiple chordwise slots on leading-edge static pressure-thrust relationship ..... .
Effects of internal contouring on chordwise slot performance .. . . . . . . . . . . . . . . ;.' .
24. Force and moment characteristics of single fence and slot-fence configurations
25. Local static pressure effects of the fence at a = 160
26. Effects of the fence and slot-fence combination on leading-edge separation (C PLE derived)
27. Effects of the fence on leading-edge thrust
28. Induced drag characteristics of single fence and slot-fence configurations .....
29. Effect of the fence on leading-edge static pressure-
" Page
66
67
70
73
77
79
81.
82
84
thrust relationship ................... ~ 86
. 30. . Effects of slot-fence combi nati ons on 1 eadi ng-edge
31.
32.
33.
34.
thrust . . . . . . . . . . . . . . . . . . . . .
Effects of slot-fence combinations on spanwise leadingedge static pressure distribution at a = 160
....
Force and moment characteristics of single and multiple pylon vortex generator configurations
Effect of the pylon vortex generator on leading-edge separation (C pLE . derived)". . . ...
Local static pressure effects of the pylon vortex generator . . . . . . . . . . . . . . . . . . .
35. Effect~ of single and multiple pylon vortex ~enerators
88
89
92
98
99
on leading-edge thrust. . . . ............ 101
vi
Page
36. Spanwise le"ading-edge static pressure distributions for single and multiple pylon vortex generator configurations at a = 140 . . . . . . . . . . . . . . . . . 103
37. Effect of the pylon vortex generator on leading-edge static pre~sure-thrust relationship ............ 104
38. Performance effects of pylon" vortex gener"ator 1 eadi ng-edge length reduction ., . . . . . . . . . . . ... 108
39. Outboard stati~ pr~ssure effects of pylon vortex generator )eading-edge length reduction . . . . . 110
40.
41.
Performance effects of pylon vortex generator chord reducti on . . . . ." . . . . . . . . ". . . . . . .
Effects of pylon vortex generator chord reduction on leading-edge thrust ........... .
42. Performance effects of py16nvortex generator
112
114
diagonal cutback . . . . . . . . . . . . . . . . . . . 116
43. Effects of pylon vortex generator diagonal cutback oh leading-edge thrust . . . . . . . . . . . . .. . . . . 117
44. Performance of an extend~d chord pylon vortex generator utilized as ~ carrier of slender external stores ...
45. Force and moment characteristics of a sharp leading-edge
46.
47.
48.
extension configuration .......... .
Effects of the sharp leading-edge extension on static pressure distributions around the leading edge at a = 120 . . . . . . . . • . . . . • . • • .
Leading-edge suction effects of the sharp leadingedge extension . . . . . . . . . . . . _ . .
Effects of the sharp leading-edge extension on static pressure distributions around the leading edge at a = 220 • . . . . • . . • • . • • . • • . .
49. Effects of SLEE extension on spanwise distribution of leading-edge thrust ..... .
50. Effects of SLEE extension on performance
1 1 " ll~
122
126
128
130
131
133
~
vii
Page
51. Effects of SLEE extension on static pressure distributions around the leading edge at a = 120 . . . . . 135
52. Effects of SLEE extension on leading-ed~e thrust 136
53. Performance comparison of the fence and chordwise slot at the SLEE apex . . . . . . . . . . . . . . . . . . . . . . 137
54. Performance effects of the fence and chordwise slot in combination with the SLEE . ..... . . -... 139
55. Upper surface tuft photographs of 0.48 cm ext. SLEE with F-2 at n = 0.25 configuration with and without an F-2 fence at n = 0.625 (a = 140 ) ..•...•.
56. Effects of the fence and chordwise slot on leading-edge thrust distribution along the SLEE ...
57. Leading-edge static pressure variation and flow development on a 300 + VF-3 configuration
_ 58. Side tuft photographs of 300 + VF-3 configuration at a = 60 and 100 . • . . • • • . • . . • • • . -. • •
59. LocaJ design point distribution (C PLE 30 + VF-3- ............ . derived) along
60. Side tuft photographs of 600 + VF-3 configuration at
· 140
· 142
· 145
· 147
· 149
a = 140 , 160 , 180 , and 230 . . . . . . . . . . . 150
61. Effects of 300 + VF-3 on wing leading-edge thrust 152
62. Effects of vortex flap deflection angle on force and moment characteristics. . . . . . . . . . . . . . 153
63. Effects of vortex flap deflection angle on longitudinal stability characteristics . . . . . . . . . . . 156
64. Comparison of 30° + VF-3 experimental data with VLM-SA theory . . .. ....... 158
65. Thrust contribution from 30° + VF-3 flap surface 159
66. Upper surface tuft photographs of 300 + VF-3 configuration at a = 140 and 170 .... 161
vii i
Page"
67. Effects of upward vortex flap deflection on force and moment characteristics. . . ..... 164
68. Upper surface tuft photograph of 30° t VF-7 "configuration at a= 6° ••.••. ~. . . . ... 166
69. Effects of "inverse" tapered flap chord distribution and apex shape on vortex flap performance . . . .. . ... 168
70. Effects of inverse tapered vortex flap chord distribution on static pressure distributions around the leading edge at a = 120 . . . . . . . . . . . . . ,.. .. 169
71. Upper surface tuft photographs of 30° f VF-2 and VF-3 configurations at a = 160 .............•... 171
72. Effects of inverse tapered flap chord distribution on local design point distribution (C PLE derived) along the vortex flap .......... ~ ...•....... 172
73. Effects of inverse tapered vortex flap chord distribution and apex shape on flap thrust parameter .......... 173
74.
75.
76.
77.
Effects of segmentation and flap geometry on vortex fl ap performance (6 = 300 ) . . ... . . . . . . . . . . .
Effects of segmentation and flap geometry on vortex flap performance (6 = 450 ) ..•.•..••••.•• .
Effects of segmentation and vortex flap geometry.on flap thrust parameter
Side tuft photographs of 300 f VF-3, VF-4, VF-5, and VF-6 configurations at a = 140 .
· 175
178
· 179
• 180
7S. Upper surface tuft photographs of 300 f VF-3, VF-4, VF-5, and VF-6 configurations at a = lSo .....•.... lS3
79.
so.
Effects of segmentation and flap geometry on local design point distribution (C PLE derived) along the vortex flap ............... .
Effect of differential vortex flap deflection on rolling moment characteristics ....... .
· lS4
· 186"
Sl. Performance effects of the fence and chordwise slot in combination with the vortex flap . . . . . .... lS8
82.
83.
84.
85.
Side tuft ph6tographs of 300 + VF-3 configuration with and without a slot in the flap at n = 0.625 (a = 14°) .... . . . . . . . . . . . . . . . . .
Effect of a chordwise slot in the vortex flap on spanwise local design point distribution (C PLE derived) ....
Effect of a chordwise slot in the vortex flap on flap thrust parameter ...... .
Comparison of sharp leading-edge extension and vortex flap performance . . . . . . . . . . .. ....
ix
Page
. . 190
191
. 192
195
x
LI ST OF SYMBOLS
A - area, cm2
AR aspect ratio
a length of pressure panel along leading edge, em
b span, cm
BW basic wing
CA axial force coefficient, Axial force/qroSref
CAo axial force coefficient at a = 0°, Axial force/qooSref
Co drag coefficient, Drag/qooSref
COo drag coeffieient at a = 0°, Drag/qooSref
CO i induced drag coefficient, Induced drag/qooSref
CL lift coefficient, Lift/qooSref
C1 rolling moment coefficient, Rolling moment/qooSrefb
Cm pitching moment coefficient, Pitching moment/qooSrefc
Cml pitching moment coefficient about original moment reference center indicated in figure 2, Pitching moment/qooSrefc
CN normal force coefficient, Normal force/qroSref
CNo
Cp
normal force coefficient at a = 0°, Normal force/qooSref
static pressure Coefficient, (p - p )/q ro ro
Cs suction force coefficient, Suction force/qooSref
CT leading-edge thrust coefficient, Thrust/qooSref
Cx fl ap thru st pa rameter, /m2
cr root chord, cm
c mean geometric chord, cm
cg center of gravity
"
o drag, N
e spanwise efficiency factor
ext. extension perpendicular to leading edge, em
F fence
K
L
L. E.
9v
PO
P
Poo
qoo
S
SC
SF
SLEE
STA
51
T. E.
Voo
VF
VG
VLM-SA
X
XI
induced drag parameter, lie
1 ift, N
leading edge
semispan leading-edge length, cm
percent drag redu~tion
static pressure, N/m2
free-stream static pressure, N/m2
free-stream dynamic pressure, N/m2
planform area, cm2
slot contour
suction force, N
sharp leading-edge extension
spanwise pressure station
chordwise slot in vortex flap
trailing edge
free-stream velocity. m/sec
leading-edge vortex flap
pylon vortex generator
Vortex Lattice Method with Suction Analogy code
chordwise distance from wing leading edge, cm
distance from vortex flap apex along wing leading edge, em.
xi
xii
x chordwise distance from wing apex, cm
Y' perpendic~lar distance from wing leading edge, cm
y spanwise distance from plane of symmetry, cm
z vertical distance fro~ chordline, cm
a angle of attack, deg,
aD departupeangle of attack, deg
8 ~ortex flap deflection angle, deg
Aleading~edge sweep angle, deg
n spanwise coordinate in fraction of semispan
Subscripts
avg average
cm center of moments
des local design point
front frontal
LE 1 ead i ng edge
loc 1 oca 1
min minimum
ref reference
sep _.leading-edge sepa~atio~
tot total
INTRGDUCTION
Modern aircraft designed for supersonic flight are often fitted
with thin, highly swept delta wings to take advantage of reduced super
sonic wave drag. At high subsonic speeds, where goo~ maneuverability
and high accelerations are desirable, thin delta wings provide added
advantages. These include high drag divergence Mach number and high-g
maneuyerability due to increased lift generated by stable leading-edge
vortices. These vortices, which result from leading-edge flow separa
tion, and subsequent flow reattachment on the upper surface produce the
added advantage of sustained performance capability at high angles of
attack. However, the large induced drag penalty characteristic of delta
wings in high-lift flight limits the aircraft's maneuver capability at
high subsonic and transonic speeds. This increase in the induced drag
is due to early 1eading~edge flow separation, which causes a total loss
of the leading-edge suction associated with attached flow around a blunt
leading edge. Consequently, excess engine thrust must be used to
counteract the drag increase; reducing the acceleration capability of
the aircraft. In addition to the drag penalty, the spread of vortex
origin from the tip to the apex of the wing with increasing angle of
attack produces severe longitudina~ instability in the mid-a range, as
the center of vortex lift moves ahead of the wing center of gravity
(ref. 1).
One method of alleviating the induced drag and stability problems
associated with highly swept delta wings is through leading-edge flow
2
control to delay the onset of leading-edge flow separation. Attached
flow ata blunt leading edge all~ws for partial recovery of leading
edge suction, providing a reduction in drag and alleviation of the
longitudinal instability resulting from leading-edge vortices. The use
of a cambered leading edge has been one method of maintaining attached
flow to higher angles of attack. However, severe drag penalties at
supersonic c~uise speeds and the added weight and mechanical complexity
involved with the use of a variable camber leading edge make this method
less desirable. Leading-edge devices such as flaps and slats have also
been used for the purpose of maintaining attached flow; however, they
have not been overly successful on highly swept wings (ref. 1). Fixed
leading~edge devices such as pylon vortex generators, fences, and chord
wise slots (notches) have previously been used for alleviation of sta
bility problems of highly swept wings. However, recent research has
shown that these devices also have potential for drag reduction by
maintaining attached fl ow to hi gher angl es of attack and,thus, seem
to bea practical solution to the induced drag problem (ref. 2).
An alternative to the conventional approach of drag reduction
through attached flow is accomplished through the use of leading-edge
vortex flaps and sharp leading-edge extensions (SLEE) (refs. 1,2,3,
and 4). These devices utilize the nat0ral tendency toward separation by
forcing vortex formation and using the resulting suction forces for drag
reduction. The downward-deflected vortex flap generates a coiled vortex
whose suction force acts directly on the forward face of the flap, pro
ducing thrust and lift force components. The sharp leading-edge
"
:l.
extension makes use of the same typ\:: of induced vortex; hO\tJf:ver, by
keeping the vortex directly ahead of the leading edge, the suction
effect produces strictly a thrust component and, thus, a reduction in
drag. The flow mechanisms of the vortex flap and sharp leading~edge
extension are shown below:
Leadi ng-edge vortex flap Sharp leading-edge extension
Sketch A
3
This report presents the results of a wind-tunnel investigation
undertaken to examine the potential for further drag reduction through
refined versions of devices such as fences, chordwise slots, pylon
vortex generators, leading-edge vortex flaps, .and sharp leading-edge
extensions. Since previous research had established the effectiveness
of these devices, the present investigation was primarily concerned with
modifications to overcome their limitations. Methods of reducing the
low-angle-of-attack drag penalty while maintaining the effectiveness of
the devices at higher angles of attack were studied. Results of previ
ous research (ref. 5) were used to design and test certain device com
binations which were believed to have drag-reduction capabilities beyond
4
those of the individual devices. In addition, several novel configura
tionswereincluded to determine the feasibility of concepts sucn as the
use of pyl~n vortex " generators as carriers .of slender external stores
(such as air-to-air missiles) and leading-edge vortex flaps for roll
augmentation and as "drag" devices for landing purposes.
Use of commercial products or names of manufacturers in this report
does. not constitute official endorsement of such products or manufac
turers, either expressed or implied, by the National Aeronautics and
Space Administration.
5
BACKGROUND
This section presents a brief discussion of the basic flow
mechanisms of the devices and results of previous research ~h;ch led to
the present investigation.
Chordwise Slots
Typical high-angle-of-attack performance improvements through the
use of chordwise slots cut into swept leading edges are presented in
reference 5. The slot flow mechanism results from the high ve.]ocity jet
sheet emanating from the slot due to the natural flow of air from the
lower to the upper wing surface (sketch B). It is believed that the
~
Spanwise boundary @ layer flow
---------------" I ---------
Root Leading edge
Sketch B
Tip
vortex shed from the outboard edge of the slot acts to obstruct the
span\lJise boundary layer flow on the upper surface, reducing boundary
layer buildup and, thus, separation tendencies near the wing tip. In
addition, the'sense of rotation of the slot vortex opposes that of the
6
primary vortex,; thereby hindering its inboard movement and growth. It
is thi s "compa)~tmentati on" effect of the chbrdwi se slotwhi ch prompted
its use in combina't;ion' withth'e vort~x flaps and sharp leading-edg.e
extension in the pres~nt investigation. The pri nci pa 1 concern , however,
was alleviation of a sudden loss of slot 'effectiveness at the higher_
angles of attack and a low-a drag penalty arising from pressure drag on
the vertical' face and friction on the internal wetted area of the slot.
Through internal contouring of the device, an attempt was made to reduce
this law-a drag and maintain sloteffectiveriess to higher a. In an
attempt at further performance improvements, a verN limited study of the
effects of slot depth and v:Jidthwas also included.
Fences
The use of fences on delta wings has tra-d-itionally been as a fix
faY' longitudinal instability. Previous research, however, ha's shown
these devi'ces to be'effective in the role of drag reducers at high
angles of attack, aiding in the alleviation ·of severe lift-dependent
drag pena Hi es. Typical performance improvements \,Iith si ngl e and
multiple fences similar to those presently tested are given in
reference 5.
Fence flml/.mechanisms are described in detail in reference 6 and
briefly summarized here. First, the fence forces the swept wing upper
surface isobars, normally parallel to th:e leading eoge, to be ut1swept
locally, reducing suction peaks and pressure gradients outboard, with
opposite effects inboard. Sl,I bsequently, outboard staTl and loss of
7
leading-edge suction are delayed. The adverse effects inboard at'e of
minor concern due to lower prevailing upwash. The fence also acts as an
obstruction to the spanwise boundary layer flow on the upper wing sur-
face, further delaying the onset of separation near the tip. However,
a loss of effectiveness is observed at high a, possibly due to an
accumulation of viscous fluid on the inboard side of the device. It
was believed that the flow through an adjacent inboard slot would IIblmv
off" this viscous accumulation (sketch C) and was, thus, tested here.
Boundary
Viscous accumUlatiOn)
----;-
_____ L~ayer ==;---- \ \111- - ---
Root
. ~
Leading edge
Fence only
Tip Root
1~ Slot-Fence combination
Sketch C
Tip
Finally, the fence impedes the inboard movement and growth of the pri-
mary vortex, allowing for the formation of a second, undisturbed
leading-edge vortex inboard, with both acting primarily on the wing
leading edge. For this reason, fences were also selected in the
present investigation ~or use in combination with the vortex flaps and
sharp 1 eadi ng-edgeexte'ns ion.
Pylon Vortex Benerators
The pylon vortex generator relies on the formation of a vortex
8
originating at its swept-forward upper edge as a result of the effective
angle of attack of the device relative to the wing leading-edge cross-
flow. Except at the lowest angles of attack, this vortex travels over
the wing upper surface and rotates in a sense so as to act as a barrier
to the spanwise boundary layer flow, while at the same time inducing a
downwash velocity outboard, with a subsequent delay in outboard stall
(sketch D). In addition, the rotating motion of the vortex promotes a
Root
/
00':1--/8'''''
Sketch 0
(ref. 5)
"
certain degree of boundary layer energization through turbulent mixing
of viscous boundary layer fluid ~.jith high-momentum fluid from the
9
external stream. Again, the result is increased resistance to separa
tion (ref. 7). However, a drag penalty is paid at low angles of attack
since the leading-edge cross-flow is not yet of sufficient magnitude to
cause vortex formation. The present investigation attempts to reduce
the vortex generator low-a drag through systematic size reductions
(basically, lower and aft edge removal), with minimal sacrifice of
high-a performance. The constant 300 forward sweep and 100 toe~in angle . .
of these devices were optimum values selected on the basis of previous
research (ref. 5). In addition, the charactedstic pylon shape of the
vortex generators suggested their possible use also as carriers of
slender external stores (such as air-to-air missiles); therefore, the
effect of a simulated missile attached to the lower edge of a vortex
generator on its aerodynamic effectiveness was investigated.
Leading-Edge Vortex Flaps
Results of previous tests on leading-edge vortex flaps and detailed
descriptions of their aerodynamic mechanisms are presented in refer-
erices 1 and 3. In essence, the vortex flap relies on the prevailing
upwash ahead of the wing leading edge to force separation and formation
of a coiled vortex whose suction effect acts on the flap, producing
aerodynamic thrust and lift components (see sketch A in INTRODUCTION).
By maintaining this sweep-stabilized vortex on the flap along its entire
spanwise extent, with the flow reattachment position ideally at the
wing-flap junction (knee), attached flow is maintained on the wing
upper surface. Flow entrainment and increasing upwash, however, cause
~
10
the flap vortex 'to grow and rni grate onto the wing surface wi thi ncreas-
ing angle- of attack and outboard di.stance. In an attempt to maintain
the ideal flow condition along the ~nt~re flap span, an inverse tapered
flap was selected in the present investigation for comparison with a
constant chord flap, under the assumption that increasing chord outboard
cou ldbetter :accominoda te the con i cal fl ap vortex. Segmented fl aps of
various planforni shapes (constant chord, parabolic, and inverse tapered)·
were also included to further assist in this matt'er through the forma
tion of a distinct vortex on· each segmeht, each acting primarily on the
flap surface. The use of fences and chordwise slots in combination with
the vortex flaps was based on 'their compartmentapoh_ effect, again with
the segmehtatian of the primary vortex into two smaller, undisturbed
vortices (sketch E). The ability of the vortex flap to modify the wing
without slot
Sketch E
with slot (or fence)
i':t
~~
~
11
spa~wise lift distribution suggested that sizable rolii~g moments might·
be obtained by means of asymmetric flap deflections at high angles of
attack, when other control surfaces are degt'aded by flow separation.
Thi s concept was tested along with the pass i bil ity of adapt; ng vortex
flaps for aerodynamic braklng at landing through appropriate control of
vortex suction forces (sketch F).
Suction
Sketch F
Sharp Leading-Edge Extension
Upward deflection
Large downward deflection
Results of preliminary research on a sharp leading-edge extension
are given in reference 5. This device, operating on the same principle
of forced separation as the vortex flap, derives its drag-reduction
capabiliti~s from a tightly coil~d vortex maintained ahead of a blunt
12
. leading ,edge, utilizing its 5uction:effect to obtain a thrust force
(ref, 8) .. Ideally, flow reattachment .occurs just aft of the wing
leading~e9ge curvature (see~ketch A in INTRODUCTION). However, a down
stream expansion of the vortex cor~ due to flow entrainment leads to
eventua 1 mi gration of the. reatta·chment poi nt onto the wing upper surface
with increasing,~ngle of attack and. outboard distance. The resulting
upper surface separated .nOVY and asso,ciated drag increase act to
partially nullify th~ thrust derived from the suction effect of the
device. The compartmentationeffect Of fences and chordwise slots
located at various positions along the SLEE was utilized in the present
investigation for the purpose of delaying this growth of the leading
edge vortex. In addition, tests were performed to determine the optimum
SLEE extension producing the ideal vortex size and position just
described.
RESEARCH MODEL AND LEADING-EDGE DEVICES
The following is a description of the research model and leading
edge devices used in the investigation. Actual photographs of the
sting-mounted model and selected devices appear in figure 1.
60-De9 Delta Wing Model
13
Figure 2 shows a drawing of the wooden, 60-deg cropped delta wing
model used in the investigation. The flat-plate wing has semi-elliptic
leading edges (ellipse ratio of 26.7 percent) with uniform leading-edge
radius of 0.231 cm.
The right-hand wing panel was equipped with six chordwise rows of
static pressure orifices around the leading edge at approximately the
20, 33, 45, 57, 70, and 82 percent semispan positions. Each station
consisted of four orifices on the upper surface, four on the lower
surface, and one near the leading edge. Thirteen additional orifices
were provided along the span at the leading edge (X = 0) and one on the
wing upper surface near the trailing edge. The pressure orifices were
fed by 0.10 cm outside diameter metal tubing, which was protected by a
removable metal base plate on the lower surface of the wing. ' The
orifice locations are given in Table I.
The research model also had six 5.08 cm long chordwise slots on
either leading edge at approximately the 25, 37.5, 50, 62.5, 75, and
87.5 percent semispan positions. The slots were meant for holding the
,leading-edge devices but also used as "devices" themselves,.being sealed
when not in use.
14
(a) Basic wing. ( b ) Open slot,.
(c) Vortex generator. (d) Vortex generator wjmissile.
Figure 1.- Photographs of research model and leading-edge devices.
(e) Down-deflected leading-edge vortex flap.
(g) SLEE w/fences.
(f) Up-deflected leading-edge vortex flap.
(h) SLEE w/slots ..
Figure 1.- Concluded.
15
Slot (0.094 wide)
6
Moment reference center
z
CHORDWI5E, SLOTS /
SLOT in-(%)
1 26.40 2. 38.68 3 ',51. 04 l\ 63.36
1 5. 3 5. 61 '" ,6 . 87.89
19.05
PRESSURE OR IFI CE'S
CHORDWI SE ...L ROW ' b/2 ('}'o)
J' 19.97 2 ,32.78 , J ;44.79 4 57.21 ' 5 69; 64 . 6; 81. 75
33.02
'-r--'+--'JJ 1.90
RESEAR CHMODEL
Planform area ........................ 3263.9 em 2
Aspect ratio ........................... 1.60 Taper ratio ............................ 0.18 Mean geometric chord ................. 52.18 em Moment reference center .............. X' 32. 38 cn
(from wi ng apex) y • 0 zoO
Figure 2.- Drawing of 60-deg delta wing research model. Dimensions in centimeters.
/ .
.~~".j
TABLE 1.- PRESSURE ORIFICE LOCATIONS 17
no 109 ----
Station 1 ISlA 1) f
106~ I
,...
105 . . 11M
103 102
ORIFICE NUMBER X(cm) y(cm)
101 0 I 6.03
102 4.07 7.29 103 2.38 104 1.41 105 0.51 106 o· 107 0.57" 108 1.46 109 2.45 110 4.34
111 0 8.38 * 112 0 10.69
113 3.94 11.95 114 2.44 .
1 115 1.52 116 0.69 117 0 118 0.56 11.94
-
X
z(cm)
0 I -0.96 -0.77 -0.61 -0.37 -0.02 0.36 0.61 0.78 0.98
0 0
-0.97 -0.79 -0.62 --OAl -0.05 0.44
~ J>y
I \~ L.E.-l I ;'-.
