Original Article

Proc IMechE Part G:

J Aerospace Engineering
Effects of pitch, yaw, and roll on delta 0(0) 1–14
! IMechE 2015

wing skin friction topology Reprints and permissions:

sagepub.co.uk/journalsPermissions.nav
DOI: 10.1177/0954410015594823
uk.sagepub.com/jaero

Sudesh Woodiga1,2, Tianshu Liu1, RS Vewen Ramasamy1

and Sai Kumar Kode1

Abstract
High-resolution skin friction fields in separated flows over a 65 delta wing and a 76/40 double-delta wing with different
junction fillets are obtained by using quantitative global skin friction diagnostics based on luminescent oil visualizations.
The topological features such as separation and attachment lines on these delta wings are identified, and the effects of the
pitch, yaw, and roll angles on the skin friction topology are studied systematically. The topological consistency of the
measured skin friction fields on the delta wings is examined by using the Poincare–Bendixson index formula.

Keywords
Delta wing, skin friction, topology, visualization, luminescent oil, pitch, yaw, roll

Date received: 20 February 2015; accepted: 11 June 2015

Introduction sections, a wing vortex is formed. At low AoA, the

A delta wing is a type of low aspect ratio wing with a
triangular planform typically defined by its sweep
angle that is the angle from the horizontal axis to
the wing leading edge (LE). Delta wings have been
widely utilized in high performance combat aircraft,
supersonic civil aircraft, hypersonic vehicles, and the
space shuttle orbiter.1–3 One of the advantages of
using a delta wing for flight vehicles operating at
these speeds is the reduction in transonic or super-
sonic wave drag due to high sweep angles. Another
advantage of a delta wing is its capability for high
angle of attack (AoA) operations. At high AoA, a
delta wing is able to operate effectively due to the
nonlinear lifting mechanism of vortex lift. A delta
wing generates a leading edge vortex (LEV) which
remains attached to the upper surface of the wing
even at high AoA, causing strong suction on the
wing upper surface subsequently augmenting lift,
allowing the delta wing to operate in flight envelopes
not attainable by conventional wings.
A double-delta wing, which is also known as a
strake/wing configuration, has the general geometric
characteristics of a delta wing with a discontinuity or
a kink in its LE. The strake is set at a higher sweep
angle than the wing. The highly swept strake acts as
an additional lifting surface forming a stable vortex
structure above it. This strake vortex carries on
through the wing section which keeps the flow
attached to the wing to a higher AoA than the case
of a wing without a strake or a conventional delta
wing. At the junction between the strake and wing
wing and strake vortices form distinct trajectories.4
Juncture fillets are additions to the double-delta
wing at the strake–wing junction with the objective
of controlling the LEVs on the double-delta wing.5,6
Two types of fillets that have been studied are the
diamond and parabolic fillets. Extensive force meas-
urements and flow visualizations have shown the
effectiveness of the diamond and parabolic fillets in
delaying vortex breakdown and augmenting lift.5–7
These observations were confirmed in numerical
simulations.8,9
The vortical structures associated with delta wings
contribute to interesting aerodynamic phenomena
that augment the performance of these
wings.1–3,10–12 Studies of flow structures on typical
delta wings (such as 65 delta wing and 76/40
double-delta wing) have been conducted experimen-
tally13–16 and numerically.17–20 In particular, the
effects of the yaw angle (YA) and roll angle (RA)
on delta and double-delta wings were investi-
gated.21–26 The effects of the YA on the vortical struc-
tures on nonslender delta and diamond wings were
also studied.27–30
1
Department of Mechanical and Aerospace Engineering, Western
Michigan University, Kalamazoo, USA
2
VEMC FD09, Ford Motor Company, Dearborn, USA
Corresponding author:
Tianshu Liu, Department of Mechanical and Aerospace Engineering,
Western Michigan University, Kalamazoo, MI 49008, USA.
Email: tianshu.liu@wmich.edu

