Sei sulla pagina 1di 23

AIAA 2015-0771

AIAA SciTech
5-9 January 2015, Kissimmee, Florida
53rd AIAA Aerospace Sciences Meeting

C m ai al Ae d amic A al i f A la Wi g
U ma ed Ae ial Vehicle

A. A. Kanoria*, K. Panchal and M. Damodaran

Indian Institute of Technology Gandhinagar,


Mechanical Engineering Discipline,
VGEC Complex, Visat-Gandhinagar Highway,
Chandkheda, Ahmedabad - 382424, GJ, INDIA

ABSTRACT
Downloaded by ROKETSAN MISSLES INC. on January 15, 2015 | http://arc.aiaa.org | DOI: 10.2514/6.2015-0771

Annular or Ring wing concept is a form of closed non-planar wing. The benefits of non-
planar wing aerodynamics are well known and have been elucidated theoretically and
experimentally. The widespread use of non-planar wings during the early years of human-
powered flight had more to do with structural expediency than augmented aerodynamic
efficiency. Non-planarity for a confined wing span may improve wing efficiency by capturing
larger volume of air to generate lift impulse. In this study, the aerodynamic characteristics
of an annular wing composed of symmetric airfoil section NACA 0012 is investigated for
various aspect ratios ranging from 0.5 to 2. The computational modeling showed higher lift
coefficient for aspect ratio of 2 and the resulting L/D ratio is observed to be best for this
aspect ratio. To study effect of asymmetry, a cambered Clark-Y airfoil is considered as a
wing section for the annular wing. The Clark-Y annular wing showed better aerodynamic
characteristics with a stall angle higher than that of the NACA 0012 annular wing and
attaining better L/D ratio at lower angle of attack. Further, a brief comparison between
annular wing and conventional wing shows that a higher lift coefficient is observed for
planar wing but annular wing admitted a higher stall angle. Computational results are
compared with available theoretical results and experimental data. Owing to the benefits
offered by cambered airfoil, an annular wing with Clark-Y airfoil is considered for an
annular wing UAV design configuration with a V-tail empennage adopted for the UAV
configuration

Nomenclature
AR = Aspect ratio
CL = Lift Coefficient
CD = Drag Coefficient
CP = Pressure Coefficient
L/D = Lift to Drag ratio
acw = Aerodynamic Centre of Annular Wing
act = Aerodynamic center of V-tail
Xg = Distance between Aerodynamic center of wing and center of gravity of UAV
Xt = Distance between Aerodynamic centers of wing and of horizontal tail
Xn = Distance between Aerodynamic center and neutral point
Vt = Volume Tail ratio,
St = Tail Surface area
Sw = Wing surface area
Cw = Wing chord length
*
Graduate Student and DEITY Junior Research Fellow, Email: a.kanoria@iitgn.ac.in
Project Engineer, Email: kartikpanchal@iitgn.ac.in
Professor, Associate Fellow AIAA, Email: murali@iitgn.ac.in
1
American Institute of Aeronautics and Astronautics

Copyright © 2015 by the American Institute of Aeronautics and Astronautics, Inc. All rights reserved.
I. Introduction

A nnular or Ring wing concept owing to its increased aerodynamic efficiency has received recent attention in the
applied aerodynamics research community. Considering its structural complexities, its implementation is
difficult for large scale passenger or military aircrafts. Flecter1 reports that initial attempts use it in 1960 for viable
air transport were not successful. The aerodynamic benefits of circular wings have recently prompted significant
interest in this concept among UAV designers such as Ko et al.2 and Traub3. The aerodynamic benefits of non-
planar bi-plane wings have been outlined in Prandtl and Tietjens 4 and Munk5. Nonplanar closed wing span may
improve wing efficiency by capturing a larger volume of air to generate the lift impulse. Despite of continuous
attempt to seek the advantages offered by annular wing it has not been fully studied and have not received
significant attention from aerodynamicists for a long time until recent times. Closed nonplanar wings in the form of
both annular (ring) or box wings have also been investigated theoretically and experimentally by Demasi6, Ribner7
and Stewart8 . Ribner7 derived the lift characteristics of annular wings analytically and concluded that the resulting
lift-curve slope is twice that of flat plate elliptical wings of the same aspect ratio. The most recent experimental
investigations on annular wings reported inTraub3 compared the effects of aspect ratio (AR), which is defined as the
Downloaded by ROKETSAN MISSLES INC. on January 15, 2015 | http://arc.aiaa.org | DOI: 10.2514/6.2015-0771