STAi }\ :\,
~ ~\ , j), >\
1 ~\. 1\
L .: '~r LE.-19
I ,~6 J
DESCRIPTION
L.E.-1
Bottom STA 1
! L. E. - 2, ST A 1
Top STA 1
1 LE.-3 L. E.-4
Bottom STA 2
! L. E . - 5, ST A 2
Top STA 2
TABLE 1.- Continued. 18
ORIFICE -. , NU~1BER'- - '>_: X (em):, -: yCem) 'z(cm) DESCRIPTION, '<
119 1. 57 l 0.64 t *' i~g "~~ ~r"- r'" ~:~~ t
,;1"" ,_','
'122 0 12.88 0 L.E.-6 123 ," 0 ,,15:06 0 L.E.-7
124 4.02' 16.28 -0.99 Bottom STA 3 , 125" 2.56 1 -0.82 ! 1~6 1.48 ,,-0.64 127 0;56 ,':, ',..0.40 128 0 '-0.04 L.E.-8, STA 3
", ,129 0.,67 ",16.32" - ,0.42 Top STA 3 130 1.64 1 0.67 ! 131 2AO ,",' 0.80 ' 132 4.28 1.01
133 0 17.28 0 L.E.-9 * 134 0 19.54 0 L.E.-10
201 3.94 '20. 77 ~ -1. 00 Bottom STA 4
202 2.50. '1 -0~82 ! 203 1..50 " -0.66' ' ' 204 0.61 '-0.42 205 ' 0 ,: .,.O~14 L.E.-ll, STA 4 206 0.78 20.84 0.47 Top STA 4
*207 1.5,8 I"!' 0.67 !' 208 2. 42 " O. 81 209 4.32 ' 1.01 '
210 '021: 77 - 0 L. E..., 12 211 ' 0 23.99 0 L. E.-l3. '
212 4.26 25.27 -,1.01 Bottom STA 5
213 ' 2 A8 -- '1 ' .;. 0 .82 l' 214 1.47 ~0.64 '
215 0.59 -0.40 216 0 "0 L.E.-14, STA 5 217 '0.69 ,'25d5 "0.45 Top STA 5
218 1.62 1 0.68 1 219 2~53 " 0:82 220 4.08 " 1.00
221 0 26.21 0 L.E.~15 222, 0 28.44,0 L.E.-16
TABLE I. - Concluded. 19
NUM!)t.K i I\\wll'j ,)'\ ... 1111 l... L\Cmj I Ut~I...Kl!" ! HJN .
I
223 4.23 29.69 -1.00 Bottom STA 6 I 224 2.36
1 .. -0.8b /
! 225 1.38 -0.61 i
226 0.51 -0.38 227 0 0.04 L.E.-17 STA 6 228 0.55 29.79 0.39 Top STA 6 229 1.43
J 0.65 ! 230 2.43 0.81
231 4.28 1.01
232 0 0 L.E.-18 . 233 0 32.85 0 L.E.-19
234 36.07 20.84 1.25 Upper Surface ---~- -------
*malfunction suspected.
20
. The balance (NASA-LRC model 846) used for force and moment measure
ments was. a six-cof)1ponent internally mOtlntedstrain-gage balance. with
maximum allowable 19adsasfonows (accurate to within 0.5 percent of . . . .
these values):
Component Load
Normal' 3113.6 N Axi al 378 .. 1 N. Side 1334.4 N Pitch 197.8 N-m Roll 36.2 N-m Yaw 84.8 N-m
The balance was located on the upper surface of the wing and shielded by
~,fuselage-li~e aluminUm housing to prevent wind interference.
The two scani -val ves .used for pressure readings were equipped wi th
3.45 N/cm2 pressure transducers, accurate tG within 0.017 N/cm2~ and
were located on the lower surface of the wing. They were a1~0 covered
by an aluminum housing identical to the balance housing.
An accelerometer, l'ocated inside the upper nose cone of the wing.
was used to measure angle of attack. It was a pendulum-type strain-gage
unit, accurate to within 0.2°.
Slot Contours (SC)
Details of the internal slot contours of the open slots are shown
in figure 3. With the exception of SC-4, the slot contours were made
from 0.079 cm thick aluminum stock and were inserted directly into the
21
SC-1 8.26 X 3.00 ellipse
Leading edge
SC-2 8.89 X 2.39 ellipse
SC-3 8.89 X 5.13 ellipse
1.78L~ Jr ~ -t- .
1 j ~,--t--1.70
. r----~81~
<~~ ... SC-4 flush with L.E.
SC-4
Figure 3.- Drawing of slot contours tested. Ellipse dimensions given are length of major and minor axes. Dimensions in centimeters.
chordwise slots. 5C-4 was made by forcing 0.159 em thick balsa wood
into the slots and s~nding it flush with the leading edge.
Fences (F)
22
The fences used in the investigation are shown in figure 4. They
were constructed from D.079 cm thick flat~plate aluminum and were also
held in position by the chordwise slots.
Pylon Vortex Generators (VG)
The geometry and dimensions of the pylon vortex generators tested
are given in figure 5. These flat-piate devices were constructed of
0.079 cm thick aluminum and were reinforced by an additional thickness
of 0.159 cm on the outboard side to prevent bending under air loads. An
additional vortex generator, with a'ft extension of the lower edge as a
possible external store-carder (fi"g. 6), was tested with and without a
1.27 cm diameter wooden dowel simulating a sidewinder missile scaled
from the F~4D aircraft~ T~e-~ortex generators were held i~ position by
the chordwise slots.
Leading-Edge Vortex Flaps (VF)
,Undeflected plan views of the vortex flaps, along vlith their corre-:
sponding planform areas, are shown in figure 7. The figure also indi
cates the deflection angles and the semispan ~overage and position of
each flap test configurati6n. The flaps were b~nt from 0.159 cm thick
aluminum and had sharp tapered edges to induce vortex formation. For
mounting, they were bolt~d difectly onto the lower surf~ce of the wing.
1 [ 5.971 +
LV ____ --~--~!~., -1-____ > 7 7 (
L=:60
86 -, t 0.64
==J I . ~3 / t
t-1.27
/' 7 7---r-L
23
~ / (I" -*-0.64
J t l
Figure 4.- Drawing of fences tested. Dimensions in centimeters.
lOa. foeHn
~
2_YQ-~ ___ _
/" VG~2' I· 7 -- ---: - -':"':;-~ ........ --'
VG-l
. 1---:5.59~;
Vortex generator
l-2.79~
I t
L90 ~
100 toe-in
h--4.19 ---l
30 81
1
l 9.14 ~I
24
Figure 5.- Drawing of pylon vortex generators tested. Dimensions . in centimeters.
L.E.
3,81 VG-8 wI missile
VG-9 wI out missile I i
~~ ~~ I. ~ 22.48 ·1
missile
SIDEVIEW FRONT VIEW
Figure 6.- Drawing of pylon vortex generator with missile tested. Dimensions in centimeters.
r
N (J1
Vf-l A =374.00 em2 .
30 - 100% sem i spa n
S "30° t
T ItOO
1-13.87
,--,-----,------'------I' ~
I IS. 62
LJ VF-3
~4S0 ~1 A a 373. 3S em
2 .
?S -100% sell1ispan "~ ~ 3(\0 /,SO 6,,0 I o u, q " v t
~.8l
SO. 52
26
"A" 362.00 em2
30-100% semispan
0"' 30°.
\~1 ~ IljOO
VP'4 2 A = 358.84 em
"25 - 100% sem ,span
8 = 30°, 4So t
Fi gure 7. ~ Drawing of 1 eading-edge . vortex ~ flaps tested. Planform areas given a~efo~ undeflect~d case and include both right arid left flaps. Dimensions in centim~ters~
I 15.62
U VF-5
. 2 A = 312.52 em
25-100% seniispan
0=30°, 45° t
VF-7-A" 329.29 tm
2
25-100% semispan
o • 30° t
T 15.62
U VF-6
A = 206.45 em2
25-100% semispah
8 = 30°, 45° t
VF""'l th rough 6
o .. ==~~/%~ZLL~ ~y .
0.4K' ..
30° t VF-7
r2•621_ --(~
5.0V
Figure 7.- tontinued.
27
I t X'
28
-<
.. YF-5 FlapCoordi nate 5 .
X' y'
0.00 0.00 0.94 0.79 2.92 1.85 7.29 3.00
l3.26 ~.50 1B.90 3.68 27,OB 3~Bl
VF-5 27.0B 0.00 28.45 1.07 31..27 236 36.B3 3.30 44.45 3.73 47.70 ·3.81 . 54.15 Q.OO
.- .
Figure 7.-Concluded.
29
Sha.r.Lle2di ~_-:: EdJL~~!tens tC!~J~lEE)
The sharp leading-edg~ extension tested appears in figure 8: The
SLEE had a variable extension ranging from a to 0.76 em mea~utedperpen
dieular to the wing leading edge. This 0.101 em thick aluminum device
had a sharp leading edge and was bolted directly onto· the lower surface
of the wi ng.
l
· 25-93% semi span
I 13.87
---,-------,I ~ '\ ~·0.094 27.08 ~I SLEE wlSlot at V 62.5% sem i span
SLEE planform
area
cm 2
75.0
'50.0
25.0.
0 0 Q.25 0.50 0.75
SLEE extension (ext.), em
l ext.
Lr .•. C07li 0.1O!< . ~""'''~
SLEE wi Fence (fence extended to SLE~ edge)
Figure 8.-. Drawing of sharp leading"-edge extension tested. Plariform area includes bath right and left SLEE. Dimensions in centimeters ..
30
WIND-TUNNEL FACILITY
The study was conducted in NASA-Langley Research Center's 7- by
10-foot high-speed tunnel. This is a continuous-flow. closed-circuit,
subsonic-transonic atmospheric wind tunnel which operates at ambient
atmospheric conditions.
31
The drive system consists of a motor generator system which powers
a 10.5 megawatt fan motor. The fan motor drives an 18-blade, 9.14 m
diameter fan at a maximum speed of 485 rpm, producing a maximum test
section Mach number of approximately 0.94.
The test section of the tunnel is 2.01 m high and 2.92 m wide, with
a usable length of 3.30 m.
The model support system used in the test is referred to as the
standard angle-of-attack sting. It consists of a vertical strut with
a variable pitch angle sting support system with a range of approxi
mately _1 0 to 230 . In addition to the pitch mode, the standard sting
also has a translation mode which allows the model to be translated
vertically from floor to ceiling, keeping it near the center of the test
section throughout the angle-of-attack range. Reference 9 contains a
detailed description of the tunnel facility.
The data acquisition, display, and control system for the 7- by
10-foot high-speed tunnel is controlled by a dedicated on-site computer.
The system includes a Xerox Sigma. 3 computer, a data acquisition unit, a
line printer, and a Tektronix 4014 graphics terminal. Reference 10
contains a detailed description of the data reduction capabilities of
the system.
32
EXPERIMENTAL PROGRAM
The present investigation was performed at nominal Mach and
Reynolds numbers of 0:16 and 2.0 x 106 (based on a mean geometric chord
of 52.18 cm), respectively~ Force, moment, and surface static pressure
data were taken at angles of attack ranging from _1 0 to 23°. A summ'ary
of the test is presented in Table II. The run number(s) corresponding
to each test configuration is the key to locating the test data pre-
sen ted in reference 11.
Surface flow visualization using a fluorescent tuft technique was
included in the investigation to aid in interpretation of balance and
pressure data. This technique involved the use of 0.02 mm diameter mini-
tufts made of a nylon monofilament material treated with a fluorescent
dye. Approximately 300 mini-tufts of 3.8 em average length were mounted
on the upper surface of the right-hand wing panel and, in some cases, on
the leading-edge devices, using a mixture of three parts Du~o cement and
one part lacquer thinner. The tufts were illuminated by ultraviolet
strobe and photographed through windows in the top and side of the test
section at various angles of attack during a test run. This mini-tuft
technique had been shown in previous testing of this model (ref. 5) to
have a negligible effect on the flow field and, thus, could be performed
simultaneously with force and pressure tests.
In comparison with the model used in the investigation of refer-
ence 5, the present model was basically identical with a few minor
exception~. The most obvious was a reduced fuselage housing in order to
33
TABLE 1I.- TEST SUMMARY
DEVICE IS) SEMISPAN POSITION (%) RUN NUMBER 25 30 ' 50 62.5 75 93 100
Basic Wing 3, 56
Open Slots (no contour) • • • 4 SC-l (short semi-elliQtic) • 47 SC-1 " • • • 36
Slots SC-2 (long semi-elliptic) • • • 7 SC-3 {quarter-elliptic} • • • 8 SC-4 (flush with L.E.) • • • 9 Double Slot Width 10.18 cm) wI SC-1 • 48
rJ~ - • 53
Slot/Fence SC-1 H n' , • 49 50 combination SC-1 F-4 , • 51
SC-1, F-5' ... • 52
VG-1 (baseline) " 40 VG-1 " " " 54 VG-l Ii ~ e e 55 VG-2 (leadi ng-edge Ie ngth reduction) " 43 VG-3 Il " 46 Vortex VG-4 (chord reduction) " 41 generators VG-5 Il " 44 VG-6 (variable chord) • 42 VG"7 Il " 45 VG-8 (extended chord w/ mi ssile) " 38 VG-9 (extended chord wI out missile) 0 39
3no t VF-l' (fu!! 'length, inverse tapered) I
10 -30v t VF-2 Il (tapered apex) 12 30u t VF-3 (full length, constant chord) 11 45
u t VF-3 11 11
60u t VF-3 " 16
30u t VF-4 (segmented, constant chord) 14 Leading- 45u f VE~4 Il 15
edge 30u ~ VF-5 (segmented, parabolic) 17 vortex l4.2". VF-5 "
I
1R 19 flaps 30u t VF-6 (segmented, inverse,tapered) 22
45u J 'VF-6 Il 23 30u ,VF-3 SC-1 SC - 29 300 t VE: 3 £-1 F 30 300 t VF-3 Slot (D. 094 cm) in flap IS') SI" 37 300 t on left' 45u t on rjqht VF-6 57 300 t VH (full length, constant chord) 21 I
0.71 cm ext. SLEE F-2 F ~ '24 I !
0.48 cm ext. SLEE F""2 F ~ 25 Sharp 0.23 cmexl SLEE F-2 F 26
leading- 0.00 em ext. SLEE F-2 F 27 J edge 0.48 em ext. SLEE SC-1 SC~ 28 i
'exten sian s 0.48 em ext. SLEE SC"'l SC~ SC 34 35 0.48 em ext. SLEE F-2 SC-1 F SC- 33 -0.48 cm ext. SLEE, F-2 F ~ F - 32
34
minimize its In{luence on the leading-edge flow development of the basic
delta wing. The accompanying reduction in profile drag acted to better
show up the effect of the leading-edge devices on reduction of the lift-
dependent drag. Indeed, the size of the housing was reduced to a
minimum compatible with the requirement to contain the balance adaptor
and scanning-valves.
In an effort towards economizing wind~tunnel time per configuration,
several pressure orifices on the original research model (ref. 5 investi-
gation) were omitted in the present investigation. The-effect of
eliminating the four most aft pressure orifices (two each on the upper
and lower surfaces at each of the six spanwise leading-edge pressure .-
stations of the original research model) on the accuracy of le~ding-edge
thrust calculations by pressure integration was checked and found to be
within 5 percent of the value obtained with all the original orifices in
use within the a range of interest. This was considered acceptable
since the present investigation was mainly concerned with relative,
rather than absolute, levels of leading-edge thrust. In addition, all
except one of the original upper surface orifices near the trailing edge
were eliminated. However, 13 additional orifices were added along the
wing leading-edge interjacent existing orifices in an attempt to better
define the movem~nt of leading-edge separ~tion in the spanwise di~ec
d~'n'-. It was estimated that, in this mann~r, the test duration per run
_ was reduced by approximately 20 percent without sacrificing the prime
objective of the study.
~
35
DATA REDUCTION
Forces and moments sensed by a wind-tunnel balance must be cor- .
rected for exte~nal interferen~es unrealistic of actual flight. Cali
brationof the wind-tunnel test section in reference 9 shows aconst~~t
streamwise static pressure distribution at the test Mach number and, .
thus, nb correction was needed for longitudinal buoyancy effect. Jet·
boundary corrections were applied to angle of attack to account for the
vertical velocity induced on the model by the test section walls
(ref. 12). To account for initial balance loads due to model weight,
wind-off \'Jeight t~re measurements were taken at various balance atti
tudes and used in the reduction of balance data (ref. 10). Balance
axial force measurements were also corrected to eliminate housing pres
sure drag using chamber (base) pressure measurements. Reference 13 was
used to calculate solid and wake blockage corrections due to the
presence of the model and wake in the test section .. Since the angle
of attack was measured by means of an accelerometer located inside the
model,. no correction for sting bending due to aerodynamic loading was
required.
Once all neces~ary corrections had been applied to the balance
data, the final results were expressed in terms of force, moment, and
pressure coefficients. Force and moment coefficients were computed
based on the basic wing (devices off) planform area. This is quite
legitimate since any addition of planform area from a device is an
essential part of that particular concept. Lift and drag coefficients
eCl' and "CO' 'res'pectivel:y;)an:e or; en ted 'along 'the conventional wind
,axis coordinate system, wi'th t'heaxial and normal force coefficients
{CAand eN' respectively) along thebod:y axis system (sketch G). . ("' ~ (. i
--)10_ ~D
Sketch G
Momentcnef'ficientsrefer to :th:ebodyaxe:s~Pitching momentcoeffi--
36
cientshav.ebeen ·modi'fiedfrom 'those reported in referencell by moving
-the moment ,center further aft ,using the :equa~tion
C m 'c AXc = +~·C
ml 'c ·N' 'r
fix 2m =0. OS for StEEand vortex 'fl aps r
=O.l65 'forother devices'
where em1 i scomputed abouttheorlgina"l moment reference center (see
fig. 2). Definingequati bns for force, moment. and pressure coeffi-, .
dents' u:s~d 'i n theanalysi s are g:;ven in 'the SYMBOLS section .
The effect'·ivenessof the devices under considera'tion depends
;crucially 'on leadi-ng~edge flm'ltontroT,~vhich jnturn is :best observed
37
through improvements in the leading-edge thrust characteristics. The
method of integration of the measured static pressures around the wing
leading edge to obtain the local leading-edge thrust is described with
the aid of the following sketch:
~~ta/CSj I .\/ ~ .Y'
STA j
f
. leading-edge curvature
Sketch H
i-I .........:.=- • . ( , z, - z f /).Z.= 1+1 i-l)
I 2
.. x
(ref. 5)
The suction force developed at the jth spanwise pressure station is
defined as
m SFj = ~(Pij - poo)a ~Zij
;=1
where i, denotes a specific orifice at the jthstation and a is the
length and ~Zij the height of the vertical i-jth pressure panel.
Assuming a constant suction force per unit length of the span (ref. 14),
'SF. J avg
''']
SF. =_J
a
the -totaTsuction-fc:irc.~ ,contr5butedby the jth pressure station is
SF. - ( SF. Jl
oc = 2,Q;)~ a
where ZQ, i sthe "to'tal 'leading..:edge 'length of the wing. Usingthe
defi nitionof ~thepressure toeffi Ci ent"
Cp, .. 'lJ
_Pi} '-POo q
00
ananond'imensi onaltz:ing using the reference force ( q~Sref) yields the
-suc'ti On forte coeffi ciient
c., Sloc
SF· Jlo'c
= Q66Sref
m , " C (2Q,'):2: Pij t;zij
Sref i =1
38
Taking the thrust component of the suction force gives 'the local leading
edge thrustcoeff'i ci ent
'c - ,m (2£)C A 'r _= "" ,p .. DZ'l· J. cos n l~c~' 1~ lL
i=l
39
Averaging the sum of the local thrust contributions and noting that
(21) cos A = b, the total leading-edge thrust coefficient becomes
CTtot
= ~ L: L: .. P~jlJ n (m be !J,z • • )
j=l ;=1 ref
40
PRESENTATION OF DATA
The force; moment,and"staticpres5ure graphs used in support of
discuss'tonof results arepriefly summarized in this section.
Forceslh the body axis coordini:i'te system (CA and CN) were the
main par~meters used for the irtitialperforinance assessment of each
device:' Since the investigation was mainly concerned with drag n~duc
tion through leading-edge flow contrOl, axial force was particularly
well-sU~ted tor this ~Grpose si~c~it providei a sensitive and dir~ct
indication of leading-edge thrust. For demonstration of longitudinal
stabil ityeffects., the pitch; ng-morrient data were transformed to a ref
erence position different from that prevailing during the tests (see
DATAREDUCTTON). This new reference center was chosen s6 as to give
approximately neutral stability at low angles of attack for a closer
approximation to the'condition expected to ,prevail on an actual a i r-
craft. L'ift-lo-drag and drag polar (CD vs CL) curves, conventi ona lly
used fo~descrfptfo~ of the aeroaynainic characteristic~ from a perfor-
, manee point of view~ were given secondary importance in the assessment
of the devices. An advantage of using the lift-to-drag parameter is
elimination of plan form area effects in case of devices such as the
vortex flaps 'and sharpleading..:.edge extension,. Since the basic intent
of this investigation was alleViation of induced drag penalties, the
effect of each devi ceon the; nduced drag of the wi ng was reflected i h
plots of an induced drag parameter; K. Details on the calculation of
this paranieter are given in the Ba'sic Wing section of RESULTS AND
41
DISCUSSION. Additional force plots such as CL, CO' and C1 as func
tions of . a were also included in support of discussion. It should be
noted, however, that the ov~rall force dat~ are representative of the
wind-tunnel model rather than of any actual aircraft. Therefore,
emphasis should be on the relative rather than the absolute magnitudes
of forces and moments and also on the leading-edge static pressure
measurements.
Leading-edge pressure data were basically used as an aid in inter-
pretation of trends in the balance data and also in assessing a device's
ability to favorably modify the leading-edge flow field. Graphs of
leading-edge static pressure (Cp ) and integrated thrust rCT ) as \ LE· '\ . loc
functions of a, at specific spanwiselocations, were useful in detec-
tion of local leading-edge separation and determination of the leading
edge sUction effectiveness of a device throughout the angle-of-attack
range. These figures were also used to plot a boundary between sepa
rated and attached flow and, thus, follow the inboard progression of
leading-edge separation. These same oarameters (C p . , . LE and CTloJ
plotted against spanwise position, at specific angles of attack, also
depicted the effectiveness of a device along the entire span. Static
pressure variations around the leading edge (at the six chordwise rows
of pr,essure. orifices) provided added insight into the specific flow
patterns existing at the leading edge and also reflected the relative
thrust and normal force contributions from various positions on the
leading-edge curvature. For comparison, basic wing (devices off) data
appear as a dashed line on selected balance and pressure data plots~
" RESULT? AND DIS~USSION
,)
60:'D~'g' Del tel ~~rng
The basic '6o'-deg 'delta' vli'ng '(d~\I;;ces'off)served as the baseli'ne
confi gurati on throughout the invest; gati on for compar; sdnand perfor:'
42
manCe assessment of the vari otis leadhig-edge devices. F8rc~a.nd moment
data obtairied for thi s basi c wingconfiguratio'h a.'r'e pres~nteagraphi:" caliy in figure 9. The negative n'ormal force and posi1:1ve pitching
moment at' CJ.~'Od :'are attributed ,to the negative camber eff~ct of an
asymmetri ca lly bevell ed wing trainrlg-edge region (see fig. 2), \<Jhlch
simulated an ,up-deflected trail irig~~dge fl"ap.
BelOw a.'~ 8°, the delta wing isuric:ler the influente of fully
attached 'flow with 100 percent leading~edg'e suction, as evidenced by the
close agreement bf the ax; iil and normal force dirves (corrected for
zero~lift forces)ytith potential flow' th~or.Y (vortex lattiCe m~thb'd with
suction analogj code (~LM-SA); ref. 15) in figure 9. At a~~roximatelj
a. = gO; referred to as the departure a'ngTe of atta.ck (aD)' flow separa':'
bon occur~ ~t the streaniwise tip~ and the subsequent vortex fOrination
spreads toward the wing apex with increasin'g a.. The low pressure pro..:
duced 1 oca lly on the wing upper surfa.c~ by theh; gh rotafiOnal vel Dci-
ties withln the V()rtex generafes additi'dnai 1 itt, which can be noted in
the departUre of the eN curve ~rDm the lrlitially linear theoretical
100 percent suction curve. However; th'ere is an accompanYi ng reducti on
in leading~edge ~ucti6n. The deViation of the 'CA cuHie from the
paraboliC theoretical potential flow <;:urve at approximately gO is
CA
Cm
.02 .~-
-.02
-.04
-.06
~ "',"\.