Downloaded from pig.sagepub.com at TOHOKU UNIV on August 21, 2015

2 Proc IMechE Part G: J Aerospace Engineering 0(0)
Surface oil flow visualizations have been used to
obtain information on the topology of skin friction
fields on delta wings,7,10,13–16 in addition to numerical
simulations providing the understanding of the flow
structures over delta wings.8,9,17–20 The global lumi-
nescent oil-film (GLOF) technique allows mapping of
topological features such as singular points, reattach-
ment and separation lines, and other subtle features
associated with vortex bursting and vortex inter-
actions.31,32 The previous study focused on the
changes in the topological structures of skin friction
fields for a 65 delta wing and a 76/40 double-delta
wing with the juncture fillets as AoA was increased.
The effects of the juncture fillets on vortex breakdown
and performance characteristics of a 76/40 double-
delta wing were also evaluated.
The objective of this study is to provide quantita-
tive high-resolution skin friction fields on a 65 delta
wing and a 76/40 double-delta wing with junction
fillets at different pitch (AoA), YA, and RA based
on GLOF measurements. The evolution of the skin
friction topology as these angles are changed is
revealed, particularly focusing on the effects of the
YA and RA. This paper is organized as follows. The
GLOF method and the experimental setup are
described in the following section. The results on
the 65 delta wing at different pitch (AoA), YA,
and RA are discussed in the subsequent section,
particularly on the topological features such as
reattachment lines, separation lines, singular points,
and vortex burst (VB) signature. Similarly, the
results on the 76/40 double-delta wing with junction
fillets are discussed in the following section, particu-
larly on the effects of the diamond and parabolic
juncture fillets on the skin friction topology. In the
following section, the topological consistency of the
extracted skin friction fields is examined by applying
the Poincare–Bendixson index formula to typical
cases.
Experimental technique and setup
High-resolution skin friction fields on the delta wings
were obtained by using the GLOF skin friction
meter.31–33 This technique is based on surface lumi-
nescent oil visualization that represents the oil-film
thickness distribution. The evolution equation relat-
ing the luminescent intensity to skin friction is derived
in the image plane, which has the same mathematical
form as the physics-based optical flow equation.34
From a pair of luminescent intensity images acquired
at two instants, the evolution equation is solved for a
‘‘snapshot’’ skin friction field as an optical flow prob-
lem. After a sequence of snapshot solutions is
obtained, a relative steady-state skin friction field is
reconstructed by a direct superposition (or time aver-
aging) of these snapshot solutions.31 Since the skin
friction topology is the focus of this work, normalized
skin friction fields are presented without a priori or in
Downloaded from pig.sagepub.com at TOHOKU UNIV on August 21, 2015
situ calibration. Furthermore, the skin friction fields
obtained by the GLOF method are time-averaged
fields since the evolution of the thin-oil thickness is
slow and the superposition of snapshot solutions is
essentially equivalent to time averaging.
The experimental setup is illustrated in Figure 1 for
luminescent oil visualization on a delta wing in a wind
tunnel. Luminescent oil was made by carefully blend-
ing an oil-based ultraviolet (UV) dye (Petroleum
Tracer Concentrate DFSB-K175 from Risk Reactor,
www.riskreactor.com) with polydimethylsiloxane
(silicone oil) with a viscosity of 100 cs. The ratio of
the UV dye to silicone oil was 0.005:1. Model meas-
urement surfaces were coated with a white Mylar
sheet to enhance the luminescent oil emission.
Luminescent oil was brushed carefully or sprayed
onto a model surface to ensure uniform oil film appli-
cation. The resulting luminescent oil emitted the radi-
ation at a longer wavelength (about 550–620 nm) due
to the Stokes shift when excited by UV illumination.
In low-speed flows with a constant temperature, the
temperature effects on the luminescent dye and oil
viscosity are not considered.
The wings used in this study were CNC machined
out of ¼ in. steel plate with a flat upper surface and
45 LE bevel. The upper surface of the wings was
covered with a 0.075-mm-thick white MonoKote
adhesive trim sheet to enhance luminescent oil emis-
sion. Figure 2(a) to (d) are images of the 65 delta
wing, baseline 76/40 double-delta wing, and 76/40
double-delta wing with diamond and parabolic junc-
ture fillets, respectively. The 65 delta wing had a root
chord and span of 203 and 185 mm, respectively while
the 76/40 double-delta wing with juncture fillets had
a root chord and span of 203 and 207 mm,
Figure 1. Illustration of luminescent oil-film skin friction
measurement setup.
Woodiga et al. 3

Figure 2. Images of (a) 65 delta wing, (b) baseline 76/40 double-delta wing, (c) 76/40 double-delta wing with a diamond fillet, and
(d) 76/40 double-delta wing with a parabolic fillet.

Figure 3. Images of (a) yaw and roll adjustment mechanism and (b) 76/40 baseline double-delta wing mounted on the adjustment
mechanism.

respectively. Figure 3(a) and (b) are images of the
wind tunnel attachment mechanism for yaw and roll
adjustments and the baseline 76/40 double-delta wing
mounted on the adjustment mechanism, respectively.
Measurements were performed at AoAs of 10 and
20 with a combination of YA and RA of 10 , 15 ,
and 20 . Table 1 is a test matrix detailing the
combination of AoA, YA, and RA performed for
each wing.
Experiments were conducted at a freestream vel-
ocity of 20 m/s in a dark environment in a low-speed
wind tunnel with a test section of 0.406 m by 0.406 m
at the Applied Aerodynamics Laboratory in Western
Michigan University. The freestream turbulence