ratio of the annular wing diameter to the chord of the airfoil section used in the annular wing, on the lift-curve
slopes with theoretical predictions. The investigation included the longitudinal aerodynamic characteristics of five
annular wings with AR of 1/3, 2/3, 1.0, 1.5, and 3.0 with the same projected areas and which used the Clark-Y
airfoil cross section (thickness-to-chord ratio of 11.7%). The Reynolds numbers (Re) of the experiment varied in the
range 0.704 2.11 million because of the variation in the root chord of the wings for different AR. The results
indicated that the lift-curve slopes of the annular wings are about twice that of the lift-curve slopes of planar
rectangular wings having the same aspect ratio, which confirmed the findings of Ribner7. Recently, Masqood and
Go9 revisited the topic by comparing their experimental data with analytical approximation on the baseline annular
wing aerodynamics. A general paucity of computational aerodynamic analysis of annular wings in literature has
motivated the present computational study. For the present study, the aerodynamic characteristics of annular wing
are assessed for possible use in Unmanned Aerial Vehicle (UAV) design concept. Three different airfoil sections,
i.e., (a) Symmetrical section NACA 0012 (labelled as AW1), (b) Cambered section geometry is positive camber
CLARK-Y on both upper and lower halves which are blended by creating a loft between the two halves using a
symmetrical NACA 0012 cross section (labelled as AW2) and (c) Cambered section geometry is CLARK-Y
section in circular loop with lower part seeing the negative camber (labelled as AW3) as shown in Fig 1, have been
used to compare and assess the respective aerodynamic benefits.

(a) (b) (c)

(d)
Figure 1: Annular Wing Configurations with Diameter (Span) of 0.3 m and AR of 2 (a) AW1, (b)
AW2, (c) AW3 and (d) UAV model with AW2 annular wing

For the present study which follows Traub4, three different annular wing geometries with AR of 1/2, 1, 2 are
considered in order to assess the effects of AR on the aerodynamic characteristics on AW1 annular wing. To study
the effects cambered wing section on the annular wing, AW2 wing geometry with an AR of 2 is analyzed and the

2
American Institute of Aeronautics and Astronautics
results are compared with those from AW1. Further, the aerodynamic characteristics of a conventional rectangular
planar wing composed of cambered airfoil, AW2 and AW3 are also compared. To validate the computational
results, aerodynamic characteristics of AW2 are compared with analytical results and available experimental data.
Finally, in view of the benefits of using cambered airfoils in the annular wing, a computational assessment of the
aerodynamic characteristics of an UAV design configuration using AW2 wings is done.

II. Mathematical and Computation Modeling


In this section, the governing equations and entire numerical procedure of solving has been discussed. The
governing equations which describes the air flow around the annular wing are the unsteady Navier-Stokes equations
given as,
U .ndS 0 (1a)
S
Downloaded by ROKETSAN MISSLES INC. on January 15, 2015 | http://arc.aiaa.org | DOI: 10.2514/6.2015-0771

d
UdV U .ndS pndS ndS f dV
dt V S S S V (1b)

where U is the local velocity, p is the pressure, is the viscous stress tensor which contains both laminar and
turbulent stress components which are modelled using a RANS turbulence model, f is the body force term and
is the density of the fluid. The convective flux in the momentum equation 1(b) has been discretized using the second
order upwind scheme and the time integration has been carried out using global time stepping, all within the
framework of the Star-CCM+10 CFD software. Fig 2(a) shows the computational domain used for modeling the
aerodynamics of AW1 and AW2 annular wings and Fig 2(b) show computational domain used for modeling UAV
model.