'~
---("...,-- Basic wing (expt. )
-- - -- Expt. CA - CAo - - - - - - VlM-SA theory
\'\. \ ""-. -- -- .
\ -- ---\ -.... \ \ \ \
Potential flow
I I I L-__ ~ ____ ~ ____ -L _____ ~_~
o 5 I 1 0 15 20 25
.015
.010
aD
.005 f--
1 o !...-- I
0 5
a. deg
Xcm = 0.589 cr
L ____ L 10 15 20
a. deg
25
Figure 9.- Basic wing force and moment characteristics.
43
CN
Cz
1.0 .. ; ;
.8
.6
.4
.2
D06[ 005
I .004
.003
.002
.001
ex.. deg
aD
I-I
Separated flow
Potential
~ Basicwing,(expt)
---Expt C .e.C . . ..• N No - - - -..;.,. - VLM:-SA theory
o I I vi o 5 10 15 20 25
ex. deg
Figure ~.~ Concluded~
44
45
characteristic of this gradual loss of leading-edge suction. This
trade-off between leading-edge suction and vortex-induced normal force
is the basis for the Polhamus suction analogy theory (ref. 15) used in
, this paper for comparison with experifuental data. In addition, the aft
position of the primary vortex at a = 9° locates the center of vortex
lift aft of the wing moment reference center (approximately at the cg),
resulting in a strong pitch-down~ as reflected by the sharp downturn of
the Cm curve (fig. 9). Although not presented here, it is important
to note that the maximum lift-to-drag ratio for th~ basic wing is
attained at approximately a = So. This implies that the additional
vortex lift is insufficient to compensate for the loss of leading-edge
thrust (drag increase) following the onset of separation.
The inboard spread of leading-edge separation is shown in figure 10
to occur very rapidly with increasing a. Plots,of leading-edge pres
sure as a function of angle of attack at selected spanwise positions
indicate a buildup of leadi~g-edge suction (negative Cp)with increasing
a up to the local onset of s~paration, followed by a sharp collapse of
suction at all but the inboard~most station. Plots of this type lead
to the angle for local onset of separation as a function of spanwise
position, shown in the same figure. The locus of data points, which is
effectively a boundary between separated and attached flow, indicates a
movement of leading-edge separation from approximately the 90 to the
30 percent semispan position within only 30 a increment (viz., 11°
to 14°). The corresponding pressute distributions around the leading
edge are presented in figure 11. 'At a = 10°', attached flow,
e", I"'LE
7J
-4
-3 '
-2
-1
Orifice III
232
~\ '2\Q, r \ q
Orifice 111 (T] " O. 232)
I b.. ,I ..' 'o..-D--D-----o ZIDI'IH.6Il2J
~ ''O---~- ---0 232 (7] = 0.847)
o gel
o
a, deg
1.0
0.8
I 'I' c9
LE.
0.6 r d LE.
Separated
Attached
0.4
0: 2,
o 'I "
o 5 10 15 20 25
a sep, deg
46
Figure 10.,.- Bas i ewing- -1 eading-edge, static pressure vari ati o'ns and Cp deri ved separation boundary.
LE
-4r-
-,3 f-
-4r-n= 0.20
-2):, ~,3r-
C Pl
n = 0.33 -2
C P2
-1
2 ,3 4 .5 0 2 ,3 4 5
-4~ X, em
-~ ~l n = 0.57
C P3
Cp 4
1L--1 1 1 2 ,3 4 5 q 1 2 ,3 4 5
X, em X, em
-4i -4r
-,3L n = 0.70 ~3r- ·n = 0.82
-·2
Cps -1l~, CP6
--0 -1~. ~-=: --0----0 °c ·1·1 I I I I I
1 2,3 4 5 1 I ° 1
2 ,3 4 5 0
X, em X, em
(a) 0 ex = 10 .
Figure 11.- Static pressure distributions around the basic wing 1eadingedge.
47
y--
CP1
CP3
c Ps
1 r -I o
-,i
n= O~20
2,,-,. 3 4
-4.~ X, em
n = 0.45
.~~~ -~~I··?
o 2 3 4-
X, em
-4 1
n =0.70 -2
--1
o ~ ~ I ... -
5
5
1 L! ____ ~ ____ L-__ ~ ____ _L __ ~
o 2 3 4 5
X, em
CP2
CP4
, C!'6
'(b) ex = 12°.
-4
-3 n = '0.33
-1
1 0 2 3 4 5
-4~ X, .em
~3 , n = (L'57
-'~ ..•. ~ •..•.•. '," O· "',' . I 1 . o 2 3 4 5
X, em
-4
-3 n =0.82
-:~ 1 LI __ ~L-__ -L __ ~L-__ -L __ ~
o 2 4 5
X, em
Figu're 11. '-Con'ti-nued.
48
49
-4r -4--
I -3,
n = 0.33
cP1 -2~ c
p2 -2~
~--D
-1
1 I I I I I 01' 2 3 4 5 0 2 345
-4 r- X, em -4, X, em
-3f--- -3 r n = 0.45 ,n = 0.57
-2
CP3 lV-----V--O----.:..o CP4 ~-'-O -1 -1~
~~II ~t 1 1 I I o 2 3 4 5 02 3 4 5
~~ ~~
-4~ -4~
-3~ -3[ n = O. 70 n = 0 82
-2~ -2 Cps - , CP6 'I ,
-1 -1~
t~IO ~~I 1° o 2 3 4 5 o 2 3 4 5
X, em X, em
(c) a = 14,0.
Figure 11.- Concluded.
50
characterized bya 10caL~uct50n pea.k near·the leading edge, due to flow
acceleration, and subsequent pressure recovery on the upper surface,
persists at each spanwise station considered. Flow visualization indi~
cates separation has begun near th~tip but apparently has not traversed
far enough to be detected. At q, = ,120,. local separation is indicated
at n = 0.82 (STA6) by the constant pressure, stagnated flow region
an the upper surface. The leading~edge flow remains attached inboard.
An increase in a to 14° shows the rapid ..inboard movement of separa
tion, as its apex now appears, to lie between n = 0.33 (STA 2) and
n = 0.45 (STA 3).
The rapid inboard spread of .le,ading-edge separation induces an
erratic wing rolling moment behavior between 90 and 150 a, as shown in
figure 9. This so-called "wi.ng rock,," characteristic of all highly
swept wi ngs, results from asymmetrical ~pam'li se movement of separati on
along the two leading edges. In addition. the forward movement of the
primary vortex leads to pitch-up at approximately nO a, as the center
of vortex lift moves ahead of the wing moment reference center.
The axial force reversal at a ~ 11° is best explained through
inspection of spanwise leadi~g-edge thrust distributions in figure 12.
At lo~."anglesof attack, ,axial force improvement with increasing a is
attributed to local thrust gains all along the span. Increasing thrust
values· toward the tip are due to increasingupwashoutboard, .resulting
in higher local effective angles of attack and, consequently,greater
flow accelerations and.suction forces around the leading edge. In the
mid-a range (11°_14°), however, a balance between thrust loss at the
C Tloe
1.0 r-0.8
0.6
0.4
0.2
a = 22° \
\ \ \
\. \ \ \
° \ 18 \ ~:\
.. \\,\\ / \\\
16° / /~Z~\ .. ,....--~~'X"'" ~\ \' / '" '\ / \s..\ /' \ "
/ ~ ,-- , ----.;:~\ ° I , , 14, ,,'- .~, '(Y = 11 / ' \ ,
\ )~\. " \
8° / --- // '---",-\\_ '~12' /
,-/ ~ "0' 12° ,/,/ >\~", ;3: " / .', "<"'14,
/',- / .... ~" lI°,- /- ....... l~' '/ " 18' 10° /' .///
./ ~·O . ./ -- "'-6 ./
80~ 6°
51
O~I ------~----~------~------~----~ o 0.2 0.4 0.6 0.8 1.0
r;
Figure 12.- Basic wing spanwise leading-edge thrust distributions. Symbols omitted for clarity.
52
outboard stations, due to local leading-edge separation, and continued
thrust gains inboard account for the relative insensitivity of the CA
curve to angl€ of attack. Above' a = 140 , additional thrust gains
inboard with relatively minor further losses outboard account for the .
CA recovery. These same leading-edge thrust data are plotted versus
angle of attack in figure 13~. Here~ .local thrust is compared VJith
balance axial force measurements correCted for profile drag (CA - CAo )'
which should represent strictly leading-edge effects and, thus, an
approximate average leading-edge thrust.. .The close agree~ent between
the local thrust and balance data prior to separation adds credibility
to the pressure integrations. Again, as in figure 12, the axial force
recovery at high a is shown to result from continued thrust gains at . I
the inboard stations as the outboard stations settle at a constant value
below the balance-derived average. Note that in the mid-a range
( 100 to 1<°, the ha1::>·':'r-e d::>+a r-"ns;s+en tl·,,;4=a" ..... '" -'-_" \.oIl ,-,IU,-lIw u.t, yv11 I v II J".I I. below the integrated
, .. , thrust values. The balance is, thus, sensing a source of drag other
than that accbuntable from loss of suttion, possibly from trailing·edge
separation or interference· from the housings.
At high angles of attack, the .rate of inboard spread of separation
diminishes, as reflected by the leveHng off of the separation boundary
in figure 10. Continued axial force improvement is observed up to the
highest angles tested. Previous research has shown that further
increases in angle of attack would eventually lead to vortex burst,
spreading rapidly toward the wing apex with an accompanying loss of
leading-edge suction and pitch~up.
53
----0-- Pressure-integrated ----0----- Sa 1 ance-deri veo -(CA-CA )
0 .10,- n •1O r-.0Sf- n = 0.20 / I
.OS[- n == 0.33
.06f- r! I
C C TI 2.06 Tloc1 CC.,. -
.04 uU·--D .04j P .-D .au
---~ . . 02 .02
0 I~ 0 0 10 15 20 25 0 5 10 15 20 25
IX, deg IX, deg
.10[ . 1°i
n = 0.45 I 11 = 0.57 .os .osl
c- 06r R~ CTloC4 .06 ~ 'Ioc~ .
.04f- Y f""'""t.-n--""'-"" 04~. ~
.02f- J:Ei 02cL . o L .... 1J .. J 1 1 -....J o . _ .. :....._L-.-L_J
0 5 10 15 20 25 0 5 . 10 15 20 25
a, deg a, deg
.10,-.10!
.081- n = 0.70 n == 0.82 .0Sr-
CTIOC5 .06 r' C I T1oc6 .06 r-
c..D 0.-0-".-0 I
.04f- / I .04 1
.02~ .02L
0·' I~ 01 L,· I -....-L ...... J 0 5 10 15 20 25 0 5 10 15 20 25
a. deg a. deg
Figure 13.- Basic wing leading-edge thrust characteristics.
54
It was previously stated that severe induced drag penalties are
characteri sti c of delta wi ngs at hi gh 1 ift. In order to see thee-ffect
of leading-edge separation on the total induced drag of the wing, a plot
of the induced drag fa~tor K, y.Jh~re
c - Co . D mln(?fAR) K = ? . C
L
1 e
i s pr~serJt~d in fi gur~ 14. Thi s formul a is derived from the equati on
for minimum induced drag,
CD = CD. + cL2K
1m n ?fAR
where K = 1 corresponds to fully attach~d flo\f.i and K > 1 to part; a 1
leading-edge separation. Very low angl~s of attack ~v~re ignored since
low lift and drag coefficients reS.l,llt in sporadic K values. Figure 14
depicts attach~d flow at low a,follow~d by a rapid increase in induced
drag as leading-edge separation spreads inboard with its accompanying
loss of leading-edge suction. - The reduction in the rate of spread of
leading-edge separation near the wing apex results in a reduction in
slope of the induc~d drag curve at approximately a =_ 14°. In the case
of an actual aircraft, induced drag would continue to increase at
approxi l11ately the mid-a rate since decreasing leading-edge radius toward
the tip cannot retain the degree of residual suctign of the uniform
radius leading edge of the present research model.
55
2.2
.... :-
2~0 ~C 1
K = _0_ (7r AR) = -C 2 e L
1.8 ~CD = Co - C
Omin
1.6
K
1.4
1.2
1.0
.~ Fully attached flow
0.8
o 5 10 15 20 25
a, deg
Figure 14.- Basic wing induced drag characteristics.
56
A comparison of test data wit~ theory ispresehted in figure 15 in
the form of a drag polar c'orrected for zero-lift drag (CO i = Co - COo),
Experi'mental values are compared with theory for zero and full
(100 percent) leading-edge suction. At low angles of attack, where
fully attached flow prevails, 100 percent leading-edge suction is
realized .. At approximately a = gO (CL ~ 0.25), deviation of the
experimental values from the theoretical 100 percent suction curve
signifies the onset of leading-edge separation. The loss of suction
continues with increasing lift and eventually settles along a constant
percent suction curve, indicating residual leading-edge suction charac-
teristic of blunt leading edg~s.
The major contribution to the integrated local leading-edge thrust
is developed by those orifices near the wing leading edge (see DATA
REDUCTION). This suggests the possibility that a si·ngle leading-edge
pressure orifice m~ght suffice for a chordwise series of orifices around
the leading edge for leading-edge thrust determination on wind-tunnel
models. If so, it would result in significant savings in model con-
struction costs and wind-tunnel test time.' Figure 16 presents staggered
plots of CPLE versus eTloc at the six spanwise ~ressure stat~ons of
the research model. The linearity of the curves is interpreted as
attached flow and the, departure from linearity as the local onset of
separation. This indicates that CTloc can be calculated from CpLE
only in the attached flow regime. No systematic relationship between
CpLE and CTlocis apparent following the onset of local separation
0.32
0.28
0; 24
0.20
Co. I
" 0.16
0.12
0.08
0.04
o o
Co.= Co - CD" I 0
/ /
/ /
/
/
/'
y""",
/'
I /
/ '
I
" I
I /
I /
!
/ /
/ /'
0% suction I
I
I I I
I " I
/ /
I I I I
/ /
/ /
/ /
j":
/
100% suction /
I /
I /,
/
~ Experiment - - - - - VLM- SA theory
0.2 0.4 0.6 0.8 1.0 .
CL.
Figure 15.- Comparison of basic wing experimental data with VLM-SA theory.
57
, c
.0
\.
STiA 1
.0 • .02
2 , ;3
4
. .0
·1
5 6
0
-..,....'.,-j .. 0 STA 1(7]='.0.2.0) ~~-4(J----2:( ; '0.33) -.' . -~"-. , 3,( ='.0.45)
, ·...I~, "4( ,=;0.571 - ..... ....,. ~,.....,... -5 c( ="O~ 70)' - .. ' -~--.-, ;6'( =.0.82)
.0 .0 .0.02
Cr loe
0.04
Figure 16.- Basic wing le~ding-edge static pressure-thrust relationship.
'.::
, "
U1 00
59
due to the significantly altered Cp distribution around the leading
edge: Further discus~ion on this type of plotting will be ptesented in
conjunction with each leading-edge device.
/
;,/' /'
60
Chordwise Slots
Single and multiple chordwise slots cut into swept leading edges
produce performance improvements as reported iii reference 5. For refer-
ence purpo~es, basic configurations of cine and three chordwise slots per
leading edge were retested in the presentirwestigation.
The results with a sing'le slot were used to illustrate the effects
of the devi ce on tne wing leading-edge flow pattern and overall perfor
mance. T6 investigate the possibility of further performance improve-
mehts j the slot was modified with an internal contour (5C-l), which will
be discussed later. Figure 17 presents results of force and moment
measurements on a configuration utilizing a single slot at n = 0.625.
the CA curve shows the expected 10w-a drag penalty discussed previously
(BACKGROUND section). However, the slot acts to delay the onset of
separation and vortex lift, as indicated by the slight loss of normal
f6tc~ in the mid~a ra~ge. The result is ~~~~~-~~ l~~d·~n·g· ~~"~ C'lllfOfl\....CU J co -I I -CU':jC suction
and, thus, CA im'provement beyond 11 0 ct. However, a sudden loss of
slot effectiveriess at high a is ihdicated by the convergence of the
slot and basic wing curveSj beginning at approximately a = 17°.
Longitudinal stab,i1ity e·ffects of a single slot (Cm curve in fig. 17)
can be noted as a slight moderation of the initial pitch-down,' as well
as a delay of pitch-up to approximately 120 a. From a performance
standpoint; a penalty in max;mumL/D value occurs, but a substantial
improvement is obtained at mid a as a result of , the drag reduction.
Static pressure distributions around the leading edge at adjacent
st~tions on either side of the slot, in figure 18 (a = 16°), illustrate
CA
em
.02
-.02
-.04
-.06
.02
.015
.010
.005
-- ---- Basicwing
---{)-- 1 x SC-l @ 7J = 0.625 ----{}---- 3x SC-l @ 7f"'0.25, 0.50, 0.75
I I__J o 5 10
a, deg
15 20
Xcm = 0.589 cr -,
I
/ I
/ I
I I
25
o I I I '~,/-,-I ~---' o 5 10 15 20 25
a, deg
Figure 17.- Force and moment characteristics of single and multiple chordwise slot configurations.
61
62
1.0
.8
.6
eN
- - - - - - Basic wing
---()--:-- 1 x SC-l @ 7] = O. 625 ---0---. 3x SC-l@7]=0.25,O.50,O.75
20 25
a, deg
10
8
6
LID 4
2
o ~I ______ -L ______ -L ______ -L __ ~ __ _L ______ ~
o 5 10 15 20 25
a, oeg
Fi gure 17.- Conel uded.
63
-4 - -- - - - - Basic wing
---0--- 1 x SC-l @ 17 = 0.625 -3
7]=0.57
-2 C
P4 ~-~-O---O=-=-=-=-.::..=-=-=o
o ~--V----=o
1 I J_J ___ L __ ~ o 2 3 4 5
X, em
-4 ---'-
--3 1] = 0.70
CP5
------~ -1
1 _~ ____ L--,--I-,------, o 2 3 4
X, em
Figure 18.- Loca1 static pressure effects of the chordwise slot at a = 16°.
5
64
the effects in the vici~ity of t~e device. While little influenc~ is
detected 'inboa~d,the slotco~~artnientati~n"~ffect (see BACKGROUND
section) acts to retain attached flow at the leading edge on the out-
board side. \tJhere the basic wing is stalled. The resulting eff.e,cts on
leading-edge suction are depicted in figure 19, where sizable increases
in negative leading-edge pressure (leading-edge suction) are indicated
on the outboard side beyond 130 Ct. The spariwise variation of CPLE at
Ct = 160, in figure 20, reflects these bas,ic trends on either side of
the slot.
While the local leading-edge flow field seems to have been
favorably modified by the slot, an overall performance improvement
depends on the device's ability to increase the total leading-edge
thrust.' Figure 21 presents local thrust variations with angle of attack.
The local onset of separation at the two pressure stations outboard of
the single slot (STA's 5 and 6) has been delayed by approximately 1° Ct,
resulting in large thrust improvements beY9~d Ct ~ 11°. The loss of
axial force at high Ct, discussed previousiy, is shown to start at the
outboard stations as the local thrust and basic wingd~ta converge.
This is believed to result from a breakdown of attached flow on the
upper wing surface, which begins at the tip in view of the higher pre-
vailingupwash. The compartmentation effect of the slot is, thus, lost
and the primary vortex allowed to spread inboard. The inboard effect
of the slot results in a minor penalty in local thrust between 140 and
20° Ct. However, this unfavorable effect is highly localized, as
c .. PLE
C PLE
-3
-2
-1
o ! cr o
-3 I
-2 I
-1 ~
5
r;
;/
Orifice 210
2ll
10
a, deg
,~
;1
..ff
- - - - - - Basicwing ----10 1 x SC-l@"'i=O.625
Orifice 210 (7]= 0.602)
15 20 25
~ Orifice 211 (7]= 0.663)
-... ....... ......... -- -
o I if o 5 10 15 20 25
a, deg
Figure 19.- Local suction effects of the chordwise slot.
65
C PLE
c PLE
-5
,...4
-3
~2
-1
00
-31-
I -2
-1
",.
q fu .~ gb~\~
",'1 '~.
,2 .4
b
~\, .......
6
- - - - - -Basic wing
, ~ lxSC-l@ 7]=0.625
-':'~--o---- 3 x SC-l @ 7]= 0.25,0.50, 0.75
a = 16°
.6, .8 1.0 I
'1]1 I I
a,: 23°
o ·I<--_--"--__ -L-__ .L..-_---L __ .....J
o .2 .4 .6 .8 1.0
'T}
Figure 20 • .:.. Spanwis'e leadirig.:..edge static pressure distributions for single and multiple chordwise s10t configurations.
66
-----.--. Basic wing STA 1 ----{)-- 1 x SC-l @ 11 := 0.625
2 ----0------ 3 x SC-l @ n '" 0.25. 0.50. 0.75
.10
.08~-
3 4
5 6
CT .D6 loel
·.b4~ = 0.20
.02
01 fJ [ o 5 10 15 20 25
a, deg
.10,
.08 n = 0.45
CT 3 .06 loe
--D 0--'
.' /
/ ~- ...
. 04
.02
o o
.10
.08
CTloes .06
.04
.02
~ I 5 10 15 20 25
ix, deg
n = 0.70
o I ..... I I J o 5 10 15 20 25
a, deg
.I0r
.081--j
~ ... }.
? /
CTIOC2 .06
n,o"tv .04
.02
o I ifl ~-',_--' o 5 10 15 20 25
a, deg
.10 r-I
.08 n = 0.57
CTIOC4 .06
R p-IT ''':0
.04
.02
OL-HI o 5 10 15 2J 25
a, deg
.10r-
I .08 L_ n = 0.82
.06 ~- .. "0
+-fo ,o:r- ~ I I I j
o '-" 5 1 0 1 5 20 25
CTIOC6
ex, deg
Figure 21.- Effects of single and multiple chordwise slots on leading-edge thrust.
67
68
stati~~s inboard of - n -~ 0.57 rem~in unaffected~ The favorable out-
board effect is more widespread .'" , ;~.
A multiple slot configuration was tested ih an attempt to spread
the slot benefit over a larger portion of the span and to delay, or
eliminate, the loss of effectiveness at high a. The s16ts were located
at the 25, 50, andl'S percent semispan positions and Were again inter
nally contoured with 5C-l. The result was further delays in local onset
of separ~tion and, significantly higher local thrust values in the mi,d
and high-a range at all but the most inboard pressure station (fig. 21).
Apparently~ the favorable iriflUence of the ",upstream" slot acts to
nullify the adverse inboard effect of a single slot, resulting in thrust,
enhancement along a large portion of the span. the lack of a slot
inboard of stAl leads to early separation and l6ss cif thrust near the
apex. Compari son of spanwi se 1 eading-edge pressure variati ons at
a = 1 &:0 IV , in figure 20, also reflects suttion improvements outboard of
n = 0.25 withrilultiple slots. Based on these results, it is believed
that several slots located along the leading edge have the ability to
redirect the spanwise boundary layer flow in the chordwise direction
before it builds up in, magnitude. Again; however, loss of effectiveness
begins near the tip, as the c6rresp6~dihg CpLE data at a = 22~
(fig~ 20) converge to the basic wing curve. For a ready assessment of
the leading-edge effects- of multiple slots, the following table presents
a comparison of total leading-'edge thrust values for single and multiple
slot configurations:
r-
I CT
to
Basic Wing (BW) 0.04 1 x SC-l .04 3 x SC-l .05
C); == 14°
Percent t over Bt~
39 -86 10.7 84 33.0
69
a = 180 a = 220 I -.
Percent I CTtot
Percent CT over BW tot over BW .
0.0466 - 0.0534 -.0512 9.9 .0555 3.9 .0656 40.8 .0645 20.8
These data indicate that high-a thrust improvements of over 40 percent
(~elative to the basic wing) are obtained with a ttiple slot
configuration.