Downloaded from pig.sagepub.com at TOHOKU UNIV on August 21, 2015

4 Proc IMechE Part G: J Aerospace Engineering 0(0)
Table 1. Test matrix for wind tunnel testing of the delta
wings.
Parameter/Case 1 2 3 4 5 6
AoA ( ) 10 20 10 10 20 20
YA ( ) 0 0 10 15 0 0
RA ( ) 0 0 0 0 10 20
intensity was about 0.2%, and the Reynolds number
based on the base chord (Rec) was 275,000 for all the
delta wings. The top and side glass windows of the test
section allowed optical access. Two UV light-emitting
diode (LED) arrays were positioned to ensure an
approximately uniform illumination field on the
wing surface. Luminescent oil visualization images
were captured at 57 frames/s using a Basler Aviator
charge-coupled device camera with a 550 nm long-
pass filter allowing only detection of the luminescent
emission centered at about 590 nm. The images were
processed by using the optical flow algorithm to
extract skin friction fields. In this study, 70 snapshot
solutions were used for superposition from successive
image pairs (1 s interval between them) to reconstruct
skin friction fields. There is no rigorous theory to
determine the number of snapshot solutions for super-
position, which depends on the evolution of the oil
film in a specific test condition.31 As a rule of
thumb, snapshot solutions are selected in a time
period in which the oil-film development in the
regions of interest is significant. When the number
of snapshot solutions is sufficiently large, the super-
posed or time-averaged field converges to a steady-
state one.
65 Delta wing
For a delta wing in yaw, the windward side of the
wing experiences an effectively reduced sweep angle,
while the leeward side of the wing experiences an
effectively increased sweep angle. This causes the
windward and leeward LEV trajectories to move
inboard toward the leeward side of the wing with
increasing YA. The windward side VB location has
been observed to move toward the apex of the wing
with increasing YA, while the leeward side vortex
tends to move off the surface of the wing with VB
location moving toward the wing trailing edge (TE).
For the delta wing in roll, the effects experienced by
the upward leading edge (ULE) and downward lead-
ing edge (DLE) are combined effective AoA and
sweep angle change. In numerical simulations,20 the
results have shown an outboard and off surface move-
ment of the LEV on the ULE side and inboard,
toward surface movement of the LEV on the
DLE side.
Figure 4(a) to (d) shows a typical raw captured
image, skin friction lines, skin friction magnitude dis-
tribution, and skin friction vectors, respectively, for a
Downloaded from pig.sagepub.com at TOHOKU UNIV on August 21, 2015
65 delta wing at AoA of 10 . Figures 5(a) and (b),
6(a) and (b), and 7(a) and (b) show the skin friction
lines on the wing for the six test cases of AoA ¼ 10 ,
AoA ¼ 20 , (AoA, YA) ¼ (10 , 10 ), (AoA,
YA) ¼ (10 , 15 ), (AoA, RA) ¼ (20 , 10 ), and (AoA,
RA) ¼ (20 , 20 ), respectively. The x (chordwise) and
y (spanwise) coordinates are normalized by the root
chord (c).
For the case of increasing AoA as shown by
Figure 5(a) and (b), the reattachment lines associated
with the primary LEV and the secondary separation
line are clearly identified. These features are symmet-
ric about the wing root. In general, the primary
reattachment line moves inboard with increasing
AoA. For example, as AoA is increased from 10 to
20 , the primary reattachment line on the right side of
the wing moves inboard from (y/c) ¼ 0.06 to
(y/c) ¼ 0.04 at the chordwise location (x/c) ¼ 0.2.
Similarly, the secondary separation line moves
inboard from (y/c) ¼ 0.36 to (y/c) ¼ 0.34 at
(x/c) ¼ 1.0 when AoA was increased from 10 to 20 .
Here, a reattachment line is defined as a skin friction
line that neighboring skin friction lines converge to.
Similarly, a separation line is defined as a skin friction
line that neighboring skin friction lines diverge from.
The skin friction lines on the wing at
(AoA, YA) ¼ (10 , 10 ) and (10 , 15 ) are shown in
Figure 6(a) and (b). The reattachment line of the pri-
mary LEV on the windward side of the wing moves
inboard with increasing YA. For YA of 10 and 15 ,
the reattachment line is located at (y/c) ¼ 0.1 and
(y/c) ¼ 0.08 at (x/c) ¼ 0.5, respectively. There is no
obvious indication of the reattachment line associated
with the LEV on the leeward side of the wing for
both cases. The skin friction footprint of VB is
observed for the windward side LEV, which is identi-
fiable as the termination of a well-defined reattach-
ment line that is typically visualized as a sudden
expansion of skin friction lines from the end of the
reattachment line. The ending position of the primary
reattachment line corresponds to the spatial location
of VB although they are not exactly same. Here, we
loosely use the ending position of the primary
reattachment line as the VB location. The VB location
is seen to move upstream toward the apex of the wing
with increasing YA. VB is observed at (x/c) ¼ 0.72
moving upstream to (x/c) ¼ 0.58 for YA of 10 and
15 , respectively. This trend of the upstream move-
ment of the VB location with increasing YA is con-
sistent with previous observations.21–23 Figure 7(a)
and (b) shows the skin friction lines on the wing at
(AoA, RA) ¼ (20 , 10 ) and (20 , 20 ). The AoA of
20 is selected in this typical case since the effect
of RA is more pronounced at higher AoAs for both
single-delta and double-delta wings. Between the
two RAs, the skin friction fields are qualitatively
similar. For both cases, the secondary separation
lines are observed to move inboard toward the
DLE side.
Woodiga et al.
Figure 4. Typical (a) raw captured image, (b) skin friction lines, (c) skin friction magnitude distribution, and (d) skin friction vectors
on a 65 delta wing at AoA ¼ 10 .
76/40 Double-delta wings
Baseline wing
The observations of the LEV behavior for a double-
delta wing are similar for a delta wing in static yaw
and roll cases although the effects are more complex
for the double-delta wing because of the interaction
between the wing and strake vortices.25,26 Figures 8(a)
and (b), 9(a) and (b), and 10(a) and (b) show the skin
friction lines on the baseline 76/40 double-delta
wing for the six test cases of AoA ¼ 10 , AoA ¼ 20 ,
(AoA, YA) ¼ (10 , 10 ), (AoA, YA) ¼ (10 , 15 ),
(AoA, RA) ¼ (20 , 10 ), and (AoA, RA) ¼ (20 , 20 ),
respectively. The x (chordwise) and y (spanwise)
coordinates are normalized by the root chord (c).
As shown in Figure 8(a) for the case of the wing at
AoA of 10 , the primary reattachment line is clearly
Downloaded from pig.sagepub.com at TOHOKU UNIV on August 21, 2015
5
visible along with the secondary separation and
reattachment lines. The reattachment lines appear to
curve outboard at (x/c) ¼ 0.7 indicating the inter-
action of the strake and wing vortices. There is no
VB observed at AoA of 10 . As AoA is increased to
20 asymmetric VB is observed. As seen in
Figure 8(b), VB occurs at (x/c) ¼ 0.68 for the left
side of the wing and (x/c) ¼ 0.78 for the right side of
the wing. A spiral feature is also observed on the right
side LE close to the strake–wing junction. The asym-
metry in Figure 8(b) could be caused by small mis-
alignment of the wing relative to the freestream in the
experimental setup and the intrinsic instability of the
vortex system.
For the cases of the wing at (AoA, YA) ¼ (10 , 10 )
and (10 , 15 ), as illustrated in Figure 9(a) and (b), the
reattachment line of the primary windward side LEV
6 Proc IMechE Part G: J Aerospace Engineering 0(0)