(a) (b)
Figure 2: Computational domain along with different boundary region (a) Annular Wing (b) UAV Model

For the Fig 2(a), an uniform free stream inflow velocity is specified at the upstream boundary, i.e., Face ABCD,
and far field boundary, i.e., Faces BFGC, CDHG and ABFE and a pressure outflow boundary, i.e. Face EFGH and
symmetry plane, i.e. Face ADHE. The no slip boundary condition is imposed on the annular wing surface. The
inflow velocity is set at 40 m/s giving a Reynolds number of 0.26 million for results discussed in section IIIA and
IIIB, while inflow velocity is set at 60 m/s and 90 m/s giving a Reynolds number range of 0.7- 2.12 million for
results discussed in section IIIC and IIID respectively. In Fig 2(b), a uniform free stream inlet velocity is specified
at the upstream boundary, i.e., Face ABCD, and far field boundary, i.e., Faces BFGC, CDHG, ADHE and ABFE
and a pressure outflow boundary, i.e. Face EFGH. The no slip boundary condition is imposed on the UAV model.
The inflow velocity is set at 60 m/s giving a Reynolds number of 0.58 million for results discussed in section IIIE.

Figures 3(a)-(b) and 3(c)-(d) show the computational mesh for AW1 of AR 1 and AW2 of AR 2 respectively.
The mesh generated for AW1 geometry has 1.17 million cells while the AW2 geometry has 1.25 million cells.
Fig. 3(e)-(f) shows the computational mesh of about 4.2 million cells around the UAV model. The mesh is refined at

3
American Institute of Aeronautics and Astronautics
the wing leading and trailing edges, as flow in these regions is complex and requires higher mesh resolution. The
computational mesh is refined in the wake region to effectively capture the trailing vortex system of the annular
wing. Polyhedral mesh is used for the computations with mesh refinement at critical points and in the wake regions
to effectively capture the trailing vortices shed from the wing surfaces. On the basis of the estimated y+ value of 5,
about 15 prism layers are stacked near wall boundaries to effectively capture the boundary layer formation.
Downloaded by ROKETSAN MISSLES INC. on January 15, 2015 | http://arc.aiaa.org | DOI: 10.2514/6.2015-0771

(a) (b)

(c) (d)

(e) (f)
Figure 3: Three dimensional polyhedral mesh and zoomed view of prism layers on the wing surface (a)-(b) of model
AW1 of AR 1 (c)-(d) of model AW2 of AR 2 (e)-(f) Views of mesh structure in the vicinity of the UAV model.

4
American Institute of Aeronautics and Astronautics
III. Results and Discussion

A. Flow past a AW1 Annular wing

1. Mesh Convergence Studies

Mesh convergence studies for the flow field computation around the annular wing model AW1 with an AR of 1
are inclined at 00 angle of attack with flow velocity of 40 m/s is shown in Figure 4 which shows the variation of
computed lift and drag coefficients with the number of mesh cells N used in the computation. It can be seen for a
mesh consisting of 1.17 million cells, both lift and drag coefficients seem to attain constant values asymptotically.

0.004 0.016

0.0035 0.014
Downloaded by ROKETSAN MISSLES INC. on January 15, 2015 | http://arc.aiaa.org | DOI: 10.2514/6.2015-0771

0.003 0.012

Drag Coefficient
Lift Coefficient

0.0025 0.01

0.002 0.008

0.0015 0.006

0.001 0.004

0.0005 0.002

0 0
0.00E+00 2.40E+05 4.80E+05 7.20E+05 9.60E+05 1.20E+06 0.00E+00 2.40E+05 4.80E+05 7.20E+05 9.60E+05 1.20E+06
Mesh Size (N) Mesh Size (N)
(a) (b)
Figure 4: Mesh convergence studies: Variation of (a) lift and (b) drag coefficients vs. number of mesh points N