Figure 17 compares force da.ta for single anq multiple chordwise
slot configurations. The CA curve indicates that multiple slots pro
vide a substantial drag advantage beyond 120 a with a more gradual CA
reversal which is delayed from 120 to 190 a. Multiple slots also pro
duce a smoothet Cm curve, delaying pitch-up to 190 a. The additional
loss of lift beyond 120 a with the use of multiple slots is attributed
to the retention of attached flow (and, therefore, loss of vortex
induced lift) over a greater portion of the leading edge. Maximum
lift-to-drag is unchanged from the single slot case, but additional
improvements are indicated beyond 100 a before the final convergence to
the basic wing data at higher a.
The ability of chordwise slots to delay the local onset of separa-
tion is reflected by the extension of the linear portions of the CPLE versus CTl curves in figure 22. Delays of up to 20 a are indicated
oc just outboard of each slot in the triple slot configuration. The
extension of the curves along their initial slope suggests that attached
-5
-4
-1
STA 1
2 3
4 5
6
- -'- - -'- - - - Basic wing (all STA's)
--0--- STA 1 (7J = O~ 20) --[}-- 2 ( = 0.33) ~-o--' - 3 ( = 0.45) ~--ic:.' ' 4 ( = 0.57) ---Itr-----' 5 ( = 0.70)
3 x SC-1 @
7J = 0.25, 0.50, Q.75
--D--- 6 ( = 0.82)
o ~~ __ ~~ __ ~~~_~~ ___ L-~_~_~ __ -L __ ~~ ___ ~
o o o o o 0.02 0.04 0.06
I.. 0.02 J I
Figure 22.- Effect of multiple chordwise slots on leading-edge static pressure-thrust relationship. Selected high.;.a data points have been omitted for clarity. Angles in parentheses refer to basic wing data (see fig. 16).
....... o
.'
71
flow has been extended to higher a without modifying the local effec--
tive upwash. This suggests that the slot action is more due to compart-
mentation rather than to any vortex mechanism.
As mentioned in the BACKGROUND section, the object of internal con-
touring ofthechordwise slots was to reduce the low-a drag penalty
while improving the high-a performance. The low-a drag effects will be
discussed first. The following table compares zero-a drag measurements
on a triple slot configuration utilizing the various slot contour shapes
tested (fig. 3):
I CD Percent Device 0 ·over BW
Basi c wing 3 x open slots· 3 x SC-l 3 x SC-2 3 x SC-3 3 x SC-4
0.01261 .0l318 .01321 .01293 .01296 .01307
4.52 4.76 2.54 2.78.
. 3 h~ l __ ~~ ____ _
Comparison of open (uncontoured) slots with those making use of SC-1
shows relatively insignificant effects of contouring on low-a drag.
Therefore, stagnation pressure on the back face of the slot appears not
to be the primary cause of the low-a drag penalty. This is not sur-
prisi~g considering the relatively small area of this vertical face.
Two additional contour shapes tested, SC-2 and SC-3, effectively
reduced the depth of the slots. Zero-a drag measurements for configura
tions utilizing these contour shapes indicate a reduction of almost
50 percent of the drag penalty associated with the slots, with a
72
relatiy·eiYmi;;or:d·i·h:erenc~ betwe'"en the two. Since these contours
differ only insha:pe·~ ··and nbtlkng'th, these data further support the
conclusion that pressure on the vertical face at the end of the slot is : .,.
not the primary sour.ce of low-a. drag penalty. Compari son of zero-a drag
for SC-l and SC-2, which differ in length but have basically the same
contour shape, suggests that sTot depth is the determining factor for
low-a drag. The reducti on in drag penalty with SCM2 and SC-3~ there.;.
fore; results from redl,lced fticti-ondrag acting on the slot side
surfaces.
Comparisons of overall performance for the slot configurations just
described appear in figure 23, .Axial force shows a slight suction
ad~antage beyond 120 a and delayed loss of effectiveness with the u~e of
SC-1 relative to the shorter slots uti1 izing SC-2 and SC.;.3. Minor dif-
ferences ex; st between SC-2.· and SC-3. Since a sl ight high-a suction
advantage is realized with the use of 5C-1 relative to the uncolitoured
slots, high-a performance may be somewhat dependent on the shape of the
back face of the slots. However, slot depth is again the primary factor
determining high-a performance. Improved performance with increasing
slot depth is explained as follolt's: Since the postulated slot flow
mechani sm is paftia lly that of a "fl ui d fence," increasing slot depth
is analogous to increase in the chordwise length of a fence. Results of
previous research (ref. 5) indicate that increasing fence length results
in improved high-a drag performanCe and delayed loss .of effectiveness,
the same trends observed here.
CA
Cm
.02
o
-.02
-.04
-.06
o
.015
.010
.005
1.1 5 10
---0-----0-------<)----------6-------.-.-~-.--
ex, deg
15 20
x em = 0.589
c r
o ~I -----.~----~----~ o
ex, deg
73
Basic wing
3 x Open slots @ 'TJ = 0.25, 0.50, 0.75 3 x sC-ll 3 x sC-21 _ 3 x SC-3J @ TJ - 0.25,. 0.50, 0.75
3 x SC-4 . . .
CJE2/1 ~
25 ~.
~w SC-4 ~
25
Figure 23.- Effects of internal contouring on chordwise slot performance.
~ ~
74
S}otdepth is also of.prime i~portance to the longitudinal st~-.
bil ity of .the configu~~ti ~n.The most nbtabl e effect is accelerated
high;..a pitch-up with reduction in slot depth (fig. 23). Effects of slot
depth and contouring on normal; f6tce. and rolling moment are iJegligib le.
An additional slot contour; SC-4,'was fested in an attempt to
reduce thE: 10w-a drag penalty by contourtrig the forward po'H, on of the
slot, as shown in figure 3. However; sta.tic pr~ssure datCiindica1:e that
the contour may have obstructed th~ flb~ through the s16t~ resulting in
practically a nonexistent effect on drag perforinance (fig. 23). In
addition, 10rigitudina1 stability is severely degraded, with accentua.ted
pitch-up and pi tch-down throughout the ang1 e-of-a.tiaCk range. parain;..
eters such as n()rma1 force ahd lift-t6-drag ratio; not presented, also
agree closely with basic wing data, indicating that & slot starting
behind the leading edge is ineffective.
Slot width was investigated as a110ther parameter through which
further perforinahce improvements migHt be Teai ized. Compari son of
slots with widths of 0.25 and 0.50 percent of the semispanshm~ed,
however, that this is not a significant sizing cOhsideration within the
range tested.
In summary, the chordwi se slot device possesses dra'g-reduction
potential in addition to its ability to improve the 10hgltLidinal sta
bil ity characteri sti cs of the del ta wi ng. Multipl eslots spaced along
the span produce further performance improvements by bringing more of
the leading edge under attached flow a.nd by eliminating the adverse
effects inboard of a single slot. The slot effectiveness is primarily
75
throughcompartmentation of the leading edge to remove the 3~D effect
which causes earlier separation on swept leading edges, rather than ftom
any vortex action. The low-a drag penalty associated with this device
results from friction acting on the side surfaces inside the slot; the
pressure drag on the vertical face at the end of the slot is of secon
dary importance. L i kewi se, hi gh-a performance is primar'ily dependent
on slot depth rather than the shape (contour) of the back face.
Increasing slot depth produces progressively higher low-a drag penalty
but improved high-a performance.
76
Fences
As noted in the .BAC~GROUN.D~ection, fences have previously been
used on highly swept leading edges primarily for improving longitudinal
stability characteristics. The em ,curve in figure 24 shows that addi
tion of a single F-3,fence atT] = 0.625 improves the linearity of the
pitch curve, with reduced pitch:-downat a = 80 and generally a more
stable Gonfiguration throughOut theangle-of~attack range. The CA
curve in figure 24 shows that,fences also have drag-reduction capab,ility
at mid and high a, although some drag penalty at low a is inturred
due to friction drag on the fence. The advantage of the device is first
seen as a slight delay in the onset of separation to a ~ 10°. Beyond
approximately 11 0 a. a significant suction advantage is indicated with
the fence, with no sign of loss of effectiveness to the highest angle~
tested. The reduction in eN relative to the basic wing at angles of
attack greater than 'at separation onset also indicates delayed separa..:
tion and vortex formation with the fence attached. This is -further
supported by static pressure distributions around the leading edge on
either side of the fence (fig. 25). Inboard of the device, the flow
remains stalled at a = 16°. However, outboard, where leading-edge
separation has clearly occurred on the basic wing (as evidenced by
nearly uniform upper surface pressures), attached" flow fs maintained by
the fence. As shown by the C1 curve in figure 24, by delaying
separation. the fence also reduces the severity of the wing rock char-
acteristic of this delta planform.
77
.02 - - - -,- - - Basic wing
--0-- F-3 @ 7] = 0.625 --~-- F-3/SC-l combination @ 7]= 0.625
o "
<-.02
CA
-.04
-.06
I o 5 10 15 20 25
a, deg
Figure 24.- Force and moment characteristics of single fence and slot-fence configurations.
.".
'c .N -";,
cz
1.0
.8
.6
.4
.2
1 "n
- - -::- ~ -B~'sic~ing
. ~F-3 @7J=O.6~ ---:.o---F-3/SC-i cOmbination @ 7]=0.625
Ol,¢: J.
a, deg
.006 I
A II I I
.005 f-- ' \ I \ 1 I" I
.004~ . \ \ \ 1
:003, \ I
01 I
o 5 10 15 20
a, deg
Fi gllre" 24. - Corkl uded. "
78
.~
79
--4
- .- -- - - - Basic wing
-3 TJ = 0.57
.---0-- F-3 @ 1] = 0.625
-2 C
P4 - ..... ~--..,..;----
-1
~[ ""
I I I I I 0 1 2 3 4 5
X, em
f\\ / , 7J. 0.70
1\ \ -2
Cps
'-
o '-
~
1 I I I o 2 3 4 5
X, em
Figure 25.- Local static pressure effects of the fence at a= 16°.
80
The sepatatiorib6undary in figure c26 shows the effect of the F-3
fence on theinboard'movement'of leading.,.edge separation. The compart
mentation effect of the device (see BACKGROUND section) is seen to delay
the onset'of ~e~~ratton onit~outboatd side by approximately 30 Q.
This reducti on: in the"rat:e; Elf inboard movement of sepa rati on accounts
for the el im;:h'ati on-of' severe pitch-up,' as the forward movement of
the center;o{'vortex Tift is also delayed. This effect also diminishes
the asymmetry 'in the- spanwise lift distribution, responsible for the
erratic rolling moment behavior of the basic wing.
Local leading-edge thrust variations with a single F-3 fence at
n ::: 0.625 are presented in figure 27. At low Q, where the basic wing
flow remaihs attached, the fence has no influence on the leading-edge
flow conditions. However, beyond the lncal onset of separation for
the basic wing, substantial thrust improvements are evident outboard
of the device due to the delay in local stall and subsequent retention
of leading-edge suction. This effect can be attributed to the unsweep~
ing of the upper surface isobars in the vicinity of the fence, as
sketched below:
TJ
81
1.0
I I
- - - ....: - - Basic wing 0.9
--0- F-3 @ 7]: 0.625 t I I I
---0--- F- 31 SC-l combi nation 0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
L.E. Attached
t I I \
@ 7]~ 0.625
, Location of device
l L.E.
Separated
-"
o LI ________ L-______ -L ________ ~ ______ ~ ______ __J
o 5 10 15 20 25
asep , deg
Figure 26.- Effects of the fence and slot-fence combination on leading-edge separation (C derived).
PLE
82
- - - - - - .Bas i c wi ng --0-- .F-3 @ n = 0.625
'T ~, I :+ .08 11 = 0.33
C :06r-- I ,( C T .06 -A ;;
Tloe1 loe2
.04 f-- 0 n = 0.20 .04
.0:1 A .02
0 0 5 10 15 20 25 0 5 10 15 20 25
lX, deg lX, deg
.1°i
n = 0.45 .1°i
n = 0.57 .08 ;08 r--
CTloe3 .06 CTloc4 .06
.04 .04
:02 .02
01 . I I
I I 0 0 5 10 15 20 25 0 5 10 15 20 25
lX,·.de.g ix, deg
.10 .10
.08 n =0.70 .08~ n = 0.82
CTloes .06 CTloe6 .06
.04 .04
,o:r I
.02
0 0 5 10 15 20 25 0 5 10 15 20 25
lX, deg a, deg
Figure 27.- Effects of the fence on leading-edge thrust.
basic delta wing with fence
-Cp
1
Fence on
-"~ Fence off \
"-"- ......
" .....
83
Inboard
~ outboard
Sketch I •
On the outboard side of the fence, suction peaks are reduced and occur
further aft of the leading edge. This reduces the tendency toward
separation by providing a more gradual pressure recovery on the upper
surface~ The opposite effects inboard facilitate earlier separation
and loss of thrust. However, this adverse effect is only localized, as
little influence of the fence is evident further inboard.
The effect of a single fence on the induced drag of the basic wing
is' shown in figure 28. At low a, before leading-edge separation has
spread to the fence location, the indicated increase in K results from
the friction drag on the fence. Beyond 100 a, however, the fence pro
vides an induced drag reduction by delaying leading-edge separation and
2.2
2.0
L8 ~
L6
K
1.4 I
L2 I--
LO
0.8
o
:''-.,
~CD . _ J.. K = -. 2~ (7T AR) - e
CL
~CD = CD - CD . min
~ I
tf / /
/ ./
II
84
... ....
- ~ - - - ~ Basic wing
~ F-3@'T]= 0.625 ---G--- F-3/SC-l combination
@ 7]= 0.625
1.. Fully attached flow
I .11 I .
5 10 15 20 25
a, deg
Figure 28.- Induced drag characteristics of single fence and slot-fence configurations. .
loss of suction. This benefit is visible up to the highest angles
tested, although the maximum relative improvement occu~s between 110 o and 13 a.
85
Figure 29 presents graphs of Cp versus LE
CTloc
at the six span-
wise pressure stations of the research model. Local separation is shown
to have been delayed by 5°, to a = 16°, at the station just outboard of
the fence, with a delay of approximately 3°, to a = 14°, at the station
nearest the tip. Inboard effects are minor. The extension of the
outboard curves along their initial slope again indicates flow modifi-
cation without alteration of the local angle of attack. As discussed
previously, the fence flow mechanism is rather one of isobar unsweeping
and leading-edge compartmentation.
Reca~l that thrust data in figure 27 indicated an early loss of
effectiveness just inboard of the fence. This was believed to be
partially the result of an accumulation of viscous fluid on the inboard
side of the device from the spanwise boundary layer flow. It was,
consequently, thought that by opening an adjacent slot inboard of the
fence, this viscous accumulation could be "blown off" by the slot jet
stream (see sketch in BACKGROUND section). The concept was tested on
the F-3 fence with the inboard slot contoured with 5C-l. Results of
force and moment measurements are shown in figure 24. Although there
is an additional low-a drag penalty as expected, definite axial force
improvements are indicated beyond 13° a. For instance, at a = 14.3°,
an 8.6 percent reduction in axial force is realized with addition of
the slot. The general nature of the pitching moment curve is
-5
-4
-3
-2
-1
5TA 1 2 .J
4 5 6'
- -' -~' - -' Basic wihg(aIISTA's)
---'-----10..-.> -~ 5T A "1 ( 7'J;' 0: 20) --'--:10': 2 ( = 0; 33)' ~~<> 3 ( = 0;45) -'-~' A' 4( =0;51) F-3@7'J=0.625 --'---iCS ," 5( = 0; 7d) ---'-ID ' 6( =0,82),
o ~~ __ ~~ __ -U¢-___ ~~~ __ ~=-__ ~=-__ -L __ ~~~ __ ~
0, o o o o o , 0.06
Figure 29.- Effect of the fence on leading-edge static pressure-thrust relationship. Selected high-a data paints have been omitted for clarity. Angles in parentheses refer to basic wing data (see fig. 16).
unaffected, but the configuration appears to be slightly more stable
beyond a ~ 110. No significant effects are evident in normal force
87
and rolling moment. Figure 28 shows that a slight improvement in
induced drag also results from addition of the adjacent inboard slot.
This advantage, however, dissipates by 190 a as the fence and slot-fence
data converge.
The reduction in drag with an adjacent slot inboard of the fence
is shown in figure 30 to result from thrust improvements outboard of
the device, rather than inboard as was expected. In fact, no influence
is evident inboard. The spanwise leading-edge pressure variation at
a = 160, in figure 31, shows the same trend. This leads to the con-
clusion that the inboard loss of fence effectiveness at high a is not
hastened by a viscous fluid accumulation, but rather is totally a result
of inviscid effects. As noted earlier, the unsweeping of the upper
surface isobars leads to increased suction peaks occurring closer to the
leading edge on the inboard side of the device. This results in a
steeper upper surface pressure recovery behind the "suctioh peak and.
consequently, earlier separation. Comparison with a single slot at
n = 0.625, in figure 31, indicates that the additional improvement out-
board is primarily a fence effect. It is believed that the spanwise
flow being lifted over the fence by the slot airstream results in the
formation of a vortex off the upper edge of the fence. This vortex lies
on the outboard side of the device and rotates in a sense so as to
induce a downwash velocity outboard. The resulting outboard reduction
88
STA ------ Basic \~;ng 1
-0---- F-3 @ T) = 0.625 2 3 ---o---~ F-3/SC-l combi nat; on}
4 -----<>----- F-4/SC-l combination @ n = 0.625
5 --------6------- F-5/SC-l combination
_10r- ,\/6 /l _10
°l I ! 0+ n = 0.33
CTloc'- _06 C Tloc2
_06
0'[ / II = 0.20 _04
_02 _02
01 o::J I b 0 5 10 15 20 25. 0 5 10 15 20 25
a., deg a., deg
~ .1Or- _10 r-
.08f- n = 0.45
°T n = 0.57 ~
C"O~0't CT .06 ~ loc4
~ .04 .04
_02 .02
01 VI I I 01 &f
0 5 10 15 20 25 0 5 10 15 20 25
a., deg a, deg
.10,.---'- .10
.08 n = U. /U 0+ n = 0.82
CT .06 C .06 loc5 Tloc6
.04 .04
.02 _02
0 0 0 .5 10 15 20 25 0 5' 10 15 20 25
a., deg a, deg
Figure 30.- Effects of slot-fence combinations on leading-edge thrust.
C PLE
~ @0J
-4.5 ~ -4.0
-3.5
-3.0
-2.5
-2.0
-1.5
-1.0
-.5
- - - - - - Basicwing
--0-- F-3 @ 7] = D. 625
----D-'--- F-3ISC-l COmbinatiOn} -----<>---- F-4ISC-l combination @ 7] = D_ 625 ---- -----&----- F-5ISC-l combination --.-~-.-- SC-l @ 7]- 0.625
~
" '-"-
Device
'. ,
"'-
4,~ \
.~
, \
89
(j ~~--~--~--~--~--~--~--~--~.~ 0 .1 .2 . .3 .4 .5 .6 .7 .8 1.0
T)
Figure 31.- Effects of slot-fence combinations on spanwise leading-edge static pressure distribution at a = 16°. -
in local effective angle of attack delays separation (see fig. 26),
with accompanying improvements in leading-edge suction and thrust.
90
In order to determine the role that the upper and lower edges of
the fence play in the flow"mechanism of the slot-fence combination,
tests were run with each edge alternately removed. Results of pressure
measurements and thr~st integrations appear in figures 31 and lO,
respectively. Removal of the lower edge of th~ fence (F-4) results in
little effect inboard but substantial high-a losses of leading-edge
suction outboard. The lower edge of the fence, therefore, may have
been helpful in guiding the lower surface flow through the slot (see
BACKGROUND section). Removal of the upper portion of the fence (F-5)
also results in outboard losses of leading-edge suction and thrust.
This is most likely attributed to elimination of the upper surface
vortex tripped by the upper edge of the fence, along with its acco~pany
ingdownwash velocity outboard. The upper surface flow characteristics,
therefore, revert back to those of a single slot, with the character
istic high-a loss of leading-edge suction and thrust.
In summary, the fence possesses significant drag-reduction poten
tial~ in addition to its ability to improve the longitudinal stability
characteristics of the delta planform. The high-a loss of leading-edge
thrust occurring inboard of the device was found to be an inviscid
effect (viz., isobar unsweeping), rather than due to viscous accumula
tion inboard as originally thought. Finally, addition of an adjacent
inboard slot resulted in a slight performance improvement, possibly
through the formation of a favorably rotating vortex outboard of the
fence.
91
Pylon Vortex Generators
The function of the pylon vortex generator relies on the formation
of a streamwise vortex at the sharp upper edge of a vertical blade
extending ahead of and below the wing leading edge (see BACKGROUND
section). Figure 32 summarizes the capabilities of this device (VG-l)
in single and multiple configurations. The VG-l had a leading-edge
sweepback angle of 300 and toe-in angle of 100 (see fig. 5). These
values were found in previous research (ref. 5} to produce the best
compromise between low- and high-a performance. The axial force and
percent drag reducti on (PD) "curves show the expected 1 ow-a drag pena 1ty
with a single VG-l at n = 0.50. This penalty continues to a = 80
(which is the angl~ for separation onset on the basic wing), at which
point the device begins to show a favorable effect. Beyond 100 a, a
single VG;l provides substantial drag reduction, with a maximum improve
ment of approximately 15 percent (PO curve) at a = 14°. In addition,
the mid-a CA reversal characteristic of the basic wing is eliminated.
A constant CA increment at higher a accounts for the tapering off
of the relative drag improvement seen in the PO curve.
The longitudinal stability effects of a single VG-J located at
n = 0.50 are reflected by improved linearity of the pitching moment
curve in figure 32. Alleviation of the initial pitch-down and a 20
delay in pitch-up, to a = 13°, are indicated. The pitch-up with the
VG is also very mild in comparison with the basic wing. The normal
force curve shows a loss of lift beyond 10° a, implying a reduction
of vortex lift contribution due to delayed separation. In addition,
CA
PD,
%
·02
o
-.02
-.04t -.06
o
mi 20 ~
10
o
-'10
-20
o
5
Drag reduction
if
I 5
92
- - - - - - Basic wing
--0-- lxVG-l @7]=0.50 ---G---- 2.x VG-l @7]=0.25, 0.50
----<>---- 3 x VG-l @7]= 0.25, 0.50, 0.75
. I 10 15 20 25
ex, deg
10 15 20 25
ex, deg
Figure 32.- Force and moment characteristics of single and multiple pylon vortex generator configurations.
93
.015
.010
Cm
.005 ''c(::~
OLI ______ ~ ________ L_ ______ ~ ______ _L ______ ~
o 5 .10 15 20 25
a, deg
Basic wing
--0-- Ix VG-I@7]=0.50 --~-- 2 x VG-l @ 7]= 0.25,0.50 -----<>---- 3 xVG-l @ 7] = 0.25, 0.50, 0.75
LO[ .8 ,/ .6
CN
.4
.2
I.. ~-----' 10 15 20 25
a, deg
Figure 32.- Continued.
f'
'-"l
.006
.005
.004
.003
- - - - - - Basic wing
----<0 'lx VG:-l@'TJ=0.50 ---0--- 2 x VG,.l@7J= 0.25, 0.50 ----<)---- 3 x VG::'l@ 'TJ= 0.25" 0.50, 0.75
A
1\ I \ I \ I \ I \ I \
\ \ \ I 1 \ \
~
.00?4-~
.001
o 1 ,I
o 5 10 15 20 25
ex, deg
Figure 32.- Concluded.
94
95
the magnitude of the fluctuations in the C1 curve of the basic wing
are significantly reduced, indicating a more gradual and symmetric
growth of separation.
The performance of the vortex generator can be further enhanced
throu~h its use in multiple arrangements. Results of balant~ measure
ments On double (n = 0.25, 0.50) and triple (n = 0.25, 0.50, 0.75) VG-l
configurations are presented in figure 32. Generally, an increa~ing
number of VGls results in progressively higher low-a drag penalty (PO
and CA curves and Table III) but improved high-a performance. Drag
reductions of 17 percent with two and 24 percent with three VG-l IS
(relative to the basic wing) are obtained at a ~ 140. However, fUrther
losses of normal force are found,as expected, due to improved suppres
sion of flow separation and vortex lift. The longitudinal stability is
similarly improved progressively with increasing number of VGls, as
shown by the improved linearity of the Cm curve. Two VG-l IS have no
further i nfl uence on the pitch-up' at a = 130 but produce a more s'tab 1 e
configuration at higher a than a single VG-l. Three VG-lls result in
a more gradual initial pitch-down together with complete elimination of
instability., This is significant since the point of maximum percent
drag reduction now lies in a longitudinally stable a range. In
general, all VG arrangements eliminate the wing rock of the basic wing
in the intermediate a range.