Figure 5. Skin friction lines on a 65 delta wing at AoA ¼ (a) 10 and (b) 20 .

Figure 6. Skin friction lines on a 65 delta wing at (AoA, YA) ¼ (a) (10 , 10 ) and (b) (10 , 15 ).

is clearly seen. The reattachment line is observed to
move inboard with increasing YA from (y/c) ¼ 0.08 to
(y/c) ¼ 0.03 at (x/c) ¼ 0.6 for YA of 10 and 15
respectively. The VB location of the windward side
LEV is located at (x/c) ¼ 0.72 and (x/c) ¼ 0.68 for
YA of 10 and 15 , respectively. This indicates an
upstream movement of VB with increasing YA con-
sistent with the previous observations.26 On the lee-
ward side, there is no well-defined reattachment line
observed for both YA ¼ 10 and 15 . Although vel-
ocity data are not available in this case, it is
speculated that LEV on the leeward side is much
weaker and less organized than that on the windward
side.
Figure 10(a) and (b) show the skin friction lines on
the wing at (AoA, RA) ¼ (20 , 10 ) and (20 , 20 ),
respectively. The primary reattachment lines are
observed to move inboard toward the ULE side of
the wing from (y/c) ¼ –0.03 at (x/c) ¼ 0.2 for the
wing at RA of 0 to (y/c) ¼ 0 at (x/c) ¼ 0.2 for the
wing at both RAs. Asymmetric VB is also observed
for both RAs. For RA of 10 , VB occurs at

Downloaded from pig.sagepub.com at TOHOKU UNIV on August 21, 2015

Woodiga et al. 7

Figure 7. Skin friction lines on a 65 delta wing at (AoA, RA) ¼ (a) (20 , 10 ) and (b) (20 , 20 ).

Figure 8. Skin friction lines on a baseline 76/40 double-delta wing at AoA ¼ (a) 10 and (b) 20 .

(x/c) ¼ 0.63 on the DLE side of the wing, and
(x/c) ¼ 0.83 for the ULE side of the wing. As RA is
increased to 20 , VB moves upstream toward the apex
of the wing at (x/c) ¼ 0.57 on the DLE side of the wing
and downstream toward the TE of the wing at
(x/c) ¼ 0.93 for the ULE side of the wing.
Wing with diamond fillet
Vortex flow control for a double-delta wing was done
previously by using the diamond juncture fillet.26 To
look at this phenomena, Figures 11(a) and (b), 12(a)
and (b), and 13(a) and (b) show the skin friction lines
on the 76/40 double-delta wing with a diamond
juncture fillet for the six test cases of AoA ¼ 10 ,
AoA ¼ 20 , (AoA, YA) ¼ (10 , 10 ), (AoA, YA) ¼
(10 , 15 ), (AoA, RA) ¼ (20 , 10 ), and
 