Apart from mesh dependency study, it is necessary to examine spatial (grid) convergence using Richardson
extrapolation and estimating the Grid convergence index (GCI) as outlined in Roach11 by considering three levels of
mesh density. As the order of convergence, which is defined theoretically for a numerical algorithm, may not remain
the same due to boundary conditions, quality of mesh, numerical schemes, etc. an estimate of the order of
1 1
convergence, p, can be obtained from the results as p ln f3 f2 f2 f1 ln r where f1 , f2 and f3 ,are
h1
the drag coefficients computed on the finest, medium and the coarsest mesh respectively and r is the ratio of
h2
sizing of the mesh on the wing surface. The GCI is a measure of the percentage the computed value differs from the
value of the asymptotic numerical value and indicates an error band on how far the solution is from the asymptotic
value. It indicates how much the solution would change with a further refinement of the grid. A small value
of GCI indicates that the computation is within the asymptotic range. Its value for the finest grid can be computed
F f2 f1
using GCI fine p
where , and F is a factor of safety, which multiplies the GCI for a more
r 1 f1
conservative reporting of the error. The Factor of safety (Fs), in the above expression is taken to be 1.25 for study
involving more than three mesh counts. A validation check is done to see if the solution is well within the
GCI 32
asymptotic range of convergence by evaluating ASC which works out to 1.03, indicating that the
r p GCI 21
computed solution is in the range of convergence.

5
American Institute of Aeronautics and Astronautics
2. Effect of Aspect Ratio on Annular Wing Aerodynamic Coefficients

The effect of AR on the aerodynamic characteristics of the annular wing is assessed by computing the lift and
drag coefficients by varying the angle of attack in the range 0 to 18 o for all the three geometries. The lift and drag
coefficients are monitored to assess convergence before accepting the steady state value. The planform area for
computing the aerodynamic coefficients is taken to product of diameter which is the wing span and the chord of the
wing. Figure 5 shows the convergence of the computed lift coefficient as the angle of attack is varied in the range
with number of iterations. The angle of attack is varied statically keeping using the previously computed flow field
as initial conditions to achieve faster convergence and more accurate value as the angle of attack is changed.
Downloaded by ROKETSAN MISSLES INC. on January 15, 2015 | http://arc.aiaa.org | DOI: 10.2514/6.2015-0771

Figure 5: Convergence of lift coefficient with No. of iterations as angle of attack is varied from 0-18o

Figures 6(a) and (b) show the computed aerodynamic lift and drag characteristics for all the 3 models of annular
wing with varying aspect ratios.

(a) (b)
Figure 6: Effect of Aspect Ratio on the computed Aerodynamic (a) Lift and (b) Drag Characteristics of Annular Wings

It can be clearly seen that the lift coefficients for the annular wing with AR of 2 is higher compared to the other
two cases of lower AR. Figure 7 shows the computed aerodynamic efficiency (Lift to Drag ratio) for all three angles
of attack. The maximum L/D ratio is estimated to be around 120 angle of attack for all the three cases. Stall is
observed for the annular wing with an AR of 2 at 160 angle of attack. The flow contours corresponding to these
angles are examined to visualize the flow over the wing.

6
American Institute of Aeronautics and Astronautics
Downloaded by ROKETSAN MISSLES INC. on January 15, 2015 | http://arc.aiaa.org | DOI: 10.2514/6.2015-0771

Figure 7: Lift to Drag ratio for varying Aspect Ratio geometries

3. Computed Flow Contours for Annular Wings at 120 Angle of Attack

Figures 8 and 9 show selected computed velocity field and wing surface shear stress contours respectively
corresponding to 120 angle of attack for all the three annular wing geometries. It can be seen from Fig. 8 that the
flow tends to separate for annular wing with AR of ½ from leading edge compared to rest of the geometries. In Fig.
9, wall shear stress contours demonstrate flow separation at trailing edges of upper and lower surfaces of the wings.
For annular wing of AR of ½, the flow tends to reattach to the surface at the end of the upper wing. The extended lift
production is because, no flow separation is observed at the lower portion at high angle of attack. These plots
highlight the benefits of delayed stall in annular wings over the conventional rectangular wings.