As briefly noted in the BACKGROUND section, the formation of a
vortex at the swept-forward leading edge of the vortex generator isa
result of the prevailing sidewash velocity ahead of the wing leading
96
TABLE Ill. - VORTEX GENERATOR ZERO-a DRAGCHARACTERI STICS
Baseline
L.E. length reduction
Chord reduction
Diagonal cut
Wi th missil e
Without missile
{ { {
{
Device
Basi c wi ng
1 x VG-1
·2x VG-l
3 x VG-l
1 x VG-2
1 x VG-3
.1 x VG-4
1 x VG-5
. 1 x VG"6
1 x VG-7
1 x VG~8
1 x VG-.9
Percent increase in wetted area
over BW
-o .s
.9
1.4
.3
. 1
•. 2
.2
.3
. 1
·2 .~6
1.3
CD Peroentin creas e
0 in CD over BW 0
.'
0.01261 -
.01344 6.58
.01391 10.31
.01448 14~83
.01323 4.92
·."01312 4.04
.01315 4.28
.01327 5.23
. 01338 6.11
.01314 4.20
.01368 8.48
.01330 5.47
97
edge. At low a, this sideVlash velocity is not of sufficient magnitude
to cause vortex formation due to the toe-in angle of the device (see
fig. 5). When the sidewash is less than the toe-in angle of the VG, a
pressure drag is produced in addition to the skin friction drag of the
device. Thls accounts for the low-a drag penalty noted in the PO and
CA curves. Increasing a, accompanied by increasing sidewash velocity,
eventually leads to vortex formation at the VG leading edge. This
vortex travels over the wing upper surface and rotates in a sense so as
to induce a downwash velocity on its outboard side. Due to .the pre
vailing sidewash velocity, the upper surface path followed by this
vortex is angled toward the wing tip. Therefore, it should be noted
that any reference to inboard or outboard effects of the VG applies to
either side of this vortex rather than the device itself.
As shown by the separation boundary in figure 33, the vortex action
of a single VG-l at n = 0.50· results in substantial delays in local
separation on the outboard side. The extensive delay adjacent to the
VG is attributed to the exponential variation of the vortex tangential
velocity and, thus, the induced downwash. Slight delays of separation
inboard are attributed to the compartmentation effect of the device.
The sense of rotation of the VG vortex is opposite that of the primary
vortex, thereby reducing its strength and obstructing its inboard move
ment. Static pressure variations around the wing leading edge, in
figure 34, indicate that once inboa~d separation does occur (between 140
and 16° a at STA 3), the upper surface flow eventually stagnates at a
'T)
1.0
0.9
0.8
D.T.
0.6
0.5 L .
0.4
0.3
0.2
0; 1
oUtboard
98
- - - - -~ Basic wing
~ VG"'l @ 7]= 0.50 --(:: ..... VG-l @ 7,/= 0.50, flow·
stHlatta.ched. (!t highest test eX
~() n c; .' . _. ~ - ..... , .. 1 .. location of V{;
LE. . 'Separated
LE. Attached
--
o ! '"
o 5 10 15 20 25
Qsep,deg
Figure 33 • .;. Effect of the pylon vortex generatorbnleading-edge separation (C derived).
PLE
99
-4 -4 ------ Basic wing
ex = 140 --'0--- 1 x VG-l @ n = 0.50
Cp,3 CP4
STA 11 11 Y3('7}=0.45)
0 2 3 4 5 0 2 3 4 5 ...---4 ('7}:0.57)
X, em X, ·en:
-4,-- -4
ex = 16° -3"
-2
C, ~ >-,.:-0 C
P4 .3 , . , , -1 ,
o ~-
1 0 2 3 4 5 0 2 3 4 5
X, Cm X. em
-4r- -4
-3' ex = 18°
-2 C .
P.3 ~ CP4
-1
1 0 2 3 4 5 0 2 ~ 4
X, em X, em -5,--
~ -4,
ex = 23° -3
-2
C~ CP4 P3 \ "'"
-1 ' ,
0
1 0 2 3 4 5 0 2 3 4 5
X, cm X. em
Figure 34.- Local static pressure effects of the pylon vortex generator. Shaded regions indicate lower vortex contribution.
100
higher constant pressure than on the basic wing (see d = 180 and 230
plots) .. This results in a smaller contribution to leading~edge suction.
An aspect of the VG flow mechanism not yet discussed is the forma
tion of a vortex at the lbwer edge of the device, as shown in sketch D
of the BACKGROUND section. At the lowest angles of attack, this vortex .' ,
passes beneath the wing, undetected by the leading edge. At higher a,
howeVer, the vortex begins to impinge on the lower surface of the wing,
as indicated by the m.inor suction peak in the lower surface pressures
at STA 4 at a = 140 ,(fig. 34). This suction peak moves progressively
closer to the leading edge with increasing a, sug~esting that at angles
higher than those considered here .• this lower-edge vortex would pass,
completely over the wing. Once this occurs, its own contribution to the
wing leading-edge thrust (cross-hatched in fig. 34) is lost. In addi
tion, this vortex may act to degrade the upper VG vortex.
Local thrust variations for single and multipJe VG-l configurations
appear in figure 35. The high'-ci CA improvements noted in figure 32
are shown tc result from thrust enhancement outboard of each VG. The
figu:re also demonstrates the ability of m.ultiple VG"s to effectively
eliminate the adverse inboard effect of the single device. Note that
at low a (attached flow on the basic wing), local thrust values just
outboard of each VG (for all VG-lconfigurations) fall below those of
the basic wing. This is attributed to the induced downwash in these
reglons~ which effectively reduces the local angle of attack and, thus,
the sucti on forces around the leading edge. At higher a, however, thi s
~
.10
.08
.~---:--- Basic wing
----{)----- 1 x VG-l @ n = 0.50 _.-0--.- 2 x VG-l @ n = 0.25, 0.50 ..:... .. -V-.. - 3 x VG-l @ 11 = 0.25. 0.50, 0.75
.12
.10 l
.08
CT .06 locI
. CTIOC2 .06
.04 n = 0.20
.02
010'1' I I I---.J o 5 10 15 20 25
lX, deg
.10
.08 n = 0.45
CTIoc3 .06
.04
°tL o I· I o 5 10 15 20 25
lX, deg
.1°i .08~ n = 0.70 .il.
CTIOC5 '.06
.04
.02
OLI_~~_-L __ L-_-L_~ o 5 10 15 20 25
lX, deg
.04
.02
o I 0"1 o
.10
15 20 25
lX, deg
.08 n = 0.57 if; ,.
CTIOC4 .06
.04
o2~L 0~~1 I o 5 10 15 20 25
lX, deg
.1°i .08~ n = 0.82
CTloc6 .06
.04
.02
01 V/!
o 5 10 15' 20 25
lX, deg
101
Figure 35.- Effects of single and multiple pylon vortex generators on leadtng-edge thrust.
102
effect acts to delay separation and~thus, retain leading-edge suction
to the highest angles tested.
The spanwise leading-edge pressure distributions at Ci. = 140
(where the maximum percent drag reduction occurs) for single and
!l1ultiple VG-l configurations are presented in figure 36. The single
. VG:..: 1 curve s'hows only a local i zed ineffeCt i venes s on the i nboa rd side of
the device. In addition, the plots sho\'{how the mutual interaction of
multiple VG's acts to enhance the leading'-edge suction along the span.
The wavelike shape of the curves just outboard of each device is
attributed to the outboard drift of the VG vortex, discussed previously.
Compar'i son of the curves -indicates that the VG has a more pronounced
effect near the tip, where the flow has a tendency toward early separa
tion. This accounts for the substantial drag and longitudinal stability
improvernentswith the addition of a VG-l at n = 0.75 (triple VG-l
corifiguration). It is believed that two VG-l's located at n = 0.50
and 0.75 would provide almost the same leading-edge thrust enhancement
of the triple VG-lconfiguration tested nere.
Plots of CpLE versus CTloc at various spanwise positions for
the single VG-l configuration are presented in figure 37(a). The
linearity of the STA 4 curve at high Ci. is consistent with an attached
flow condition. The reduction in slope of the curve from its initial
value is att~ibuted to the suction peak produced by the lower VG vortex,
which additionally contributes to the leading-edge thrust but is not
detected in the CpLE measurements until it has passed over the leading
edge at higher Ci.. Without the suction contribution of this lower VG
...... -
C
103
- - - - - - Basic wing
-0-- lxVG-l @ 7]-0.50 ---0--- 2 x VG-l @ 7]= 0.25, 0_50 --.--0---- 3xVG-I@ 7]=0.25,0.50,0.75
-4.5
-4.0
-3.5
-3.0
-2.5~ NB
/ +------''-, PLE
-2.0
-1.5
-'1.0
.8 .9 1.0
7J
Figure 36.- Spanwise leading-edge static pressure distributions for single and multiple pylon vortex generator configurat~ons at a = 14°.
-5
-4
-3
-2
-1
0 0 0
I-
STA 1 2
3 4
5
0
0.02 -,
6
(a)
- - - - ..,... - Basic wing (all STNs)
0
---{o.---..,-----'l[)I----~O>---
a. ----1:1:33-----,
~---'lD~-
0 0
C Tloc
ST A 1 (7] = 0.20) 2 ( = 0.33) 3 ( = 0.45) 4 ( = 0.57) 5 ( = 0.70) 6 ( = 0.82)
0.02 0.04
Overall spanwise effect.
1 x VG-l @ 7] = O. 50
0.06
Figure 37.- Effect of the pylon vortE!X generator on leading-edge static pressure-thrust relationship. Selected high-a data points have been omitted for clarity. Angles in parentheses refer to basic wing data (see fig. 16).
~
C PLE
-5
-4
-3
-2
-1
o '/>(
o
.; ';
If
0.04
/
~ /.
Lower VG vortex th rust contribution
®/ /
/
~ Linear extension of / low-a portion of curve
--"-76 STA 4 (7]= 0.57)
® Data pOint recomputed based on elimination of lowe r VG vortex suction
0.06
C Tloe
0.08 0.10
(b) Local outboard effect of lower VG vortex.
Figure 37.- Concluded.
105
0.12
-... ,. ..... '.
106
vortex, therefore, the CpLE versus CTloc curve would continue along
its initial slope. To check this hypothesis,. the influence of the lower
surface suction Was removed by reintegrating to obtain leading-edge
thrust using lower surface pressures measured on the basic wing at STA 4
for a = 180 and 230. The resulting data points are plotted in fig-
ure 37(b), along with a replotting of the STA 4 data from part (a).
Indeed, these recomputed pOints are found to shift much closer to the
linear extension of the low-a portion of the curve, reflecting the sig
nificant thrust contribution from the lower-edge vortex.
In an attempt to all evi ate the low-a drag penalty associ ated with
vortex generators, possible reduction of the VG size without adversely
affecting high-a performance was investigated. The size reductions were
in the form of progressive lower-edge cut-off, which effectively reduced
the leading-edge le~gth of the VG, and back-edge cut-off, which reduced
the chord of the device. In addition, a diagonal cutback (in effect
removing the chordwise tip) was included as another means of VG size
reduction. Geometric details of these size modifications are presented
in .figure 5.
Zero-a drag data for single VG configurations utilizing the various
geometries tested appear in Table III. As expected, lower-edge (VG-l
to VG-2, 3) and diagonal cutbacks (VG-l to VG-6, 7) resulted in sub
stantial zero-a drag reductions. The initial chord reduction (VG-2 to
VG-4) also produced a low-a drag improvement; however, further cutback
(to VG-5) resulted in a drag increase. This anomalous result may be due
to the construction method used. As previously noted, an additional
107
thickness was used on the outboard side of the VG for added stiffness to
avoid excessive deflection. As sketched below, cutback to VG-5, thus,
resulted in a thick base, with the additional pressure (base) drag
appearing in the overall zero-a drag of the configuration.
L. E.
VG-2 Added thickness
./
./ ./
I.E.
L.E •
VG-5
T.E. VG-4
VG-2
Sketch J
Figure 38 shows the overall drag and longitudinal stability effects
of variations in VG leading-edge length (lower-edge cutback). A 30 per
cent reduction from VG-l (to VG-2) results in the appearance. of a CA
reversal at a ~ 13°. In addition, substantial drag increase and severe
pitch-up are indicated at high a. Further reduction in leading-edge
length (to VG-3; 60 percent reduction from VG-l) results in little
further effect on drag up to 16° a but a loss of effectiveness at
higher a~· This is also reflected by increased severity of pitch-up at
a ~ 190.
CA
Cm
·02.l
- - - - - - Basic wing
~ ---0- I'VG-I} Of-------~-- Ix VG-2. @ 7]=0.50 ----<>---- I x VG-3
I b -.02
~""7::~ ---
~--=:B -.04 '
-.06
I o 5 ,10 15 20 25
lX, deg
.02~~
.015
.010
.005
Xcm = 0.589 cr
? f //-- .... tp
/ f
)
, I
\J
o LI ____ L-____ ~ ________ ~ ______ ~ ________ ~
o 5 10 15 20 25
lX, deg
IDS
Figure 3S.- Performance effects of pylon vortex generator leading-edge . 1 ength reducti on.
109
The primary function of the VG leading edge is to provide a sl'larp
edge along which the VG vortex may form and build in strength before
passing over the wing. A decrease in the length of this edge, there
fore, would be expected. to reduce the vortex strength. The-accompanying
reduction in downwash on the outboard side then results in earlier
separation. This effect is evident in the static pressure distribu
tions around the wing leading edge at a = 140 (STA's 5 and 6), in
figure 39(a). Figure 39(b) presents pressure distributions just out-
board of the VG (STA 4) at a = 16° and 190. With VG-3, there is no
evidence of the lower-edge vortex suction on the lower surface pressures
as was the case with VG-l. The lower vortex may, thus, be passing over
the wing leading edge at a lower a. The opposite sense of rotation of
this lower vortex would induce an upwash velocity outboard, leading to
earlier separation; in addition, the associated suction peak on the
lower surface and its contribution to the leading-edge thrust (shaded in
fig. 39(b)) is lost. A second advantage of a long VG leading-edge
length may, thus, be its ability to keep the lower VG vortex below the
wing and acting near the leading edge to higher a.
The effects of VG chord reduction on drag and longitudinal sta
bility are shown in figure 40. For structural reasons, VG-2 (30 percent
leading-edge length reduction from VG-l) was used as the baseline geome
try for analyzing this parameter. Axial force indicates that a25 per
cent reduction in VG-2 chord (to VG-4) produces a drag improvement
beyond 12° a, in addition to the low-a drag reduction noted earlier.
In addition, the severe pitch-up at a '" 19° of the VG-2 has been
Cps
~
-1
o
7] = 0.70
--D---rl LS-- ___
--0
- - - - - - Basic wi ng
---0-- lxVG-l @ '7=0.50
__ ~-- 1 x VG-3 @ '7= 0.50
VG-l
1 LI ______ ~ ____ ~ ______ _L ______ L_ ____ ~
o 2 3 4 5
X, em
-4, -3~ 7] = 0.82
Cp6
-1
o
1 o 2 3 4
X, em
(a) a = 14°.
110
5
Figure 39.- Outboard static pressure effects of pylon vortex generator leading-edge length reduction.
CP4
- - - - -- Basicwing
a z 160 . -0-- lxVG-1 @ 1]=0.50
---Q--- lxVG-3 @ 1]=0.50
VG-l
1 IL-__ -L ____ ~ __ ~ __ --~--~
o 234 5
X, em
\
~ a -190
~ C
P4
o .1 2 3
X, em
(b) a = 16° and 19°.
Figure 39.- Concluded.
III
4
.02
~
o
-.02
CA
-.04
-.06
o
.02 ~v.,..;i(J
.015
.010
Cm
.005
00
5 10
Xcm =.0.589 Cr
5 10
- ~ - - - - Basicwing
-----0-- 1 x VG-2} ---D---" 1 x VG-4 @ '1]= 0.50 ----(>---- 1 x VG-5
'-
15 20 25
ex, deg
15 20 25
ex, deg
Fi gure 40. - Performance effects of pylon vortex generator chord reduction.
112
113
eliminated. Further chord reduction (to VG-5; 50 percent reduction from
VG-2), however, results in increased high-a drag and reappearance of
pitch-up at a ~ 190. Local thrust variations in figure 41 indicate
that these trends result primarily from f16w modifications outboard of
the device. Substantial high-a thrust enhancement is indicated at
STA's 5 and 6 with the initial chord reduction (VG-2 to VG-4), but this
improvement is lost with further cutback (to VG-5). Therefore, there is
a specific VG chord within the range tested which will produce the
optimum combination of low- and high-a performance.
The original design of the vortex generator presumed that the flow
mechanism depended totally on the vortex formed at its leading edge and.
passing over the wing upper surface. However, it now appears that the
vortices formed at the lower and back edges of the device may be of
importance. As the VG chord is reduced, the proximity of these edges to
the VGleading e~ge eventually becomes s~ch so as to cause interference
between corresponding vortices. In addition, at a high enough a, these
lower- and back-edge vortices may pass completely over the wing leading
edge and interfere with the upper surface flow. Specific effects on
performance would seemingly depend on' the. strength of these vortices,
which depends partially on the length of the corresponding edges. How
ever, at this stage there are insufficient data to provide definite con
clu~ions on the VG flow mechanisms producing the trends observed with
VG chord reduction. The geometries tested, however, did provide a
general idea of the VG proportions required for optimum performance.
·.;.,.
.10
.os
"'.' Crloe1 .06
C Tloe}
.04
.02
01.0'1 o 5
.10
. 0sL
.06
.04
.02
.;: .... ,
= 0.20
10 15 .20 ·:;25.
IX, deg
n = 0.45
I /i .
01 QI o 5 10 15 20 25
IX, deg
.10
.OS n = 0.70
CTloe5 .06
.04
.02
0 0 5 10 15 20 25
IX, deg
- - - - - - Bas i c wi ng
··~.lx VG,..2} '. -. .......u:-.- 1. x VG-4@Tl=0.50 -:-;-.. -:<>-.. - l' ,x VG-5
.10
.os n = 0.33
CTloe2 .06
.04
.02
0 0 '5 10 15 20 25
',12~ IX, deg
.10 .
T .OS
C . ,Tloe4 .06
,04
.02
0 0 5 10 15 20 25
:10
.os
Cr ' ' . . Ci6-loe6
,04
.02
01 v'l o 5
IX,deg
n = 0.82
10 15 20 25
IX, deg
,Figure 41.- Effects of pylon vortex generator chord reduction on leading-edge'tnrust.
114
115
Performance effects of simultaneous cutbacks of the lower and back
edges of the VG, in the form of diagonal cuts (see fig. 5), will now be
considered. Figure 42 presents force and moment data for single VG-l,
6,and 7 configurations. Axial force indicates a reduction in suction
effectiveness beyond 130 a with increasing cutbackr In addition, the
initial cutback (VG-l to VG-6) results in pitch-up at a ~ 190, which
becomes more severe with further reduction (to VG-7). Inspection of
local thrust variations, in figure 43, reveals that these effects on
high-a performance are again attributed to outboard flow modifications.
The losses resulting from cutback to VG-6 are most likely attrib
uted to an effective elimination of the lower edge of the VG. This
eliminates the lower surface suction peak produced by the lower-edge
vortex and, thus 1 its contribution to the leading-edge thrust. Another
possibility is that the proximity of the leading and back edges near the
tip of the VG-6 may cause a weakening of the VG leading-edge vortex as a
result of interference with the counter-rotating back-edge vortex. The
additional loss of effectiveness with cutback to VG-7 is believed to be
primarily the result of leading-edge length reduction. As previously
noted, this reduces the strength of the VG leading-edge vortex by
reducing the distance over which it forms.
The possible use of pylon vortex generators also as carriers of
slender external· stores (such as air-to-air missiles) will now be con
sidered. The VG tested was sim~lar to VG-l but had an extended chord
in order to provide a mounting position for the wooden dowel si~ulating
the external store (see fig. 6). This device was tested with (VG-8)
CA
Cm
.02
-0
-.02
-.04
-.06
~
Basic Wing
~ lXVG-l} ~~-o--- 1 x VG-6 @ 7]=0.50 _ .. --"<)-._- 1 x VG-7
~.~
o 5 10 15 20 25
.02~
.005
Xcm = 0.589 cr
Lx.; deg
o LI ___ ~ ______ ~L-______ L-____ ~~ ______ ~
o 5 10 15 20 25
a. deg
Fi gure 42. - Performance effects of pylon vortex generator di agona 1 cutback.
116
117
STA ------Basic wing 1
~1 x VG-11 2 3 ----0---1 x VG-6 @ 11 = 0.50
4 ----<>---- 1 x VG-7
.10..- '\ /5 .10 6
.08 LJ .08 L- n = 0.33
C cT T1oc1 .06
loc2 .06
.04 = 0.20 .04
.02 .02
0 0 0 5 10 15 20 25 0 5 10 15 20 25
a, deg .12,- a", deg
.~ .10..- .10 I-- /; /j
n = 0.45 .08~ >5;'
.081-- n = 0.57 b CTloc3 .06
C " Tloc4 .06
.04 .04
.02 .02
01 01' 0 0 5 10 15 20 25 0 5 10 15 20 25
a, deg a, deg
.10 .10
.08 n = 0.70 .08~ n = 0.82
CTIOC5 .06 C T loc6 .06
.04 .04
.02 .02
0 0 0 10 15 20 0 5 10 15 20 25
a, deg a, deg
Figure 43.- Effects of pylon vortex generator diagonal cutback on leading-edge thrust.
, '
'. i
118
and without (VG-9) the store in single VG configurations (at n = 0.50).
Table III indicates a low-a drag penalty due to the chord extension and
store addition; however, this is not of concern since a drag increase , .'
would in any case be obtained with a pylon. Selected results of farce
and, momen't measurements on VG-8 and VG-9 canfi guratians are presented
and campared with the baseline VG~l configuration (at n= 0.50) in
figure. 44. The CA curve indicates a loss of drag perfarmance beyand
120. C/. with VG-8 as campared wi th VG~ L Stati c pressure data~ nat pre-
sented, indicate that the external stare prevents the farmation of a
vortex at the lawer edge af the VG, thereby reducing the lawer surface
cantribution to' the wing leading.:.edge thrust. However, a, significant
high~C/. drag advantage aver the basic wing is s:till realized. In addi
tion, VG-8 delays the pitch-up af the basic wing by approximately 20.,
to c/., ~ 130 ~
Figure 44 shows that once the slender external store has been
re 1 eased, the, extended, chard VG:-9 conti nues to perfarm as a drag
reducer. Hawever, the drag reductian effectiveness af the device
appears to' have been hampered by the chard extensi on (from VG-l).
Sur,face pressure data indicate that this may be a result af the per-
sistence af the lawer-edge vortex along the extended edge, eliminating
its suctian effect an the wing leading~edge region. However, the
resulting drag reductian would, still be an improvement aver a conven-
~ianal pylon canfiguratian, which would cantinue to' praduce a drag
penalty even at higher C/.. Further research should be performed an
madificatians that will allow far, therealizatian of a greater portian
~
.02
o
-.02 --
CA
-.04
-.06
o
Cm
.005
0 0
---0-----0----_.-(>-._-
1 ·1
Basic wing
1 x VG-l} 1 x VG-8 @ 7]=0.50 1 x VG-9 .
.~
119
5 10 15 20 25
Xcm = 0.589 Cr
5 10
n, deg
15
n, deg
7""'-< (/ / ( ( / ( (,
,\. __ -- VG-8 w/mi ssile i
.... "'// VG-9 w/out missile
20 25
Figure 44.- Performance of an extended chord pylon vortex generator' utilized as a carrier of slender external stores.
120
of the VG drag-~eduction potential, both with and without the external . .,:
store. Variations in VG chord length and chordwise location of the
s~ore along the lower edge of the VG may be initial steps .