(AoA, RA) ¼ (20 , 20 ), respectively.
As shown in Figure 11(a), at AoA of 10 , the pri-
mary reattachment line is clearly seen with the second-
ary separation and reattachment lines. A set of
separation and reattachment lines is also observed
near the diamond fillet junction related to the vortex
structures created by the additional discontinuities
associated with the diamond-shaped junction. The
reattachment lines curve outboard at (x/c) ¼ 0.65

Downloaded from pig.sagepub.com at TOHOKU UNIV on August 21, 2015

8 Proc IMechE Part G: J Aerospace Engineering 0(0)

Figure 9. Skin friction lines on a baseline 76/40 double-delta wing at (AoA, YA) ¼ (a) (10 , 10 ) and (b) (10 , 15 ).

Figure 10. Skin friction lines on a baseline 76/40 double-delta wing at (AoA, RA) ¼ (a) (20 , 10 ) and (b) (20 , 20 ).

indicating interaction of the strake and wing vortices.
The curving of the line occurs earlier compared to the
baseline wing possibly related to additional vortices
created by the diamond fillet. As AoA is increased
to 20 symmetric VB is observed at (x/c) ¼ 0.73 in
Figure 11(b).
In the case of the wing at (AoA, YA) ¼ (10 , 10 )
and (10 , 15 ), the reattachment line of the primary
LEV of the windward side of the wing is observed to
move inboard toward the leeward side of the wing
with increasing YA, as seen in Figure 12(a) and (b).
The reattachment line is located at (y/c) ¼ 0.08 and
(y/c) ¼ 0.05 at (x/c) ¼ 0.6 for YA of 10 and 15 ,
respectively. The primary LEV is observed to burst
at (x/c) ¼ 0.7 and (x/c) ¼ 0.68 for YA of 10 and 15 ,
respectively. The VB location also moves upstream
toward the apex of the wing with increasing YA.
The delay of VB is not observed when compared to
the baseline wing for the wing with YA.
Figure 14(a) and (b) shows the skin friction lines on
the wing at (AoA, RA) ¼ (20 , 10 ) and (20 , 20 ).
The primary reattachment lines are observed to
move inboard toward the ULE side of the wing
from (y/c) ¼ –0.03 at (x/c) ¼ 0.2 for the wing at RA
of 0 to (y/c) ¼ 0 at (x/c) ¼ 0.2 for the wing at RA of
both 10 and 20 . Asymmetric VB is also observed for
both RAs. For RA of 10 , VB occurs at (x/c) ¼ 0.63
on the DLE side of the wing, and (x/c) ¼ 0.80 for the

Downloaded from pig.sagepub.com at TOHOKU UNIV on August 21, 2015

Woodiga et al. 9

Figure 11. Skin friction lines on a 76/40 double-delta wing with a diamond juncture fillet at AoA ¼ (a) 10 and (b) 20 .

Figure 12. Skin friction lines on a 76/40 double-delta wing with a diamond juncture fillet at (AoA, YA) ¼ (a) (10 , 10 ) and
(b) (10 , 15 ).

ULE side of the wing. As RA is increased to 20 , VB
moves upstream toward the apex of the wing at
(x/c) ¼ 0.53 on the DLE side of the wing and down-
stream toward the TE of the wing at (x/c) ¼ 0.94 for
the ULE side of the wing in a similar trend to the
baseline wing.
Wing with parabolic fillet
Figures 14(a) and (b), 15(a) and (b) and 16(a) and (b)
show the skin friction lines on the 76/40 double-
delta wing with a parabolic juncture fillet for the
six test cases of AoA ¼ 10 , AoA ¼ 20 , (AoA,
YA) ¼ (10 , 10 ), (AoA, YA) ¼ (10 , 15 ), (AoA,
RA) ¼ (20 , 10 ), and (AoA, RA) ¼ (20 , 20 ),
respectively.
As shown in Figure 14(a), at AoA of 10 , the pri-
mary reattachment line is clearly visible along with the
secondary separation and reattachment lines. As dis-
cussed earlier, the purpose of the parabolic fillet is to
create a single dominant vortex system and it appears
to achieve that goal as illustrated by the skin friction
lines in Figure 14(a). The reattachment lines curve
outboard at (x/c) ¼ 0.6 indicating interaction between
the strake and wing vortices. As AoA is increased to
20 asymmetric VB is observed at (x/c) ¼ 0.75 for the
left side LEV and (x/c) ¼ 0.68 for the right side LEV
as seen in Figure 14(b).

Downloaded from pig.sagepub.com at TOHOKU UNIV on August 21, 2015

10 Proc IMechE Part G: J Aerospace Engineering 0(0)

Figure 13. Skin friction lines on a 76/40 double-delta wing with a diamond juncture fillet at (AoA, RA) ¼ (a) (20 , 10 ) and (b) (20 ,
20 ).

Figure 14. Skin friction lines on a 76/40 double-delta wing with a parabolic juncture fillet at AoA ¼ (a) 10 and (b) 20 .