Figure 10 shows the computed velocity contours on selected Trefftz planes aft of the annular wing which appears
to diminish at a distance approximately four times the chord of the wing from the trailing edge of the wing. It can be
seen that for annular wing with AR of ½, high disturbances are observed in the wake regions because of
flow separation from the upper part of the wing, while the disturbances appear to be low for the annular wing with
AR of 1. Figure 11 compares the distribution of wing surface pressure coefficient (C p) and on the symmetry
plane. For annular wings with AR of ½, low pressure is observed at the suction side of upper part of wing compared
to that on the annular wing with AR of 1 and 2, resulting into lower values of lift coefficient observed in Fig.6 (a).
Also, Cp of +1 at leading edge of all three geometries defines the stagnation points for the flow past the annular
wings.

(a) (b) (c)


Figure 8: Computed velocity field contours for annular wings with AR of (a) ½, (b) 1 and (c) 2

7
American Institute of Aeronautics and Astronautics
(a) (b) (c)
Figure 9: Computed wing surface shear stress contours for annular wings with AR with (a) ½, (b) 1 and (c) 2
Downloaded by ROKETSAN MISSLES INC. on January 15, 2015 | http://arc.aiaa.org | DOI: 10.2514/6.2015-0771

(a) (b) (c)


Figure 10: Computed velocity contours on Trefftz planes for annular wings with AR of (a) ½, (b) 1 and (c) 2

(a) (b) (c)


Figure 11: Computed Pressure Coefficient contours for annular wings with AR of (a) ½, (b) 1 and (c) 2

4. Computed Flow Contours in the Vicinity of Annular Wings at 180 Angle of Attack

Figures 12 and 13 show computed velocity field contours and pressure contours in the flow field, symmetry
plane and wing surface at 180 angle of attack for all the three annular wing geometries. In this case, compared to
previous case results, flow separation is clearly visible from the upper wing portion for all the three AR cases. The
dominant lift at higher angle is produced by the lower wing portion where the separation has not occurred yet. Fig.
14 shows the computed skin friction lines on the body of the geometries. Three-dimensional separation topologies
are visible on the upper wing of geometry, while flow separation is also visible near the trailing edges of the lower
wing portion. It can be seen that the flow separation for annular wing with AR of 1, is quite delayed compared to
remaining two configurations, thereby producing more lift, which is seen in the lift characteristics graph in Fig. 6(a).
The computed streamline patterns are shown in Fig. 15 for all the three geometries which show that the flow past the
lower wing portion of the annular wing is quite uniform compared to the upper wing portion.

8
American Institute of Aeronautics and Astronautics
(a) (b) (c)
Figure 12: Computed Velocity contours for annular wings with AR of (a) ½, (b) 1 and (c) 2
Downloaded by ROKETSAN MISSLES INC. on January 15, 2015 | http://arc.aiaa.org | DOI: 10.2514/6.2015-0771

(a) (b) (c)


Figure 13: Computed pressure coefficient contours for annular wings with AR of (a) ½, (b) 1 and (c) 2

(a) (b) (c)


Figure 14: Computed skin friction lines over surface for annular wings with AR of (a) ½, (b) 1 and (c) 2

(a) (b) (c)


Figure 15: Computed streamlines of flow past annular wings with AR of (a) ½, (b) 1 and (c) 2

9
American Institute of Aeronautics and Astronautics
B. Flow past the Annular Wing Configuration AW2

1. Mesh Dependency Study

Mesh convergence studies for the flow field computation around the annular wing model AW2 consisting of
cambered airfoils with an AR of 2 inclined at 00 angle of attack at a free stream velocity of 40 m/s is shown in
Figure 16 which shows the variation of computed lift and drag coefficients with the number of mesh cells N used in
the computation. It can be seen for a mesh consisting of 1.25 million cells, both lift and drag coefficients seem to
attain constant values asymptotically.