. In summary, the pylon vortex generator produces substantial
improyements in high-a drag and longitudinal stability when utilized on
highly swept leading edges. The lower VG vortex apparehtly plays an
important role through its own contribution to the leading-edge thrust
and, thus, should be considered in the design of the VG shape. The
mutual interaction of multiple VG's acts to further enhance the leading-
edge suction along the span. However, a low-a drag penalty is charac
teristic of the vortex generator. VG size reductions in the form of
leading-edge length and diagonal cutbacks reduce the low-a d~ag penalty
but also result in a loss of performance at high a. Reduction in VG
chord to a certain degree produces performance improvements at both low
and high a. The vortex generator has potential also as a carrier of
slender external stores. Although the performance of the extended chord
VG is not as efficient as that of the baseline VG-l, substantial drag
and longitudinal stability improvements over the basic wing are produced
both with and without the external store.
121
Sharp Leading-Edge Extension
The common basis of the devices discussed thus far has been that of
maintaining attached flow at the wing leading edge in order to retain
leading-edge suction to higher a. An alternative to this conventional
approach of drag reduction is provided by the sharp leading-edge exten
sion (SLEE), or vortex plate, which manipulates the natural tendency
toward flow separation and vortex formation at a swept leading edge.
A sharp-edged plate lying in the plane of the wing lower surface and
projecting ahead of the wing leading edge forces separation along its
leading edge and subsequent vortex formation. Ideally, this vortex is
maintained just ahead of the wing leading edge along the entire span,
with its associated suction acting on the leading-edge thickness to
generate a thrust force. Flow reattachment just aft of the leading
edge curvature helps to maintain attached chordwise flow on the wing
upper surface (see sketch A in INTRODUCTION).
Balan~e data for a SLEE configuration utilizing an 0.71 cm exten
sion (see fig. 8) is presented in figure 45. This device had a spanwise
coverage of n = 0.25 to 0.93 and was tested with an F-2 fence at its
inboard edge (n = 0.25). The fence was intended to obstruct the span
wise boundary layer flow originating b~tweeri the wing apex and the
inboard edge of the SLEE, to allow the formation of a clean SLEE vortex.
This fence was shbwn in previous testing (ref. 5) to delay a mid~a
longitudinal instability produced by the SLEE and to maintain CA
improvement to significantly higher a. At low a, the vortex suction
is insufficient to offset the additional drag of the SLEE, accounting
--- ..... "-
o ~
" , ~
" -.02
CA
-.04
-.06
o 5
.04
o
-.04 em
-.08
-.12
o 5
,
-- - -- - Basic wing
---0-- 0.71 em ext. SLEE; wI F-2 @ 7]= 0.25
I· I· 10
a. deg
I
15 20
x . ~ = 0.474
cr
'-
"
25
'-
10 15 20 25
a, deg
122
Figure 45.- Force and moment characteristics of a sharp leading-edge extension configuration.
eN
LID
1.0
.8
.6
.4
.2
10 I
./ /
\ \
\ 8
6 11 4~ ,I I
2
a, deg
\
/
/ I
/ /
/
/ /
/
/ /
/
- - - - - - Basiewing --0- 0.71 em ext. SLEE;
wI F-2 @ 7]= 0.25
"-"-
"-"-
"-"-
OLI ________ L-______ ~ ________ L_ ______ _L ______ ~
o 5 10 15 20 25
a, deg
Figure 45.- Continued.
123
C'Z
.006
.005
.004
.003
.002 , .
. 001
- - - ....;; ~......;. ~asic wing ~,O.71cm ext. SLEE;
, I I
!
I
I
wI F2 @7J= 0.25
A
1\ 1 \ 1 \ I \ I 1
- \
\
\ .\
I I I
OLI ______ ~ ______ ~ ________ L_ ______ L_ ____ ~
o 5 10 15 20 25
a, deg-
Fi gure 45. - Cone 1 uded.
':.'
124
125
for the net CA increase. Therefore, to avoid a cruise drag penalty,
the SLEE should be designed as a retractable device. Beyond a ~ 12°,
the vortex strength and its optimum position produce substantial CA
improvements over the basic wing. The ability of the SLEE to maintain
its vortex ahead of the leading edge is also reflected in delayed onset
of vortex lift, reducing the CN beyond a ~ 10°. From a performance
standpoint, the low-a drag increase results in severe losses in LID.
Beyond 120 a, however, substantial improvements are noted, with the LID
increment tapering off gradually with increasing a. In addition, the
rolling moment instability of the basic wing between gO and 140 a is
eliminated. The longitudinal stability of the configuration is slightly
r~duced fro~ that of the basic wing, but the linearity of the Cm curve
is significantly improved. Here, the moment reference center has once
again been selected to better show up the effects of the device. The
new location (see DATA REDUCTION) is 3.82 cm (5 percent of root chord)
aft of the original moment reference center given in figure 2.
Static pressure distributions around the wing leading edge for the
0.71 cm ext. SLEE configuration at a = 120 are presented in fig-
ure 46. Attached leading-edge flow inboard of STA 6 on the basic wing
results in high suction peaks near the leading edge with subsequent
pressure recovery on the upper surface. However, the vortex-induced
reattachment due to the SLEE produces a stagnated flow condition, with
elimination of the peak negative pressures near the leading edge, but
with high negative pressures over a greater portion of the leading-edge
curvature. Note, especially, the high negative Cp on the lower
STA 1
2
- -'- - - -'Basi ewing
3 4
-0-----, Lo~er surface} 0.71 cm ext. SLEE; --'-'-0-----' Upper surface wI F-2 @ Tl = 0.25
- -4
-3
5 6
n =,0.20
CP1
-1
11 1 1 I I 0 2 3 4 5
-I x, em
-3 , n 0.4-5 \
\
-2~\ C I ",
P3 \ .--n..:--_
or "-1 _ I---~I----I----[
0 2 3 4 5
x, em
-I -3
, n = 0.70 \ \
-1: Cps \ ',---- ____
-1 \ \ ,
~r " ---------
0 2 3 4 5
X, em
Cp:z
CP4
:, CP6
-4
-3
1
, \
0
-4~
-3
k _ -2
-1
f" 0
-4
-3
I -2
0
1 0
n = 0.33
2 3 4 5
x, em
n = 0,.57
------------
2 3 4 5
X, em
n = 0.82
--------
-----------
2 3 4 5
X, em
126
Figure 46.- Effects of the ~harpleading7edge extension on static 0 pressure distributions around the leading edge at a = 12 .
127
surfa~e in contrast with the positive values measured on the basic wing.
With the 6asic wing at a lifting condition, the stagnation point occurs
just below the leading-edge curvature. The associated positive pressure
acting on the forward sloping surface (see sketch K, below) contributes
a positive axial force and, therefore, drag. On the other hand,-when
the stagnation point lies on the lower surface of the SLEE, the positive
pressure produces no axial force (sketch K). In addition, the high flow
velocity at the SLEE edge develops a high suction level to generate a
thrust force.
~
Basic wing wI SLEE
Sketch K
Leading-edge pressure variations, in figure 47, indicate that at
low a the leading-edge suction produced by the SLEE vortex is sub
stantially less than the potential flow suction of the basic wing.
However, once potential flow is lost, the SLEE eventually provides an
improvement over the basic wing (beyond a ~ 150 at n = 0.724,
a ~ 190 at n = 0.416). Static pressure distributions around the wing
C PLE
C PLE
-3 r-
-2 t -1
/
I ~~ 0
0 5
-3
-2
-1
o \ l o 5
128
Orifice 123 - - - - - - Basic wing
0 0.71 cm ext. SLEE; 221 wI F-2 @ 7] = 0.25
~ "'"' Orifice 123 (7]= 0.416)
/ ., / \
I '--
I /
/ /
/ /
/
10 15 20 25
a,deg
;---, I \
Orifice 221 (7] = 0.724)
/ \ / \
10 15 20 25
d, deg
Figure 47.- Leading-edge suction effects of the sharp leading-edge extension.
129
leading edge at ~ ~ 220 (fig. 48), indicate that the greatest contribu
tion to the leading-edge thrust improvement comes from the low pressures
induced on the lower surface. The accelerated loss of thrust at STA 1
may be attributed to the unsweeping of the upper surface isobars by the
fence (see Fence section). The spanwise CTl distributions, in. oc figure 49, are consistent with the CpLE data just presented. At
a = 120, the leading-edge thr0~t produced along the SLEE falls below
that of the basic wing; however, by 230 a, substantial improvements are
noted along the entire length of the device. Decreasing thrust incre
ment toward the tip suggests an expansion or diffusion of the SLEE
vortex core due to increasing upwash and entrainment effects. Means to
alleviate the loss of SLEE effectiveness due to vortex diffusion will
be discussed later.
Optimizing the SLEE extension \>li11 now be considered as a possfble
means of further performance improvements. The possibility of such
improvements is suggested by the fact that over-extension of the SLEE
would result in reattachment on the device itself, with local suction
peaks and the possibility of separation near the \!Jing leading edge. In
addition, the vortex suction would mainly act in the normal direction,
with little effect on axial force. This condition is depicted in
sketch L, below:
~
Cp1
CP3
Cps
STA 1 1 ------ Basic wing
3 4
5 6
-----cr----. _. Lower surface} 0.71 em ext. SLEE; ---~--- ~pper_surface wi F-2 @ n = 0.25
-5
-4
-3
-1
o
, ,
n = 0.20 --0---0-------0
1 LI ____ ~ ____ L_ __ -J ____ _L~~
o 2 3 5 4-
-4 X, em
-3 n
-1 ,
o ", [- "
1 I - I
-4
-3
-2
o
o
"' ,
2 3 4
X, em
n = 0.70
5
1 LI ____ -L ____ ~ ____ ~ ____ L_ __ ~
o 2 3 4 5
X, em
--'4 n = 0.33
-3
---------------'0::.:- --2 r- ~---o----_o c '-. P2 \
CP4
CP6
-1
o , ,
1 IL-____ L-____ L-____ L-____ L-__ ~
o 2 3 4 5
-4 X, em
-3 n = 0.57
---0-----
-1 , "
o
1 LI _-'--'-_--'-----'-_-'-__ '--_-'
o 2 3 4 5
X, em
-4
-3 n = 0.82
-2
-1
o
1 I'--_~ ____ _L ____ ~ ____ L-__ ~
o 2 3 4 5
X, em
l30
Figure 48.- Effects of the sharp leading-edge extension on static pressure distributions around the leading edge at a = 22°.
C Tloc
CT loc
.10 - - - - - - Basiewing
.08
--0--- 0.71 em ext. SlEE} I
---0--- 0.48 em ext. SlEE wi F-2 @ "l= O. 25 _ .. -(>-.. - 0.23 em ext. SlEE
- .. ---ts-.--- 0.00 em ext. SlEE
.06 a'12°
.04
.02
O~I ----~ ____ L-__ ~ ____ _L ____ ~
o
.12
.10
.08
.06
.04
.02
.2 I
\1 ,
~ SlEE apex;
F-2
.4
7J
" \ \
\ \
.6 .8
.. '~
" I ,
" "
,~
1.0
a = 23°
I I I OL- .1..6.8 1.0 o .2 _ I
7J
Figure 49.- Effects of SLEE extension on spanwise distribution of . leading-edge thrust.
131
( 132
Over-extension Under-extension
Sketch L
However, considering the range of extensions tested here, the over
extended condition is unlikely to be a factor except in a very narrow
a range. Nev~rtheless, this conditibn does provide a conceptual upper
bound for the SLEE extension. On the other hand, under-extension of the
SLEE may not allow the stagnation point to move to the SLEE even at
high a, leading to a large drag contribution, as discussed previously
(see sketch L). Therefore, the minimum reguirement for the SLEE exten
sion would appear to be that it project ahead of the stagnation point at
the highest a of interest.
Selected balance data for corifigurationsutilizing SLEE extensions
ranging from 0.71 to 0 cm (with an adjacent F.:..2 fence at the SLEE apex)
are presented in figure 50. Little effect from SLEE extension is evident
in CA up to a~· 160 . At higher a, how~ver, decreasing extension
results in progressively improve8 axial force characteristics, with the
CA
em
.0
o
-.02
-.04
-.06
o 5
.04
o
-.04
-.08 t·· -.12
. 0 5
10
---10)-------0---
----<>----_.---6---.. -
a, deg
"'
15 20
Xem = 0.474 c -r
"'-~
lO 15 20
a, deg
Basic wing
0.71 em ext. SLEE} 0.48 em ext. SLEE I
w F-2 0.23 em ext. SLEE
0.00 em ext. SLEE
25
I 25
Figure 50.- Effects of SLEE extension on performance.
133
@ 7]= 0.25
134
improvements continuing down to zero extension. In addition, the longi
tudinal stability of the configuration is slightly improved. Although
rtot presented here, effects on normal force and lift-to-drag are minor.
Comparison of static pressure distributions around the wing leading
edge for the 0.71 and 0 ·cm SLEE extensions (fig. 51) indicates a delay
in onset of vortex suction on the leading edge as the SLEE extension
is reduced. At. ~ = 120, the 0.71 cm ext. SLEE configuration is
.characterized by stagnated leading-edge flow; whereas with the 0 cm
extension, attached flow still persists on the upper surface. Appar
ently, the vortex size with zero extension is such so as to produce a
stagnated condition only on the lower portion of the leading-edge region
(see STA 5 inset sketches). This delay is shown in figure 52 to payoff
in improved leading-edge thrust at higher ~. At the highest angles
tested, a distinct advantage with reduced SLEE extension appears only
at STA's 5 and 6. However, at higher ~ngles, thrust improvements would
also be expected at the remaining spanwise stations, as the SLEE vortex
moves inboard. While the adverse inboard effect of the fence is again
seen at STA 1, the favorable outboard effect at STA 2 allows for con
tinued thrust improvement to the highest angles tested.
As already noted, the purpose of locating a fence inboard of the
SLEE was to aid in the formation of a clean SLEE vortex. The possi
bility of. substituting a chordwise slot (w/SC-l) for the fence was
investigated on the 0.48 cm ext. SLEE configuration. Selected
results appear in figure 53. Axial force is slightly improved in the
mid-~ range with the use of the slot; however, early loss of slot
CP1
c P3
Cps
STA 1
------ Basie wing
2 3 ~ 0.71 em ext. SLEE; wi F-2 @ n = 0.25 ---~--- 0.00 em ext. SLEE; wi F-2 @ n = 0.25
-4
-3
1 0
~:~ \ \
-2$:::
4 5
6
n = 0.20
-=---
2 3 4
X, em
n 0.45
Of-- '" ; I 1 -- - -,- - --,- - - -I
0 2 3 4
X, em
5
5
~ ~
~ ~\, \
' ''ri i '----_ n = 0.70 -1
~~~,~~--i-0 1 2 3 4 5
X, em
-4
-3 n = 0.33
CP2
1 0 2 3 4
-4~ X, em
-3 " r""7 , , n = U.Of
CP4
o ',_ 1'2- -o-'--D-:~:~-1· I ----1-----1-----1
o 2 3 4
X, em
~:I -2 b
. n = 0.82
C"_1~-~ \-y_---D---- ---o-~-
O~------------------1 1 1 1 1 o 2 3 4
X, em
135
5
5
5
Figure 51.- Effects of SLEE extension on static pressure distributions around the leading edge at a = 12°.
,
136
STA' ------Basie wing
1 ----cr- 0.71 em ext. SLEE; wI F-2 @ T) = 0.2.5 .·,2 ---G--- 0.00 em ext. SLEE; wI F-2 @ T) = 0.25
3
''[ ;8 4 5 ! .10r- ~../6 .10
.,
~ , n = 0.33 .os f---
, .os
CT .06 C .06~ )}'-/' loc1 Tloc2 I
.04 0.20
.04
.02 .02
0 0 0 .5 10 15 20 25 0 '='5 10 15 20 25
.12 r- a, deg p
a, deg
.10 //0 .10
.OS .OS
C .06 C .06 Tloc3 , / Tloe4
.04 .04
.02
/1 .02
01 0 0 5 10 15 20 25 0 5 10 15 20 25
a, deg a, deg
'T .10
.OS n = 0. 7Or:: .0sL n = 0.82 ---0
C .06 CT .06 Tloe5 loe6
.04 .04
.02 .02
0 0 0 5 10 15 20 25 0 5 10 15 20 25
a, deg a, deg
Figure 52.- Effects of SLEE extension on leading-edge thrust.
CA
LID
.0
o
-.02
-.04
-.06
o 5 10
':[ 6
4
2
137
- - - ---, - - Basic wing
-0-- 0.48 em ext. SLEE; w/F-2 @ 7]=0.25
---D--- 0.48 em ext. SLEE; w/SC-l @ 7]= 0.25
"
15
a, deg
_-0
20
'- ' - "
25
O~I ___ ~ __ -~---~---~---~ o 5 10 15 20 25
a, deg
Figure 53> Performance comparison of the fence and chordwise slot at theSLEE apex.
~:.
138
effectiveness (see Chordwise Slots section) results in a sudden loss of
SLEE performance beyond a ~ l~o. The same trends are evident in lift
to-drag. Although not presented, no significant effects are noted in
pitching moment. The chordwise slot, therefore, seems to be slightly
more effective than the fence in obstructing the spanwise flow near the
wing apex up to approximately 160 a. However, highly maneuverable air
craft encounter a wide range of angles of attack and, thus, the high-a
performance capability with the fence may be well worth the small
penalty at mid a.
As noted previously, the tendency of the SLEE vortex to expand and
migrate onto the wing surface will ultimately 1 imit the drag-reduction.
effectiveness of the device. The previous success of fences and chord
wise slots in compartmentation of the wing leading edge suggested that
their use along the SLEE would limit the growth of the SLEE vortex and,
thus, improve its outboard effectiveness. Each device (F-2 and slot
w/SC-l) was, thus, alternately tested at n = 0.625 (midpoint of SLEE)
on a 0.48 cm ext. SLEE configuration also utilizing an F-2 fence
at n = 0.25. Selected force and moment data appear in figure 54.
While the additional F-2 fence at n = 0.625 produces the expected
low-a drag (or CAl penalty, substantial improvement in SLEE suction
effectiveness is indicated beyond a ~ 16°, with no adverse effects on
longitudinal stability. Tuft photographs in figure 55 support the view
that this improvement is due to a compartmentation of the SLEE. In
interpretation of these photographs, a herringbone tuft pattern repre-
- sents a vortex core (shown as a dashed line), whereas a divergence of
c -A
Cm
D
-.02
-.04
-.06
-.08
o
.04 r-::
o
-.04
-.08
-.12
o
5 10
I 5 10
139
- - - - - - Basic wing
--0- 0.48 cm ext. SLEE; w/F-2 @ 7]=0.25 ---0---- 0.48 em ext. SLEE; w/F-2 @ 7]= 0.25, 0.625 - _.-<)----- 0.48 cm ext. SLEE; wI F-2 @ 7] = 0.25 and
"-
I J
IX, deg
15 20
Xcm = 0.424 c r
15 20
a, deg
25
25
- 5C-1@ 7]= 0.625
Figure 54.- Performance effects of the fence and chordwise slot in combination with the SLEE.
Figure 55.- Upper surface tuft photographs of 0.48 cm ext. SLEE with F-2 at n = 0.25 configuration with and without an F-2 fence at n = 0.625 (a = 140).
~
141
adjacent tufts indicates a flow reattachment line (solid line). In the
absence of the ~utboard fence, the SLEE vortex is free to expand, having
migrated completely onto the wing surface at a = 14°. A fence at
n = 0.625, however, splits the SLEE vortex into two smaller vortices;
one emanating just outboard of the fence and the other near the wing
apex. Thi s 1 imits the growth of the outboard voritex and keeps the
leading-edge region near the tip under the influence of vortex suction
to higher a. The inboard vortex is, subsequently~ forced onto the wing
surface by the fence and acts as a barrier to the streamwise migration
of the outboard vortex.
The effect of this additional F-2 fence at n = 0.625 on the
leading-edge suction is depicted in figure 56. At a = 160, a slight
suction improvement outboard of the fence roughly balances the charac-
teristic loss of suction on the inboard side. However, at a 23 0, a
substantial increase in the outboard thrust increment accounts for the
net improvement noted in the CA curve.
The chordwise slot, on the other hand, is shown in figure 54 to be
ineffective in the role of SLEE compartmentation. The spanwise iocal
thrust distribution at a = 160, in figure 56, shows signs of compart
mentation through a slight thrust increment just outboard of the slot.
However, by 230 a, this improvement has vanished due to loss of slot
effectiveness .. This slot at n = 0.625 was also tested in combination
with a slot (as opposed to F-2) at n = 0.25 . on the 0.48 em ext.
SLEE configuration, but with the same results (not presented).
- - -' - -- Basiewing
.10,-~ 0.48 em ext. SlEE; wi F-2 @ 7]' 0.25 _.--0-.- 0.48 em ext. SLEE; wi F-2 @ 7]'0.25,0.625 _._--<>-_.- 0.48 em ext. SLEE; wI F-2 @ 7]' 0.25 and
SC -1 @ 7]. O. 625
.08
I ~/rM'o 06t ' t ' C
\"'L, Tloc
.04 a -160
"-"- ,
I .02
0 0 .2 .4 .6 .8 1.0
I I 7}
I .12
I 110
.10f--- I~~, I I
I .08
I L~, / "- IB -
a' 230 06[ /
, 0
\'
CT \
\
Icc , .04 "- ,
F-2
g~>' F-2 02r <> F-2 SC-1
0 0 .2 .4 .6 _8 1.0
7}
Figure 56.- Effects of the fence· and chordwise slot on leading-edge ·thrust distribution along the SLEE.
142
· 143·
In summary, the sharp leading-edge ~xtension is effective in drag
reduction and rolling moment stabilization at high a, with no adverse
effects on longitudinal stability. The drag reduction is insensitive
to SLEE extension at low and mid a; however, there is a distinct
advantage to having a shorter extension at high a, assuming the exten
sion is beyond the stagnation point. When utilized at the apex of the
SLEE, the chordwise slot is slightly more effective than the fence in
obstructing the spanwise flow near the wing apex at mid a. However,
far better performance at high a makes the fence more desirable for
use on highly maneuverable airciaft. Unlike the chordwise slot, the
fence also possesses the ability to compartment the SLEE when used along
its length, thereby retaining leading-edge suction to higher Ci,.
_" ~ • , •• ~~~. 110_
144
.. Lea:di ng- Edge Vortex Flaps
As an ext"erision of thE{ SLEE concept; 1 eading-edge vortex fl aps were
tested as a possible means'6f obiaining furtnerimprovements in leading-
edge thru~t at high a. This device is similar in appearance to a con
ventional leading-edge flap which is deflected to align it w5th the
local upwash in order to provide a smooth onflow at the wing leading
edge and, thus, av6id5eparation. The vortex flap, however, relies upon
a separation vortex forced~by underdeflection relative to the oncoming
flow to generate a suction force on the flap and, thus, provide a thrust
component. Ideally, this vortex induces reattachment just aft of the
leading-edge curvature, thereby allowing attached flow on the upper
surface. The thrust component generated on the vortex flap, thus, leads . to" alleviation of the lift-dependen't drag (see sketch A in •
INTRODUCTION) .
The flow mechanism of the vortex flap will now be discussed in
detail with respect to the 300 downward-deflected (+) VF-3. This was a
constant chord, flap (flap chord 7.3 percent of mean geometric chord)
beginning at n = 0.2~ and extending out to the wing tip (see fig. 7).
Elimination of the first ~~ percent of the flap near the wing apex
(n = 0 to 0.25) was based on its relative ineffectiveness due to low
prevailing upwash in that region (ref. 3). Figure 57 presents wing
leading-edge static pressure variations at selected spanwise positions
for a configuration utilizing a 300 + VF-3. As an aid in the discussion,
a qualitative summary of the various stages of leading-edge flow
development with iricreasing ~,at constant deflection angle, is also
C PLE
Orifice 117
_ 205
227 ©zpm
-2.0
-1. 5
-1. 0
.
~
~\ ."
:9 .. \\i
\ I
-u ,>I I
-0.5 /1D1 .... q? I c( \0 I
, /
~~-~ ~\ ''0; q
o J \ I
o 5 10 15
a, deg
p--
117 ("1= 0.33)
--0- _ P 227 ("1 = 0.82)
@
20 25
145
Figure 57.- Leading-edge static pressure variation and flow develop~ent on a 30° + VF-3 configuration. Stages of flow development are arbitrarily indicated on the orifice 117 curve.