In the case of the parabolic wing with YA of 10 ,
the windward side LEV reattachment line is seen as
illustrated in Figure 15(a). A reattachment line and a
separation line are also observed on the leeward side
of the wing from approximately (x/c) ¼ 0.65 to the TE
of the wing. When YA is increased to 15 , the
reattachment line of the primary LEV on the wind-
ward side moves inboard to (y/c) ¼ 0.05 from
(y/c) ¼ 0.08 at (x/c) ¼ 0.6 for YA of 10 . The reattach-
ment line previously observed on the leeward side of
the wing is not present for YA of 15 as illustrated by
Figure 15(b). Similar to the baseline and diamond
juncture fillet wings, the VB location moves upstream
from (x/c) ¼ 0.68 to (x/c) ¼ 0.63 as YA is increased
from 10 to 15 .
Figure 16(a) and (b) shows the skin friction lines on
the wing at (AoA, RA) ¼ (20 , 10 ) and (20 , 20 ). The
primary reattachment lines are observed to move
inboard toward the ULE side of the wing from
(y/c) ¼ –0.03 at (x/c) ¼ 0.2 for the wing at RA of 0 to
(y/c) ¼ 0 at (x/c) ¼ 0.2 for the wing at RA of both 10
and 20 . Asymmetric VB is also observed for both RAs
similar to the baseline and diamond fillet wings. For
RA of 10 , VB occurs at (x/c) ¼ 0.6 on the DLE side of
the wing, and (x/c) ¼ 0.81 for the ULE side of the wing.
As RA is increased to 20 , VB moves upstream toward
the apex of the wing at (x/c) ¼ 0.5 on the DLE side of
the wing and downstream toward the TE of the wing at
(x/c) ¼ 0.82 for the ULE side of the wing in a similar
trend to the baseline and diamond fillet wings.

Downloaded from pig.sagepub.com at TOHOKU UNIV on August 21, 2015

Woodiga et al. 11

Figure 15. Skin friction lines on a 76/40 double-delta wing with a parabolic juncture fillet at (AoA, YA) ¼ (a) (10 , 10 ) and
(b) (10 , 15 ).

Figure 16. Skin friction lines on a 76/40 double-delta wing with a diamond juncture fillet at (AoA, RA) ¼ (a) (20 , 10 ) and
(b) (20 , 20 ).

friction vector field, the Poincare–Bendixson index

Topological consistency
Surface visualization techniques using silicone oil with
titanium dioxide mixture and color dye have been
used traditionally to obtain information of skin fric-
tion topology in wind tunnel tests. Since there is ambi-
guity in inferring the topological features from flow
visualizations, the topological rules have been pro-
posed as useful constraints.35–38 According to the
topological rules, singular points on delta wings
were tallied in previous oil streak visualizations.10
To examine the topological consistency of a skin
formula (the P–B formula in short) can be applied,39
which represents a conservation law for the numbers
of isolated singular points enclosed by a closed
boundary and boundary switch points. The P–B for-
mula is given by #N  #S ¼ 1 þ ð #Zþ  #Z  Þ=2,
where #N denotes the number of the nodes, #S the
number of the saddles, #Zþ the number of the posi-
tive switch points, and #Z  the number of the nega-
tive switch points. A switch point Z is a point dividing
neighboring inflow and outflow segments on a bound-
ary. When vectors around Z move inward first and