0.2 0.04
0.18
0.035
0.16
0.03

Drag Coefficient
0.14
Lift Coefficient
Downloaded by ROKETSAN MISSLES INC. on January 15, 2015 | http://arc.aiaa.org | DOI: 10.2514/6.2015-0771

0.12 0.025
0.1 0.02
0.08 0.015
0.06
0.01
0.04
0.02 0.005

0 0
0.00E+00 2.50E+05 5.00E+05 7.50E+05 1.00E+06 1.25E+06 0.00E+00 2.50E+05 5.00E+05 7.50E+05 1.00E+06 1.25E+06
N (Mesh Size) N (Mesh Size)
(a) (b)
Figure 16: Mesh convergence studies: Variation of (a) lift and (b) drag coefficients vs. number of mesh points N

Apart from mesh dependency study, it is also necessary to examine the spatial (grid) convergence using
Richardson extrapolation and hence finding the Grid convergence index. The calculation for the same has been
carried out in similar fashion as outlined for the earlier annular wing AW1 and the solution is well within the
asymptotic range of convergence which for this case is estimated at 1.05, indicating that the computed solution is
within the range of convergence.

2. Comparing Computed Aerodynamic Coefficients of Annular Wing Configurations AW1 and AW2

In order to assess the effect of annular wing sections on aerodynamic characteristics, Figures 17 (a) and (b)
compares the computed aerodynamic lift and drag characteristics of the annular wing with AR of 2 for the AW2
configurations which use cambered airfoils. It can be seen that the AW2 configuration gave a higher lift coefficient
with a higher maximum lift coefficient and also a higher stall angle of 24o. The computed drag coefficient is low at
smaller angles of attack up to 140 angle of attack resulting in aerodynamic efficiency of AW2 configuration being
higher than that of the AW1 configuration as can be seen in Fig.18 which shows the variation of aerodynamic
efficiency with angle of attack.

10
American Institute of Aeronautics and Astronautics
1 0.3
0.9
0.25
0.8

Drag Coefficient
0.7
Lift Coefficient

0.2
0.6
0.5 0.15
0.4
0.1
0.3 AW1
0.2 AW1
AW2 0.05
0.1 AW2

0 0
-4 -2 0 2 4 6 8 10 12 14 16 18 20 22 24 26 0 2 4 6 8 10 12 14 16 18 20 22 24 26
Angle of Attack Angle of Attack
Downloaded by ROKETSAN MISSLES INC. on January 15, 2015 | http://arc.aiaa.org | DOI: 10.2514/6.2015-0771

(a) Lift Coefficient (b) Drag Coefficient


Figure 17: Comparison of Aerodynamic Coefficient for AW1 and AW2 Annular Wing

8
AW1
7
AW2
6

5
L/D Ratio

0
0 2 4 6 8 10 12 14 16 18 20 22 24 26
Angle of Attack
Figure 18: Comparison of L/D ratio for AW1 and AW2 Annular Wing

3. Computed Flow Contours of the AW2 Annular Wing Configuration at 180 and 240 angle of attack

The computed flow contours in the vicinity of the AW2 annular wing configuration at 180 and 240 angle of
attack are examined in Figures 19-22 which shows computed velocity contours, pressure coefficient contours on the
wing surfaces, symmetry plane and flowfield, skin friction lines on the wing surface showing topology of flow
separation on the surfaces and velocity contours in the wake on selected Trefftz planes aft of the wing. It can be seen
that flow separation occurs at the upper wing portions as well as from trailing edges of the lower wing portion. The
same can be observed from the skin friction line patterns on the surface of the annular wing shown in Fig. 21. A
reduction in lift is observed at 240 angle of attack because of flow separation on both the upper and lower portions of
the annular wing. The vorticity contours on the selected Trefftz planes shows the strength of vortices being shed
from the trailing of annular wing and distance at which it tends to diminish. It can be seen the vortices at 18 0 angle
of attack tend to diminish at distances three times the chord of wing while not in case of vortices shed at 240
angle of attack.