146
included. Due to entrainment effects and the spanwise upwash distribu
tion ahead of the leading edge, these successive stages of flow develop
ment will also be found with increasing outboard distance for constant
flap deflection and angle of attack. At low a (stage A),the vortex
flap is ~ffectively overdeflected relative to the small upwash and,
thus, induce~ ~6rtex formation on its lower side. The resulting orien
tation of the flap normal force vector produces a drag increase and loss
of lift. Eventually, a smooth onflow at the flap edge (stage B) occurs,
as the flap angle matches the upwash. Note that to this stage, CpLE
has become more negative with increasing a due to the rapid develop
ment of potential flow suction around the leading edge. Therefore,
separation at the knee is likely due to the high degree of leading-edge
curvature. At stage C, the upwash angle nominally exceeds the flap
angle, resulting in vorte~ formation and reattachment on the flap upper
surface. Tuft photographs (fig. 58) indicate that the transition of
the flow pattern from over- to under~deflection occurs between a = 60
and 100 for the 300 + VF-3. At a = 60 , tufts near the flap leading
edge are hidden, indicating separation on the lower side. At a = 100,
howeVer, these tufts lie on the face of the flap, indicating favorable
separation. With further increases in a, or outboard distance,
reattachment moves toward the leading edge, eventually coinciding with
the leading edge at stage D. The relatively high pressure at the
reattachment position corresponds to the low point on the -CpLE dip
in figure 57. This may be regarded as the "design" point since the full
chord of the flap is under vortex suction. The locus of the minimum
1"-, o:::j-.--~
+-' rO
c: 0
-r-+-' rO $-:::s en
-r-4-c: 0 u
(Y)
I I.l.. > -7-
0 0 (V)
4-0
U1 .c: 0. CO >-en 0 +-'0 0 0
.c: rl 0-
-0 +-' c: 4- CO :::s +-'0
\.0 OJ
-0 II -r-V) (j
ro I.!)
QJ >-:::s 01
-r-I.l..
148
-CpLE points along the span may, thus, be used as a guide to design an
optimum twisted flap for a given average deflection angle (see next " ,
paragraph). As the flow reattachment position and, subsequently, the
flap vortex, move onto the wing upper surface (stage E), the flap thrust
effectiveness diminishes.
The locus,of CpLE derived ,deSign points along the 300 + VF-3 is
plotted in figure 59. As previously noted, the ideal flow condition is
first met near t,he wing tip and moves, inboard with increasing a. This
curve also provides a general idea of ,the relative flap twist necessary
to produce the ideal flow condition simultaneously along the entire
flap. However~ this twisted flap, apart from being impractical, may not
be desirable since loss ~f effectiveness would also occur at once along
the entire span.
Figure 59 also presents the locus of the initial relative maximum
-C PLE points (see fig. 57) along the 30° + VF-3. This curve may be
thought of as a boundary between over- and under-deflection of the flap
relative to the oncoming flow. Again, due to the spanwise upwash dis
trib~tion, vortex formation initially occurs near the tip and spreads
inboard with increasing a. This is also evident in the tuft photo
graphs in figure 60. Although these were taken of the 600 + VF-3~ the
same basic trends would be expected with a 300 + deflection but at
lower angles of attack. The chordwise orientation of the tufts on the
flap at a = 140 is indicative of attached flow. By 160 a, however,
the flap vortex has formed and advanced to approximately the midpoint
of the flap (position 'b'). Due to expansion of the vortex core
0,
deg
20
18 .
16
14
12
10
8
6
o
-Cp LE
r
a
----(0)----'-----iDI---
149
Low point on -CPLE dip Initial -CPLE maximum
J,9 Attachment / - \'
on flap t--'
Attachment on wing
0.2 0.4 0.6
7]
Q. . P ___ Smooth -'Dr::;r:Y .. ~ on flow
.. / ~
0.8 1.0
Figure 59.- Local design point distribution (Cp derived) along 30° + VF-3 (see fig. 57). LE
Figure 60.- Side tuft photographs of 60° + VF-3 configuration at a = 14°, 160, 180 , and 23°.
150
151
toward the tip, reattachment between positions 'b' and 'c l occurs on the
flap face (indicated by the diverging tufts on the flap), with nearly
. an ideal flow condition aft of position 'cr. Further increases in a
are accompanied by continued inboard movement of the vortex, v/ith its
apex lying near the flap apex at a= 230. At this stage, reattachment
aft of position 'c' occurs on the wing upper surface, as evidenced by
the realignment of the tufts along the length of the flap.
Leading-edge thrust variations for the 300 ~ VF-3 configuration are.
presented in figure 61. No effect is evident on the leading-edge thrust
at low a since separation occurs on the lower side of the flap. At
mid a, however, the flap induces a thrust loss outboard of STA 1, cor
responding to the leading-edge suction loss noted in figure 57. However,
as the ideal flow condition is reached near the tip, the flap effective
ness begins to appear as a substantial thrust improvement. At angles of
attack higher than those considered here, further inboard movement of
the ideal flow condition would result in thrust improvements also at
the inboard stations.
The effects of th~ 300 ~ VF-3 on the overall performance character
istics of the configuration are given in figure 62. A large axial force
improvement over the basic wing is found at high a and is sustained to
the. highest angles tested. The severe low-a drag penalty is not of con
cern since at cruise the flap would be undeflected or retracted; as is
the loss of normal force at low a, which is attributed to the effective
negative camber induced by the downward flap deflection. The migration
of the primary vortex onto the wing upper surface leads to onset of
152
STA B' . . ~-----aSlC wlng 1 -0-- 30° ... VF 3 2 -
.10 r- ~ _3 .10 4
5 / - . . n - 0 33 .08 f--\,\ 6' .08· - •..
. 02 • . .02
o - 1 0 o 5 1 0 1 5 20 25 0 5 1 0 1 5 20 25
ex, deg IX, deg
.10
t .10
.08 n = 0.45 .o8L n = 0.57
cT .06·,' CT .06
'0'' :04 [ ,/ V-- - '00< ' .. o.;r /,/\ .02 .02
o ,'1 ·1 0 o .5 10 15 20 25 0 5 10 15 20 25
ex, deg IX, deg
.10r- .10
1 ,08/:::- n= 0.70. .08f- n = 0.82
CT
[OC5 .06t /"'. ~ c"o,' .06L .04 / Y .04
.02~r - .02
o 1 ,'1 1 1 0 o 5 1 0 1 5 20 25 0 5 1 0 15 20 25
ex, deg ex, deg
Fi~ure 61.- Effects of 300 + VF~3 on wing leading-edge thrust.
C A
CN
.08
o
-.04
"'-.08
-.12
o
1.0
.8
. 6
.4
.2
- - - - - - Basic wing.
~ 300 .VF-3 ----0---- 45°. VF-3
-----<>---- 60° t VF-3
I . I 5 10 15
lX, deg
20
I~ i)/~ / '
/¥ " .
25
C );;r4/ ~ 25
lX, deg
153
Figure 62.- Effects of vortex flap deflection angle on force and moment characteristics,
'.
154
10 I
/ Basic wing /
--- -I
8f- I ----0- 30° t VF-3
I J
---0--- 45° t VF-3 J _c_-(>-___ (I:P t VF-3
I 6
LID
4~ <;> "-
"-"-
"-"-
---I 2
0 0 _2 A _6 _8 LO
CL
-006r ~ 1\ 1\ r _005 f- I \ I \ f
I \ / J \
.004 r- \ \
\ \
I _003
Cz
_002
_001
5 10 15 20 25
0:, deg
Figure 62.- Concluded.
155
vortex lift at approximately a = 140 , accounting-for the convergence of
the CN curve toward the basic wing data at hi~her a. The resulting
effect on LID is a loss of performance at low a but substantial
improvement beyond CL ~ 0.3. The C1 data indicate that the vortex
flap also eliminates the severe mid-a rolling moment instability of the
basic wing.
Longitudinal stability effects of the 300 -t VF-3 are deP5cted in
figure 63. The moment reference cent~~ has once again been moved to the .,". ~.
same position used with the SLEE data (3.S2 em aft of the original ref
erencecenter indicated in fig. 2) to better show up the effects of the
device. The 300 -t VF-3 produces: a slight reduttion in longitudinal
stability at mid and ~igh ~ but eliminates the nonlinearities -of the
basic wing at a ~ So. The chordwise location of the center of preSSll;re
for thi~ same configuration (derived from Cm and eN data) is plotted
as a fUnction of a in figure 63. The forward-movement of xcp rela
tive to the basic wing beyond a ~ 60 (the angle at which vortex forma
tion is initiated on the flap face) is attributed to a forward movement
of the center of pressure along the flap. Whereas attached flow on the
basic wing positions the suction peak near the leading edge, the vortex
suction effect on the flap moves this suction peak forward onto the flap
surface (sketch M). Since pitch-up is the result of a forward movement
of the center of pressure, those flap configurations producing the most
aft center of pressure will be considered the most desirable as far as
longitudinal stability is concerned.
em
x ~
c r
0.04
o
-0.04
-0.08
-0.12 .
I o 5
x ~ = 0.474
cr
10
-- - - - - Basicwing
~--{Ol--- 30° t VF-3
° ---0--- 45 t VF-3 ----<>---- (fJ0 t VF-3
...... .......
....... ....... .......
L J ___ I 15 20 25
a, deg
a, deg
o 5 10 15 20 25 o I I . . I
0.1
0.2
0.3
0.4
0.5
Figure 63.- Effects of vortex flap deflection angle on longitudinal stabil ity character; s ti cs.
156
-c p
Basic wing
Sketch M
157
wi Vortex flap
Comparison of experimental results with theory (ref. 15) for the
300 + VF-3 configuration appears in figure 64. The downward shift in
the experimental lift data as compared VJith theory is attributed to the
effective neg.ative camber induced by the asymmetric trailing-edge region
(see fig. 2). Otherwise, the experimental lift and drag data show
relatively good agreement with theory throughout the angle-of-attack
range.
For a more basic assessment of the aerodynamic effectivene~s of the
various flap configurations tested, an analysis on a per unit flap area
basis will also be used. This will provide an indication of the effi
ciency with which the surface area of a particular flap configuration
is being utilized for drag reduction. Figure 65 presents balance and
leading-edge pressure-integrated thrust data for the 300 + VF-3 configu
ration. The axial force curve has been shifted downward to zero to
CL
Co
1.0
0.8
0.6
0.4
0~2
10
a, deg
0;3 r-
0.2
0.1
/
/ /
/
/
-'-0------ Expe ri me nt
- - - - - VLM-SA theory
15 20
/,.
'l' I'
25
o I -,~ .. 1.-1 II o 5 10 15 20 25
a, deg
Figure 64.- Comparison of 300 + VF-3 experimental data with VLM-SA theory. . .
158
-c Ttot
and
flCA
Cx,
1m2
a, deg
o 5 10 o ITP;;: I
...... 'D... t... t..v
...... ,Q
'Q Wing leading-f-" 'Du" ,dgo th,," -0.01 . ~~ -0.04
-0.06
-0.08
-0.10
-0.12
2.0 I
1.6
1.2
0.8
0.4
_ (-CTtot) - (flCA) Cx - A
VFtot
"-h.
...... -- -0 -Crtot Vortex flap
contribution
f Total thrust
Onset of vortex lift
,6CA = CA - CAo
Oe1------L-----~-------L------~----~ o 5 10 15 20 25
a, deg
Figure 65.- Thrust contribution from 30° + VF-3 flap surface.
159
160
eliminate the profile drag component .. CTtot represents the contribu
tion of the wing leading-edge suction to the axial force obtained by a
spanwise summation of CTloc
The difference in the two curves, there-
fore, represents directly the thrust contribution of the flap. This
difference is divided by the total flap surface area and plotted in fig-
ure 65. The curve indicates an improvement in flap efficiency with
increasing a. The sudden reduction in slope at a ~ 160 indicates
the onset of loss of flap thrust effectiveness due to vortex migration
onto the wing upper surface. This is consistent with the onset of vor-
tex lift noted in the CN curve (fig. 62) at approximately the same a.
Additional evidence is provided by the photographs in figure 66. At
a = 14°, the mini~tufts near the wing leading edge are generally chord-
wise, implying reattachment either on the flap or at the knee, with
attached flow prevailing on the wing upper surface. At a = 170,
however, these same tufts are angled toward the wing tip, suggesting
vortex action and reattachment on the upper surface.
The effects of flap deflection on overall performance are shown in
figures 62 and 63. Axial force shows an improvement at high a with an
increase in downward deflection of VF-3 from 30° to 45°, but a loss of
performance with further deflection to 60°. The effective increase in
negati ve ca·mber with increasing fl ap defl ecti on results in a reducti on
of normal force at low a. The same trends are evident at higher a
due to a delay in vortex migration onto the wing upper surface. Lift-
to-drag appears relatively insensitive to flap deflection beyond
CL ~ 0.4. This is attributed to the self-compensating for thrust effect
Figure 66.- Upper surface tuft photographs of 30° + VF-3 configuration at a = 14° and 17°. I-' m I-'
163
A major concern of the aircraft design engineer is that of produc
ing adequate lift to reduce the landing speed. In addition, decelera
tion of the aircraft from cruise to landing speed has been a problem
requiring the use of speed brakes and high approach angles of attack.
An up-deflected (t) vortex flap was tested in the present investigation
as a possible aid in alleviation of these problems (ref. 17). It was
believed that the suction effect of a vortex formed on the lee side of
the flap (see sketch F in BACKGROUND) would produce both lift and drag
increments ideal for the landing approach. The vortex flow induced on
the wing upper surface would provide additional lift.
Results of balance measurements on a 300 t VF-7 configuration (see
fig. 7) are presented in figure 67. Normal force indicates onset of
vortex lift almost simultaneously with departure from a = 00. The
strong spanwise inclination of the tufts near the wing leading edge at
a = 60, in figure 68, is indicative of this upper surface vortex flow.
Note the well-defined reattachment line with chordwise flow downstream.
The resulting improvement in CN implies that the required lift at
landing may be obtained at a lower flight speed. The rolling distance
after touch-down, which is proportional to the square of the landing
speed, can thereby be significantly reduced. At a 00, the 300 t VF-7
produces a relatively small increase in axial force since there is no
addition of frontal area (see fig. 7). However, at mid and high a,
large increases in drag are available for deceleration of the aircraft
during the landing approach. In addition, t1e linearity of the pitching
moment curve is significantly improved over that of the basic wing. The
.08
.04
CA
-.04
-.08
1.2
1.0
.8
.6
CN
.4
.2
o 5
"
/ /
/
10
- - - - Basic wing
---0- 30° t VF-7
15 20 25
lX, deg
I I
I I
I
I
/ /
I
I I
I /
/
/
I /
I I
I
I I
/
lX, deg
164
Figure 67.- Effects of upward vortex flap deflection on force and . moment characteristics.
em
C2
.04
o
-.04
-.08
-.12
o
.006 r--
,005
.004
.003
.001
'--,
Xcrn = 0.474 C
r
5 10
, '-
ct, deg
"" ,I
h f\ I' I \ I \ I \
\
\
\ I
- - - - - - Basic wing
--0---- 30° t VF-7
, , ,
15
, ,
I 20
, ,
25
0, ~ o ,~
cx. deg
Figure 67.- Concluded.
.. '.
165
" -J.
to c ..., (0
Q)
00
OJ c: rl--O
U Q n)
-') II
Ul O)C 0-,
-n Q.I (') (0
rl'" C -n rl'"
-0 ::y 0 rl'" 0 to ..., Q.I
"0 ::y
0 -n v..) C)
0
...,..
< -n !
'-.J
(') 0 :::::; -I)
to C -'l OJ rl'" -J. ~
0 Q)
::l Q)
~
167
rolling mOment characteristics are excellent up to approximately
ex. = 16°, at which point a minor instability appears. However, the
severe wing rock of the basic wing between 80 and 14° ex. is el-lminated.
These characteristics are significant since any unstable behavior at
near-stall landing conditions may be critical.
An inverse tapered vortex flap was also tested in the present
investigation to take advantage of certain flow phenomena characteristic
of the device. It was reasoned that increasing the flap chord toward
the wing tip would better accommodate the expanding vortex core and,
accordingly, produce a more efficient drag reducer. This inverse
tapered flap (VF-2) was designed with approximately the same surface
area as the constant chord VF-3 (see fig. 7), allowing for direct per
formance comparison. Balance data for the 300 + VF-2 and VF-3 configu
rations are presented in figure 69. The axial force data indicate a
slight improvement in drag-reduction effectiveness beyond ex. ~ 16° with
the inverse tapered VF-2. However, there is a slight loss of normal
force within this same ex. range, The longitudinal stability character
istics (hbt pr~sented) are not significantly different from those of the
constant chord flap.
Static pressure distributions around the wing leading edge
(fig. 70) confirm that drag performance improvements with the inverse
tapered VF-2 are indeed attributable to its ability to better maintain
the vortex on the flap and delay its migration onto the upper surface
in the outboard region. Stagnated pressures at n = 0.70 and 0.82 wIth
the constant chord VF-3 at ex. = 120 suggest that reattachment has moved
.08
.0
o CA
-.04
-.08
-.12
o 5
'T .8
.6
CN
.4
.2
10
Basic wing
~ 300~VF-3 ---D--- 30°. VF-2 -----0---- 30°. VF-l
---
15 20 25
a.. deg
168
Figure 69.- Effects of inverse tapered flap chord distribution and apex shape on vortex flap performance.
STA 1
2 -4
-3
3 4
5 , 6
-- -- -- - - Basic wing -0-- 30° {- VF-3 ----0---- 30° {- VF-2
-4-
-3
(Constant chord) (Inverse tapered)
n = 0.33
Cp1 = 0.20 C
p2
-1 -1
1~~ 2 3 4 5 0 1 2 3 4 5
X, em X, em
-4 r- -'[ -3!- -3
\ n = 0.45
-2 "'" n = 0.57
-2 l \ l \ , l
CP3
, , CP4
\
\
-1 -1
---------- -------
1 I 1 0 2 3 4 :5 0 2 3 4 5
X, em X, em
-4
-3 ,
1 = 0.70 \
~. -2 \
Cp " \'------ CP6 5, _
1 1 I I 11 1 1 I.~ o 2 3 4 5 02345
X, em X, em
169
Figure 70.- Effects of inverse tapered vortex flap chord distribution . on static pressure distributions around the leading edge
at a = 12°.
170
to the wing surface. The inverse tapered VF-2, however, has retained
attached flow at the leading edg~, with reattachment still on the flap.
Tuft photographs taken at a = 160 (fig. 71) indicate the same trends.
The chordwise orientation of the tufts near the wing leading edge with
the VF-2 indicates attached upper surface flow, whereas the spanwise
inclined tufts with the VF-3 configuration are the result of vortex
spillover. The delay in onset of vortex lift with the VF-2 accounts
for the reduced normal force noted in figQre 69.
A comparison of the CpLE derived locus of design points along the
.300 + VF-2 and VF-3 appears in figure 72. The relatively smaller flap
chord near the apex of the VF-2, as compared with the VF-3, reduces
the a at which the design point is met. The opposite effect is evi
dent outboard of n ~ 0.60. Delays in vortex spillover of up to 30 a
at these outboard stations with the VF-2 account for the high-a improve
'ments noted in the CA characteristics.
Comparison on a pe~ unit flap area basis, in figure 73, indicates
that the inverse tapered flap area distribution is somewhat more effi-
cient than the constant chord ~ariation up to o .
a ~ 16 . -More pronounced
improvements are indicated at higher a, as the VF-3 begins to lose out
board suction effectiveness. An implication of this result is that
savings in system ~'ieight are possible through appropriate geometric
design of the vortex flap.
An additional inverse tapered vortex flap (VF-l), characterized by
a chordwise-cut apex, as opposed to the sweptback apex of the VF-2, was
tested to determine the importance of flap apex shape. It was thought
Figure 71.- Upper surface tuft photographs of 30° + VF~2 and VF-3 configurations at a = 16°. I-' "'-J ......
20
i
18
.-
14
ades ,
deg 12
10
8
6
(} " 0.2
-.~ - - -:--
----IO~-
.~ ,.
\ \ \
300 t Vf-3
300 t VF-2
172
I nverse tapered
\. .
" / Constant ~hord " / - ./
TJ at which the chords of (VF-2 and VF~3 are iden.tical
0.4 0.6 0.8
'7]
l.0
Figure 72.- Effects of inverse tapered flap chord distribution on local design point distribution (C derived) along th~vbrtex flap~ _. PLE
; . .;.
r \...IX,
1m2
2.0
1.6
(-C l C = Ttot - (!',CAl x
AVFtot
1.2
0.8
0.4
/'
./' --./
-- ---- - - 30° t VF- 3
--0- 30° t VF-2
-----G---- 30° t VF-l
./
/ /
./
173
o ~ry/~/ ____________ ~ ______________ L-____________ ~L-____________ ~ ______________ ~
5 10 15 20 25
a, deg
Figur~ 73.- Effects aT lnverse tapered vortex flap chord distribution and apex shape on flap thrust parameter.
174
that a counter-rotati~g, vortex formed at this chordwise edge would
interfere with the formation of the primary fl ap vortex. Balance data
in figure 59, however, indicate th~t this particular apex modification
has insignificant effects on overall performance. The close agreement
of the flap thrust coefficient data in figure 73 provides added support.
As evidenced by the vortex flap data presented thus far, streamwise
migration of the flap vo~tex with increasing outboard distance is the
primary 1 imi tati on to effici ent fl ap performance at hi gh Ci.. Segmenta
tion of the flap was suggested ~s one method of alleviating this problem
through the formation of an independent vortex on the outboard segment,
thereby delaying thrust loss near the tip. The geometries of the seg-
mented flaps tested appear in figure 7. VF-4 was derived from VF-3 ,
simply by segmenting the flap at its midpoint and, thus, afforded the
opportunity to isolate the performance effects attributed solely to seg-
mentation. VF-5 and VF-6were included to investigate the possibility
of further performance improvements through variations in flap chord
distribution.
Result~ of balance mea~urements on configurations utilizing the
segmented vortex flaps at 300 downward deflections are presented in fig
ure 74. Comparison of VF-3and VF-4 axial force data indicates little
overall effect from segmentation. However, slight reductions in normal
force and longitudinal stabillty result at high Ci.. Comparison with
VF-5 and VF-6 data suggests that a substantial amount of flap area may
be eliminated with little sacrifice of CA performance up to Ci. ~ lSo.
Beyond Ci. ~ 180, theinvers~ tapered V~-5, with 42 percent less surface
CA
CN
.08..-
o
-.04
-.08
-.12
o
1.0
1 .8
.6
.4
.2
0 1,,/'
0/:;
175
Basic wing
--0-:- 30° t VF-3
----0--- 30° t VF-4 ------'0---- 30° t VF-5 - ... --6--... - 30° t VF-6.
5 10 15 20 25
0:, deg
0:, deg
Figure 74.- Effects of segmentation and flap geometry on vortex flap performance (8 = 30°).
o
-.04 em
-.08
-.12
o
--'·~"O.-"
- - - - - --'- Basic wing
--'----(0)-----.,. 30° t VF- 3
° ----0---. 30 t VF-4
----<)---- 30° t VF-5 " ---- -6------ 30v t VF-6
" ~.~ .. ::~:-\ .. ~'" '--- --........ ....... ....
" ", '- .. '-
'-'-
5
Xcm = 0.474 ~r
10 15
<1, deg
Figure 74.- Concluded.
..... '-
'- .......
20
176
25
177
area than the constant chord VF-4, loses effectiveness, while the para
bolic VF-5(13 percent reduction in area) continues to perform on par
with the VF-4. Reduction in flap area actually results in an improve
ment in longitudinal stability beyond a '" So, v.Jith little effect on
normal force. Balance data for 450 downward deflections of these seg
mented flaps (fig. 75) indicates the same trends. However, the 300
deflections will be used for description of the underlying flow
mechani sms.
The performance of the various segmented vortex flaps on a per unit
flap area basis is summarized in figure 76. As previously noted, the
flap thrust coefficient, Cx, provides an indication of CA effects
attributed solely to vortex suction on the flap surface. Segmentation
of the full length VF-3 into VF-4 results in substantial improvement in
flap area efficiency beyond. a:'" 160, as loss of flap suction has
apparently been delayed. The parabolic (VF-5) and inverse tapered
(VF-6) segmented flaps provide further improvements at lower a, but
with earlier loss of effectiveness with decreasing flap area. Each
particular segmented flap geometry seems to be the most efficient of
those considered within a specific a range, implying that the actual
geometry selected would depend on the design angle of attack.