Downloaded from pig.sagepub.com at TOHOKU UNIV on August 21, 2015

12
then outward across the boundary, the switch point is
negative, and it is denoted by Z . Otherwise, Z is
positive, which is denoted by Zþ . The positive and
negative switch points on a boundary are illustrated
in Figure 17.
Near-wall streamline patterns on a low sweep angle
delta wing were measured by Yavuz et al.40 using
near-wall particle image velocimetry. The detailed sur-
face streamline patterns enclosed by the leading and
TEs of the delta wing as a closed penetrable boundary
can be analyzed using the P–B formula. For instance,
in the time-averaged surface streamline pattern on the
stationary delta wing (Figure 2 in Yavuz et al.40), two
Figure 17. Illustration of the positive and negative switch
points.
Figure 18. Skin friction lines on a baseline 76/40 double-delta wing at (AoA, YA) ¼ (a) (16 , 6 ) and (b) (24 , 6 ).
Downloaded from pig.sagepub.com at TOHOKU UNIV on August 21, 2015
Proc IMechE Part G: J Aerospace Engineering 0(0)
saddles and two sink nodes are identified near the LE
on the wing upper surface. Two positive switch points
associated with the large-scale swirling pattern of the
streamlines are identified on the TE. Two negative
switch points located at the sharp wing tips along
with two more negative switch points on the TE are
identified between the saddle and the wing tip on each
side of the wing. Counting the number of singular
points and boundary switch points we have, #N ¼ 2,
#S ¼ 2, #Zþ ¼ 2, and #Z– ¼ 4. Since #N 
#S ¼ 1 þ ð #Zþ  #Z  Þ=2 ¼ 0 according to the
P–B formula, the conservation law is satisfied.
Figure 18(a) and (b) are skin friction lines on the
baseline 76/40 double-delta wing at (AoA,
YA) ¼ (16 , 6 ) and (24 , 6 ), respectively. The distinct
features visible are saddle (S1) and the spiraling node
(N1) that occur on the left side of the wing due to the
YA. The combination of S1–N1 indicates a typical
closed type separation. Complex inflow and outflow
behavior along the right side LE is observed for both
cases. To simplify topological analysis, a polygon
ABCDEFA is considered, which encloses all the sin-
gular points excluding the right side LE. The singular
points and boundary switch points are indicated in
Figure 18(a) and (b), where #N ¼ 2, #S ¼ 2, #Z– ¼ 1,
and #Z– ¼ 3 for both cases of AoA of 16 and 24 .
Thus, for the measured skin friction field in the
domain enclosed by the polygon ABCDEFA for
both cases, the conservation law is satisfied. The mag-
nified views of the singular points and boundary
switch points are shown in Figure 19(a) and (b) for
AoA of 16 and 24 , respectively. The conservation
law is satisfied in these local domains as well.
Woodiga et al.
Figure 19. Magnified views of singular points on the skin friction fields of a baseline 76/40 double-delta wing at (AoA, YA) ¼
(a) (16 , 6 ) and (b) (24 , 6 ).
Conclusions
Skin friction fields are evaluated for different combin-
ations of the AoA, YA, and RA. For the 65 delta
wing, the inboard movement of the reattachment and
separation lines as increasing AoA is observed. As YA
is introduced, VB is observed for the windward side
LEV which moves upstream toward the apex of the
wing with increasing YA. As RA is introduced, no
distinct changes are observed in the skin friction
fields except for the inboard movement of the
reattachment and secondary separation lines toward
the DLE side of the wing.
For the 76/40 double-delta wing, the effects on the
skin friction fields generally follow the same trend for
all three wing configurations. As YA is introduced to
the wings, the primary reattachment line of the wind-
ward side LEV is observed to move inboard toward
the leeward side of the wings. The VB location is
observed to move upstream as YA is increased. The
asymmetric VB is observed when RA is introduced.
The VB location for the DLE side moves upstream
toward the apex of the wing while the VB location for
the ULE side moves downstream toward the wing
TE with increasing RA. The skin friction fields of
the 76/40 baseline double-delta wing at (AoA,
YA) ¼ (16 , 6 ) and (24 , 6 ) are analyzed as typical
cases by applying the Poincare–Bendixson index for-
mula. The conservation law is found to be satisfied for
all cases.
The GLOF method can provide rapid global
skin friction diagnostics for complex vortical
flows over delta wings at different attitudes, and the
results are useful for the design of delta and delta-like
wings since the skin friction topological features
are faithful footprints of three-dimensional flow phe-
nomena. Future work will be a parametric study on
Downloaded from pig.sagepub.com at TOHOKU UNIV on August 21, 2015
13
the effects of combined RA and YA in a broad range
of AoAs.
Funding
This research received no specific grant from any
funding agency in the public, commercial, or not-for-profit
sectors.
Conflict of interest
None declared.
References
1. Gursul I. Recent development in delta wing aerodynam-
ics. Aeronaut J 2004; 108: 437–452.
2. Gursul I, Gordinier R and Visbal M. Unsteady aero-
dynamics of nonslender delta wings. Prog Aerosp Sci
2005; 41: 515–557.
3. Narayan KY and Seshadri SN. Types of flow on the lee
side of delta wings. Prog Aerosp Sci 1997; 33: 167–257.
4. Hebbar SK, Platzer MF and Fritzelas AE. Reynolds
number effects on the vortical-flow structure generated
by a double-delta wing. Exp Fluids 2000; 28: 206–216.
5. Hebbar SK, Platzer MF and Khozam A. Investigation
into the effects of juncture fillets on the vertical flow over
a cropped, double-delta wing. AIAA Paper 94-0626,
1994.
6. Erickson GE and Gonzalez HA. Pressure-sensitive paint
investigation of double-delta wing vortex flow manipu-
lation. AIAA Paper 2005-1059, 2005.
7. Verhaagen NG, Jenkins LN, Kern SB, et al. A study of
the vortex flow over a 76/40-deg double-delta wing.
NASA/ICASE-95-5, 1995.
8. Kern SB. Numerical investigation of vortex flow-control
through small geometry modifications at the strake/wing
junction of a cropped double-delta wing. AIAA Paper
92-0411, 1992.
9. Kern SB. Vortex flow control using fillets on a double-
delta wing. J Aircraft 1993; 30: 818–825.
14 Proc IMechE Part G: J Aerospace Engineering 0(0)