11
American Institute of Aeronautics and Astronautics
(a) (b)
0 0
Figure 19: Computed Velocity Contours of AW2 at (a) 18 and (b) 24 angle of attack
Downloaded by ROKETSAN MISSLES INC. on January 15, 2015 | http://arc.aiaa.org | DOI: 10.2514/6.2015-0771

(a) (b)
Figure 20: Computed Pressure Coefficient Contours of AW2 at (a) 18 and (b) 240 angle of attack
0

(a) (b)
0 0
Figure 21: Computed Skin friction lines on AW2 at (a) 18 and (b) 24 angle of attack

(a) (b)
Figure 22: Computed velocity contours on Trefftz planes at (a) 18 and (b) 240 angle of attack
0

C. Comparison between Planar, AW2 and AW3 Annular Wing Configurations

In the AW2 annular wing configuration, the cambered airfoils maintain the same positive camber on both the
upper and lower portions of the wing to the external flow and this requires a lofting at the intersection of the two
portions via a NACA 0012 airfoil while in the AW3 annular wing configuration the upper portion has positive
12
American Institute of Aeronautics and Astronautics
camber while the lower portion has negative camber to the external flow. In order to assess these two configurations,
aerodynamic characteristics are computed for the same flow conditions mentioned in previous section and the
computed results are compared with the corresponding equivalent rectangular planar wing in Fig. 23.

2 0.5
0.45
0.4
1.5

Drag Coefficient
0.35
Lift Coefficient

0.3
1
0.25
Planar
0.2
AW2
0.5 0.15 Planar
AW3
0.1 AW2
Downloaded by ROKETSAN MISSLES INC. on January 15, 2015 | http://arc.aiaa.org | DOI: 10.2514/6.2015-0771

0.05 AW3
0
-4 -2 0 2 4 6 8 10 12 14 16 18 20 22 24 0
0 2 4 6 8 10 12 14 16 18 20 22 24
-0.5
Angle of Attack Angle of Attack
(a) (b)

0.8 18
Planar
16 Planar
AW2
0.6 AW2
AW3 14
Moment Coefficient

AW3
0.4 12
L/D ratio

10
0.2 8
6
0
0 2 4 6 8 10 12 14 16 18 20 22 24 4
-0.2 2
0
-0.4 0 2 4 6 8 10 12 14 16 18 20 22 24
Angle of Attack Angle of Attack
(c) (d)
Figure 23: Comparison of Aerodynamic (a) Lift (b) Drag and (c) Moment Coefficients and (d) L/D Characteristics of
Annular Wing Configurations AW2 and AW3 with that of the equivalent Planar Wing

From Fig. 23(a) the computed maximum lift coefficients for planar and AW3 configurations are approximately
the same value while the AW2 configuration has a higher maximum lift coefficient. The computed stall angle of the
AW2 configuration is much higher compared to other two configurations. The moment coefficient, computed at the
quarter chord point of the all the wing configurations, for annular wing configurations AW2 and AW3 differ with
that of the planar wing as shown in Fig 23(c). As observed in Fig 23(d), the lift to drag ratio of the planar wing is
higher than both AW2 and AW2 but a wider operating angle of attack and higher stall angle makes AW2 wing more
suitable for UAV application as they are designed to have higher maneuverability.

D. Validation of Aerodynamic Coefficient with Leading-Edge Suction Analogy

A simplified concept of leading-edge suction analogy proposed by Polhamus12 for the low-aspect ratio planforms
(specifically delta wings) in the late 1960s served as a useful model for characterizing and explaining the non-linear
lift of low AR planar wings. The aerodynamic surfaces at moderate angles of attack suffer from flow separation at
or near the leading edge, which significantly alters the pressure distribution on the upper surface of the wing. The
approach assumes that the total lift in the pre-stall regime can be calculated as the sum of the component of lift from

13
American Institute of Aeronautics and Astronautics
potential flow analysis (based on fully attached pressure distribution) and another component associated with the
separated leading-edge vortices.