Tuft photographs in figure 77 confirm that Cx improvements
resulting from segmentation are indeed attributed to retention of vortex
suction on the outboard flap segment. The mini-tuft pattern and static
pressure distribution at STA 6 with the full length VF-3 suggest flow
reattachment on the upper surface. However, the aft portion of the
.08
.0
o CA
-.04
-.08
-.12
.04
o
em -.04
-.08
-.12
u Basic wing
-0---:--- 45° t VF-3
° ----fr---- 45 t VF-4 -----<)---- 45° t VF-5 - ... ---L:r--.-.- 45° t VF-6
-~ - --
.--~
~~
'-------"----'---------'------'----,. o
o
5 10
5
Xcm = 0.474 Cr .
10
15
a, deg
15
a, deg
20 25
..... ...... ......
"
20 25
178
Fi gure 75. - Effects of segm.enta~i on and fl ap geometry on vortex fl ap performance (8 = 45 ).
ex, 1m2
179
2.0
1.6
1.2
0.8
0.4
:--_-0
V' / /<"'0-- / /
d~:/X/ /~~ ~/ '
" //--- . '",~
:;:>p' / '->
C = (-CTtot1 - (.6.CAl x
AVFtot
1d #
/:! !~ <1/7
,'10(//
A,)-l---Y
~
/;~ /~/
- - - - - - 30° t VF-3
--0-- 30° t VF-4
---0--.- 30° t VH
----<)---- 30° t VF-6
O~~S/~ ____ ~---------~------------~---------~------_-J 5 10 15 20
a, deg
Figure 76.- Effects of segmentation and vortex flap geometry on flap thrust parameter.
25
~:L -2~
I
-i~--~------~
ob--1 I ~----I----I----I- ) o 3 4
180
-4F -3 ------ Basicwing
-2k
Figure 77.- Side tuft photographs of 300 + VF-3, VF-4, VF-5, and VF-6 configurations at a = 14°.
-4, -:It -2
.~ i~
-4,---I I
o , 2 :I
-3~
~~~I, __ ~ __ - ,, __________ __
(-" ,,:::. 0
1 '-----
,1 I -;----1-- --;-023 ~
Figure 77.- Concluded.
181
_4
f -3 - - - - - - Basic wing
-2 _ ~----------------,
, -, -~ ,1 1--'-1----; ---I
o
182
segmented VF-4 appears to still be fully under the influence of vortex
suction (as shown by the mini-tufts on the flap slanting away from the
wing leading edge) with reattachment in the vicinity of the knee
(attached flow indicated on the wing leading edge). The flow through
the break has apparently forced the vortex on the forward flap segment
to peel off and, most likely, travel over the upper wing surface. The
parabolic VF-5 and inverse tapered VF-6 show the same baSic trends but
with accelerated forward movement of the outboard vorte~. Compar~son
with static pressure distributions at STA 3, where segmentation would
appear to have little influence, suggests that this is purely a geo-
-metric effect. As a result, the tuft photographs in figure 78 indicate
a reduction in upper surface vortex flow in the outboard region upon
segmentation of VF-3 (to VF-4), but with the opposite effect with cut-
back from VF-4 to VF-5 and, subsequently, VF-6. As noted in prev; ous
sections, an increase in upper surface vortex flow tends to increase the
overall lift of the wing but also the severity cif the lift-dependent
drag.
Figure 79 presents -the CPLE ~erived locus of- design points alnng
each of the 300 down-deflected segmented flaps. Although -the data
points are scattered, certain trends are consistent with those observed
in the tuft photographs. Sinc~ effects inboard of n 0.625 are-
pu~ely geometric, the VF-3 and VF-4 data points approximately coincide. - . Reduction in flap chord with VF~5 and VF-6 results in satisfaction of
the design condition at progressively lower a alorig'this forward
segment. The convergence of the data in the vicinity of the break
183
Fi gure 78. - Upper surface tuf'tphotographs of 30° -} VF -3. VP.;;'4'; VF-5. .. and VF-6 configurations at a = 18°.
Odes,
deg
- - - - - 30° t VF-3 --~o 30° t VF-4
---0--- 30° t VF-5
20;....,......- ----<>--.-. 30° t VF-6
18
16
14
12
10
8
Segment
'- / ....... _/
6 ., I_ I I . . . . o 0.2 0.4 0.6 0.8
'TJ
184
1.0
Figure 79.- Effects of segmentation and flap geometry on local design point distribution (C derived) along the vortex flap. . ~E
185
(n = 0.625) i~ attributed to the equality of the respective flap chords
(see fig. 7). The effect of segmentation (VF-3to VF-4) is evident
along the outboard segment in the form of a delay of several degrees in
vortex migration onto the upper surface. The effects of chord reduction
noted on the inboard flap segment are also evident here.
The utilization of vortex flaps for roll augmentation was briefly
investigated using the inverse tapered segmented VF-6. It was antici
pated that asymmetric flap deflections would alter the spanwise lift
distribution so as to produce adequate rolling moments at high lift,
when conventional control surfaces, such as ailerons, are degraded by
flow separation. Rolling moment data for a configuration utilizing a
300 i VF-6 on the left-hand and 450 + VF-6 on the right-hand wing panel
are presented in figure 80. The positive shift in the ~t curve cis com
pared with the basic wing is purely a planform effect. The imbalance of
planform area addition produces an additional lift increment on the
left, with a resulting right-wing-down rolling moment. Beyond ex '" 160 ,
the flap deflection effect becomes evident. The strong positive rolling
moment initiated at this ex is due to the earli~r migration Qf the
300 i flap vortex onto the left wing panel as compared with the vortex
emanating from the 450 i VF-6 on the right. The"additional vortex lift
on the left produces a maximum positive rolling moment of 0.0041 at
ex '" 180. The subsequent migration of the 450 i VF-6 vortex onto the
right wing panel, with its lift acting at an n value greater than that
of the left-hand vortex, produces a reversal in the rolling moment at
higher ex. The fact that strong positive and negative rolling moments
C'l
- - :- - ~ . ..,- Basicwing'
--'-'-_. """""""0 30° t VF-6~n left; 45° ~ VF-6 on right
.006 ~ '\ 1\
.0051--- I \
.003
.002
"'--- .... , f '" ,
.001
o
-.001
o 5
I . I . I
I
I I
I
I
10
\ \ \ \
15 20 25
lx, deg
186
Figure 80.- Effect of differential vortex flap deflection on rolling moment characteristics.
r,'.
..
187
are separated by only 40 or 50 a may produce undesirable handling
qualities at high lift. In addition, the magnitudes of the rolling
moments produced by the configuration tested here are inadequate to be
: of practical value on an actual aircraft. However, a greater differen
tial in flap deflections may provide the required rolling moments.
Based on the previous performance of fences and chordwise slots,
individually and in combination with the SlEE, their potential for
improving the performance of the vortex flaps was also investigated.
The 300 + VF-3 was tested with a single slot (w/SC-l) and then with a
single F-3 fence at its apex (n = 0.25), again in an attempt to aid in
the formation of an undisturbed flap vortex by blocking the leading-edge
cross-flow near the wing apex. An additional test run was made with a
single chordwise slot cut into the flap itself (SI) ~t n ~ 0.625, with
the slot in the wing leading edge remaining sealed (see fig. 7). As
with segmentation, the flow through the slot was expected to aerody
namically compartment the flap and, thus, induce the formation of an
independent primary vortex on its outboard side. Ideally, the entire
. flap surface would thereby remain under the influence of vortex suction
to higher u.
Axial and normal force data for the vortex flap-slot/fence con
figurations just described are presented in figure 81. Both graphs
indicate the overall ineffectiveness of each concept .. Effects on pitch
ing and rolling moment, not presented, are also negligible. Static
pressure data for the configurations with either a slot or fence at
n = D.25 show no evidence of a loc~l effect from the additional device.
.08 - - - - - - Basic wing
---0-- 3fP t VF-3 . . ° ---D---o~-. 30 t VF-3; wI SC-l@7]=0.25 ----0---- 30° t VF-3; wI F-l @ 7]= 0.25 -_._-£r_._- 300 t VF-3;w/S' @ 7]=0.625
o CA ----
-.04
-.08
-.12
o 5 10 15 20 25
ex. deg
'I I .6
CN
A
_2
a,
Figure 81.- Performance effects of the fence and chordwise slot in combination with the vortex flap.
188
"
--
189
The fact that these devices did not extend out ahead of the wing leading
edge, where the formation of the flap vortex occurs, may be responsible.
In that case, the chordwise slot cut into the flap at n = 0.625 would
be expected to perform effectively, assuming the flow through the slot
is sufficient for compartmentation. Indeed, the tuft photographs in
figure 82 show evidence of a second flap vortex outboard of the slot.
As depicted by the static pressure distributions in the same figure, the
wing leading-edge flow at STA 5 (n = 0.70) in the case of the VF-3 with
out a slot has already reached a stagnated condition, implying reattach
ment on the upper wing surface. The additional vortex formed as a
result of the slot at n = 0.625, however, retains attached flow at the.
wing leading edge, with the ideal flow conditi~n to be met at higher a,
as indicated in figure 83. Apparently, however, the sum of the suction
effects from the two smaller flap vortices (with a slot at n = 0.625)
is approximately equal to the suction produced by 'the single expanded
. vortex (without a slot), since no CA effect is evident. Additional
support is provided in figure 84, which shows that on a per unit flap
area basis, the 5uCtioneffectiveness of the flap surfaces· are
equivalent.
Finally, since the SLEE and vortex flap make use of basically the
same flow mechanism, a brief comparison of their suction effectiveness
is appropriate. Comparison of axial force data for the 450 + VF-3 and
o cm ext. SLEE (each found as the most effective in its respective
class), in figure 85, shows a significant advantage beyond a ~ 100
with the vortex flap. However, on a 6CA (relative to the basic wing)
190
-3f- ------ Basic wing
-2
-......\-.
o r '------- ___ h ___ h_
1 I I i I o 2 3 4
-4
f --
-3
-2
-- - -- ----- ---- -- ---
-'~ , ,
o~ 1 o 3
X, em
Figure 82.- Side tuft photographs of 30° + VF-3 configuration with and without a slot in the flap at n = 0.625 (a = 140 ).
.,
a des,
deg
'_~,7"'~~ ,
20
18
16
141 12~
10
8
6
191
- - - - - 30° t VF-3
----0-- 30°f VF-3 wI slot in
\A~ ~ I b-o
\ i
"-"- /
....... ./
Slot in flap
, flap '@ 7]= 0.625 (see fig. 7)
I. L------L_-.l.----L--~---'
a 0.2 0.4 0.6 0.8 1.0
7]
Figure 83.- Effect of a chordwise slot in the vortex flap on spanw·ise local design point distribution (C derived).
PLE
ex, 1m2
2.0
1.6
1.2
0.8
0.4
~ #'
~
~ ~
(-CT ) - (t.CAJ C = tot
x AVFtot
/,
~ ~
/,
V I On . o 5 10
'; J
I
Q., deg
Y
/ "7
- - - - - 30° ~ VF-3
--0-- 30° t VF-3 w/slot in
15
flap @ 7]= 0.625 (see fig. 7)
20
Figure 84.- Effect of a chordwise slot in the vortex flap on flap thrust parameter.
192
25
C>
CA
Sref (,0..C A) Afront
.08
-.04
-.08
-.12
.4
-.4
-.8
-1. 2
o 5
o 5
10
10
193
- --- -- Basicwing
---0- 0.00 em ext. SLEE; wI F-2 @ 7"J = 0.25 o
---~-- 45 t VF-3
15
ex, deg
15
20 25
fleA = CA - CA BW
Sref = Basic wing pianforrr area
Afront = Total frontal area (excluding housing)
20 25
ex, deg
Figure 85.- Comparison of sharp leading-edge extension and vort2x flap performance.
194
and frontal are~ basis., the two. devices perform identically. Therefore,
although the high-a drag performanc~ of th~ vortex flap is significantly
better than that of the SLEE, the additional weight of the vortex flap : I.
and the fact that the. SLEE may be rapidly deployed must be considered.
In summary, the vortex flap.prov~des substantial leading-edge
thrust improvements, resulting i\l 5izable axial force and lift-to-drag
increments at high a and elimination of the severe mid-a rolling
moment instability of the basic wing. In order to eliminate the charac-
teristic low-a drag penalty, however, the vortex flap would have to be
retracted or undeflected at cruise. Forty-five degrees appears to be
the optimum flap deflection angle based strictly on drag-reduction per
formance; however, lift-to-drag is relatively insensitive to flap
deflection in the range CL = 0.4 to 0.9. An upward deflected vortex
flap, by producing aerodynamic drag and vortex lift increments, is also
effective as a decel~ration and high-lift device for approach and land-
ing. A large downward deflection of the vortex flap upon touch-down can
provide continued. aerodynamic braking, with the added advantage of
increased downpressure on the wheels. High-a performance may be
improved through the use of an inverse tapered, as opposed to a constant,
flap chord distribution, which is better able to accommodate the expand-
ing vortex core and, thus, delay vortex spillover in the outboard
region. A two-segment constant chord flap improves the flap thrust
efficiency at high a through the formation of a separate vortex on
the outboard segment, which delays the outboard loss of leading-edge
suction. Elimination of the relatively ineffective portions of the
195
constant chord flap, leading to the parabolic (13 percent area ~eduction)
and inverse tapered (42 percent reduction) segmented flap configurations,
results in little or no loss of drag-reduction performance except at the
highest test a. Therefore, on a per unit area basis. these segmented
flap geometries are more efficient than the constant chord flap. Dif
ferential deployment of the vortex flap appears to also have roll aug
mentation potential at high lift. The use of a chordwise slot cut into
a full length flap at its midpoint generates an aerodynamic segmentation
effect by splitting the primary vortex into two smaller vortices but
with no significant influence on the total leading-edge thrust.
196
CONCLUSIONS
This report has presented the results of a wind-tunnel investiga-
tion undertaken to examine the potential for further drag reduction
through refined versions of leading-edge devices such as chordwise
slots, fences, pylon vortex generators, sharp leading-edge extensions,
and leading-edge vortex flaps. The results were based on low-speed
balance and static pressure measurements taken on a 60-deg cropped delta
wing model in the NASA-Langley Research Center 7- by lO-foot high-speed
tunnel. The intended use of the devices tested would be to reduce the
severe lift-dependent drag penalties associated with highly swept wings
at high lift. This section highlights what are believed to be the most
significant findings of the investigation.
The chordwise slot, fence, and vortex generator devices produce
substantial high-a drag and longitudinal stability improvements when
utilized on highly swept leading edges. The use of these devices in
mUltiple arrangements further enhances the leiding-edge suction along
the span by alleviating the adverse inboard effect of each particular
device. However, each device is characterized by a low-a drag penalty.
In the case of the chordwise slot, the main contribution to this drag
increase is from friction acting on the internal side surfaces, with the
pressure drag acting on the vertical face at the end of the slot of
secondary importance. Sealing the slot at cruise vJOuld be one method of
avoiding this drag penalty. Likewise, high-a performance is improved
with increasing slot depth, but with no effect from the contour of the
back face.
c· I
197
,
J In the case of the fence, the characteristic high-a loss.of
leading-edge thrust on the inboard side of the device is attributed to
its unsweeping of the upper surface isobars, rather than to any viscous
accumulation on the inboard side.
Vortex generator size reduction in the form of leading-edge length
and diagonal cutbacks reduce the low-a drag of the device but also
adversely affect high-a performance. Reduction in VG chord to a certain
extent results in performance improvements throughout the a range;
however, further reduction actually produces an increase in low-a drag.
The utilization of variable VG toe-in along the span to match the pre-
vailing sidewash ahead of the leading edge may be one method of simul-
taneously reducing the low-a drag and improving high-a performance. The
lower VG vortex apparently plays an important role through its own con-
tribution to the leading-edge thrust and, thus, should also be con-
sidered in the design of the VG shape. An extended chord VG provides
the added advantage of possible use also as a carrier of slender exter-
nal stores, providing substantial drag and longitudinal stability
improvements both with and without the external- store.
The vortex suction effect of the sharp leading-edge extension and
leading-edge vortex flap devices produces substantial lift-to-drag
increments at high a, in addition to improving the linearity of the
pitching and rolling moment curves. The drag-reduction effectiveness
of the SLEE is insensitive to extension at low and mid a; however,
there is a distinct advantage to having a shorter extention at high a.
as long as the device extends at least beyond the stagnation point.
198
The performance of the SLEE is enhanced by the utilization of either a
fence or chordwise slot at its apex to obstruct the spanwise flow
originating near the wing apex and allow for the formation of an undis-
turbed primary vortex. The fence also possesses the ability to compart-
ment the SLEE when used along its length, delaying outboard loss of
effectiveness.
In the case of the vortex flap, the lift-to-drag is insensitive to
deflection angle in the range. CL ~ 0.4 to 0.9. High-a drag-reduction
effectiveness is improved through the utilization of an inverse tapered,
as opposed to a constant, flap chord distribution to better accommodate
.the expanding vortex core. In addition, segmentation of the vortex flap
improves the high-a performance of the device on a per unit flap area
basis through the formation of a separate vortex on the outboard segment.
A chordwise slot in the vortex flap has the same aerodynamic effect ! .
but has no influence on the total leading-edge thrust. Parabolic and i . inverse tapered segmented. flap chord distributions provide f~rther ! -
improvements in flap area efficiency. The performance of these flaps
~uggest that -similar variations in SLEE extension may ~nprove the
device's ability to hold the primary vortex ahead of the leading edge.
Twisting the SLEE and vortex flap in future tests may also aid in that
respect.
By means of an upward deflection, the vortex flap is also effective
as a deceleration and high-lift device for approach and landing. Large
downward deflection upon touch-down provides continued aerodynamic
braking, along with the added advant~ge of downpressure on the wheels.
199
In addition, differential deployment of the vortex flap appears to pro
vide roll augmentation at high a.
Finally, it should be noted that the devices tested here are still
in the early stages of development and, therefore, future tests must be
performed on scale ~odels of actual aircraft to determine the effects of
parameters such as camber, sweep, twist, and fuselage interference prior
to making any final decisions on their effectiveness. In addition, it
is recommended that future research include testing of the lateral/
directional characteristics of the devices to complete a data base to
be used in the final design.
200
, 'R~FERENCES
. "~' '. ,
1. Rao, D. rv'!.: Leading-Edge IlVortex Flapsll for Enhanced Subsonic
AerodYnami.cs of Slender Wings. ICAS-80-13.5, 12th Congress of ( ..... ;<PO ....
the International Council of the Aeronautical Sciences (Munich),
1980. ':,'
2. Rao, D. M. and Johnson,T .. ,D."Jr.: Subsonic Wind-Tunnel Investi-
gation of Leading-Edge Devi~es.0n Delta Wings (Data Report).
NASA CR 159120, ,1979.
3. Rao. D. M.: Leading Edge Vortex-Flap Experiments on a 74-Deg.
Delta Wing. NASA CR 159161,1979.
4. Bobbitt, P. J.: Modern Fluid Dynamics of Subsonic and Transonic
Flight. AIAA Paper 80-0861, 1980.
5. Johnson, T. D., Jr. and Rao. D. M.: Experimental Study of Delta
Wing Leading-Edge Devices for Drag Reduction at High Lift.
NASA CR 165846, 1982.
6. KUchemann, D., FRS: Types of Flow on Swept Wings. J. Royal Aero.
Soc., Vol. 57, pp. 683-699, November 1953.
7. Harris, C. D. and Bartlett, D. W.: Wind Tunnel Investigation of
Effects of Underwing Leading-Edge Vortex Generators on a Super-
critical-Wing Research Airplane Configuration. NASA TM X-2471,
1972.
y
8. Rao, D. M. and Johnson, T. D., Jr.: An Investigation of Delta
Wing Leading-Edge Devices. J. Aircraft, Vol. 18, No.3,
pp. 161-167,1981.
201
9. Fox, C. H., Jr. and Huffman, J. K.: Calibration and Test Capabili
ties of the Langley 7- by 10-Foot High Speed Tunnel. NASA
TM X-74027, 1977.
10. Fox, C. H., Jr.: Real Time Data Reduction Capabilities at the
Langley 7- ~y 10-Foot High Speed Tunnel. NASA TM 78801, 1980.
11. Rao, D. M. and Tingas. S. A.: Subsonic Balance and Pressure
Investigation of a 60-Deg. Delta Wing with Leading-Edge Devices
(Data Report). NASA CR 165806, 1981.
12. Gillis, C. -L., Polhamus, E. C., and Gray, J. L., Jr.: Charts for
Determining Jet-Boundary Corrections for Complete Models in 7- by
10-Foot Closed Rectangular Wind Tunnels. NACA NR L-123, 1945.
13. Herriot, J. G.: Blockage Corrections for Three-Dimensional-Flow
Closed-Throat Wind Tunnels, with Consideration of the Effect of
Compressibility. NACA Rep. 995, 1950.
14. Ridder, S.: On the Induced Drag of Thin Plane Delta Wings. An
Experimental Study of the Spanwise Distribution of the Leading
Edge Forces at LOlli Speeds. KTH AERO TN 57, Royal Institute of.
Technology (Stockholm), 1971.
202
15. Margason, R. J. and Lamar, J. E.: ~ Vortex-Lattice Fortran Program
for Estimating Subsonic Aerodynamic Characteristics of Complex
P1anforms. NASA TN 0-6142, 1971.
16. Stadler, E. L. and Trammler, S. J.: F-4C Performance Data and
Substantiation. McDannel Aircraft Corporation Report G 853,
Vol. I, 1968.
17. Marchman, J. F.: The Aerodynamics of Inverted Leading Edge Flaps
on Delta Wings. AIAA Paper 81-0356, 1981.
-'
..,.
1, Reoo;;:-No. -- 2. Governmen~Acces~ion No. ! 3. Recipient's Cataloo No.
NASA CR- 165923 ! . -4. Title and Subtitle f 5: RePort Date
Subsonic Balance and Pressure Investigation of a 60-Deg ~ . Ma~ 1982 Delta Wing with Leading-Edge Devices I o. Performing OrganIzatIOn Code , I 505-31-43-03
7. Author(s) I 8. Performing Organization Report No.
Stephen A.Tingas* and Dhanvada M. Rao** I------------~------------I--_____________________ -,-__ '--___________________ . ____ ,~ 10, Work Unit No.
9, Performing Organization Name and Address !
I North Carol ina State University Ii------------~--' I Mechanical and Aerospace Engineering Department ! 11. ~C~rI~406 Grant No,
I Raleigh, NC 27650.
I 13. Type of Report and Period Covered
12. Sponsoring Agency Name and Address Contractor Report I National Aeronautics and Space Administration
"
Washington, DC 20546 14. Sponsoring A,gency Code
I 505-31-43-03 15. Supplementary Notes
*North Carolina State University, Raleigh, NC 27607 **Vigyan Research Associates, Inc., 28 Research Drive, Hampton, VA 23666
116. Abstract
'I' Low supersonic wave drag makes the thin highly swept delta wing the logical choice for use on aircraft designed for supersonic cruise. However, the high-lift maneuver capability of the aircraft is limited by severe induced-drag penalties attributed
I I I
·1
I I i i I
I I
I I
i I
I \17.
I
to loss of potential flow leading-edge suction. Th'is drag increase may be alleviated through leading~edge flow control to recover lost aerodynamic thrust through eith~r retention of attached leading~edge flow to higher angles of attack or exploitati'on of the increased suction potential of separation-induced vortex flow. A low-speed wind-tunnel investigation was undertaken to examine the high-lift devices such as fences, chordwi'se slots, pylon vortex generators, leading-edge vortex, flaps, and sharp leading-edge extensions. The devices were tested individually and i~ combinations in an attempt to improve high-alpha drag performance with a minimum of ! low-alpha drag penalty. This report presents an analysis of the force, moment, and !
static pressure data obtained i'n angles of attack up to 230, at Mach and Reynolds
numbers of 0.16 and 3~85 x 106 per meter, respectively, The results indicate that all the devices produced drag and longitudinal/lateral stability improvements at high lift with, in most cases, minor drag penalties at low angles of attack.
Key Words (Suggested by Author(s)) 18, Distribution Statement
Leading-edge devices Delta Wing , Unclassified ~ Unlimited
,. I Subsonic wind-tunnel test Ba)ance and pressure data High-lift drag reduction
19. Security Oassif. (of this report)
Unclassified 20. Security Classif. (of this page)
Unclassified
Subject Category - 02
21. No. of Pa;;-es
216 1 22, Price
! A.10
N-305 For sale by the National Technical information Service, Sprinefieid, Virginia 22161
End of Document