10. Huang XZ, Hanff ES and Jobe CE. Surface flow top- 29. Sahin B, Yayla S and Canpolat C. Flow structure over
ology on a delta wing at high incidence for a range of the yawed nonslender diamond wing. Aerosp Sci
roll angles. AIAA Paper 96-2398, 1996. Technol 2012; 23: 108–119.
11. Délery JM. Physics of vortical flows. J Aircraft 1992; 30. Canpolat C, Yayla S and Sahin B. Observation of the
29: 856–876. vortical flow over a yawed delta wing. J Aerosp Eng
12. Tobak M and Peake DJ. Topology of three-dimen- 2012; 25: 613–626.
sional separated flows. Annu Rev Fluid Mech 1982; 14: 31. Liu T, Montefort J, Woodiga S, et al. Global lumines-
61–85. cent oil-film skin friction meter. AIAA J 2008; 46:
13. Huang XZ and Verhaagen NG. Vortex flow behavior 476–485.
over slender delta wing configurations: experimental stu- 32. Woodiga S and Liu T. Skin friction fields on delta
dies, numerical and analytical solutions. Berlin: Springer, wings. Exp Fluids 2009; 47: 897–911.
2008. 33. Liu T. Extraction of skin-friction fields from surface
14. Verhaagen NG. Effects of Reynolds number on flow flow visualizations as an inverse problem. Meas Sci
over 76/40-degree double-delta wings. J Aircraft 2002; Technol 2013; 24: 124004.
39: 1045–1052. 34. Liu T and Shen L. Fluid flow and optical flow. J Fluid
15. Verhaagen NG and Jobe CE. Wind-tunnel study on a Mech 2008; 614: 253–291.
65-deg delta wing at sideslip. J Aircraft 2003; 40: 35. Lighthill MJ. Attachment and separation in three-
290–296. dimensional flows. In: Rosenhead (ed.) Laminar bound-
16. Van Bossuyt BC. Experimental investigation of the ary layer theory, Sect. II 2.6. New York, UK: Oxford
vortex flow over the apex of a delta wing. MS Thesis, University Press, 1963, pp.72–82.
Department of Aerospace Engineering, Delft University 36. Hunt JCR, Abell CJ, Peterka JA, et al. Kinematical
of Technology, 2005. studies of the flows around free or surface-mounted
17. Fuji K and Schiff LB. Numerical simulation of vortical obstacles; applying topology to flow visualization.
flows over strake-delta wing. AIAA J 1989; 27: J Fluid Mech 1978; 88: 179–200.
1153–1162. 37. Foss JF. Surface selections and topological constraint
18. Rizzi A and Miillerf B. Large-scale viscous simulation evaluations for flow field analyses. Exp Fluids 2004; 37:
of laminar vortex flow over a delta wing. AIAA J 1989; 883–897.
27: 833–840. 38. Foss JF. Topological considerations in fluid mechanics
19. Ekaterinaris JA, Coutleyt RL, Schifft LB, et al. measurements. In: Handbook of experimental fluid
Numerical investigation of high incidence flow over a mechanics, Berlin, Germany: Springer, 2007, Chapter
double-delta wing. J Aircraft 1995; 32: 457–463. 13.
20. Gordnier RE. Numerical simulation of a 65-degree 39. Liu T, Woodiga S and Ma T. Skin friction topology in a
delta-wing flow. J Aircraft 1997; 34: 492–499. region enclosed by penetrable boundary. Exp Fluids
21. Verhaagen NG and Naarding SHJ. Experimental and 2011; 51: 1549–1562.
numerical investigation of the vortex flow over a yawed 40. Yavuz MM, Elkhoury M and Rockwell D. Near-sur-
delta wing. AIAA Paper 88-2563, 1988. face topology and flow structure on a delta wing. AIAA
22. Verhaagen NG and Jobe CE. Topology of the flow on a J 2004; 42: 332–340.
65-deg delta wing at sideslip. AIAA Paper 2002-0095,
2002. Appendix 1
23. Verhaagen NG and Kunst VF. Topology of the flow on
delta wings at sideslip. AIAA Paper 2005-60, 2005. Notation
24. Gordnier RE. Computation of delta wing roll maneu-
vers. J Aircraft 1995; 32: 486–492.
c root chard (m)
25. Grismer D and Nelson R. Aerodynamic effects of N node
sideslip on double delta wings. AIAA Paper 93-0053, #N number of nodes
1993. Re Reynolds number
26. Hebbar SK, Platzer MF and Chang WH. Juncture S saddle
fillets for vortex flow control on double-delta wings #S number of saddles
undergoing sideslip. AIAA Paper 96-0663, 1996. x, y chordwise and spanwise coordinates
27. Canpolat C, Yayla S, Sahin B, et al. Dye visualization (m)
of the flow structure over a yawed nonslender delta Zþ positive switch point
wing. J Aircraft 2009; 46: 1818–1822. #Zþ number of positive switch points
28. Yayla S, Canpolat C, Sahin B, et al. Yaw angle effect on
Z negative switch point
flow structure over the nonslender diamond wing.
AIAA J 2010; 48: 2457–2461.
#Z  number of negative switch points

Downloaded from pig.sagepub.com at TOHOKU UNIV on August 21, 2015