1.60
Computational AR= 1/3
1.40
Computational AR= 1
1.20
Computational AR= 3
Lift Coefficient

1.00 Experimental AR= 1/3


0.80 Experimental AR= 1

0.60 Experimental AR= 3


Downloaded by ROKETSAN MISSLES INC. on January 15, 2015 | http://arc.aiaa.org | DOI: 10.2514/6.2015-0771

Theoretical AR= 1/3


0.40
Theoretical AR= 1
0.20
Theoretical AR= 3
0.00
0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32
Angle of Attack

(a)

0.35

0.3 Computational AR= 1/3


Computational AR= 1
0.25
Drag Coefficient

Computational AR= 3
0.2 Experimental AR= 1/3
Experimental AR= 1
0.15
Experimental AR= 3
0.1 Theoretical AR= 1/3
Theoretical AR= 1
0.05 Theoretical AR= 3
0
0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32
Angle of Attack

(b)

Figure 24: Comparison of Aerodynamic Coefficient between Computational, Experimental and Theoretical Results
(a) Lift Coefficient (b) Drag Coefficient (c) Moment Coefficient (Continued)

14
American Institute of Aeronautics and Astronautics
0.20

0.15 Computational AR= 1/3


Computational AR= 1
0.10
Moment Coefficient

Computational AR= 3

0.05 Experimental AR= 1/3


Experimental AR= 1
0.00 Experimental AR= 3
0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32
-0.05 Theoretical AR= 1/3
Theoretical AR= 1
Downloaded by ROKETSAN MISSLES INC. on January 15, 2015 | http://arc.aiaa.org | DOI: 10.2514/6.2015-0771

-0.10 Theoretical AR= 3

-0.15
Angle of Attack
(c)
Figure 24: Comparison of Aerodynamic Coefficient between Computational, Experimental and Theoretical Results
(a) Lift Coefficient (b) Drag Coefficient (c) Moment Coefficient

On the basis of this model, Masqood and Go9 developed a similar model for the annular wings to properly model
the division of overall lift contribution from fully attached flow and vortex induced flow and compared their
estimates of aerodynamic coefficients with experimental data. Annular Wing of three different AR of 1/3, 1 and 3
were used for comparative study. Based on the present computational estimates of the aerodynamic characteristics
for these three cases considered in Masqood and Go9 are compared with the experimental and theoretical estimates
in Fig. 24.

There appears to be a good agreement for the estimates of lift coefficient and moment coefficient about quarter
chord point of annular wing as shown in Fig 24(a) & (c) respectively. However, drag coefficient shown in Fig 24(b)
appears to be over predicted in computational analysis compared to the experimental. This can be assumed due to
limitation of RANS turbulence model used for computational study and will be explored in due course using LES
models.

E. UAV model with AW2 Annular Wing

As it can be observed from the previous sections that the AW2 annular wing configuration has better
aerodynamic characteristics than the AW1 and AW3 annular wing. Hence, AW2 annular wing is used on a complete
UAV design concept configuration to study the possible enhancements in aerodynamic variables as compared with
conventional rectangular wing configuration. Two AW2 annular wing configurations are used with V-tail
configuration at the end of the fuselage as shown in Fig. 25. The AR of the annular wings is 2 considering the
advantages of this particular annular wing over all other annular wings with other aspect ratios as discussed in the
previous sections.

1. UAV Static Stability Analysis

The chord length of the annular wing is 0.15 m; the fuselage length is 1.0 meter. The tail sizing is based on the
method outlined in Tulapurkara13. The tail volume obtained is estimated to be 0.52 which is found to be within the
acceptable range. A quick stability check is carried out by calculating static stability margin. Figure 25 shows
isometric view of UAV configuration with estimated location of aerodynamic center and center of gravity. The
center of gravity for each component is estimated using CAD software CATIA and also verified using hand
calculation. Table 1 shows position of center of gravity for overall UAV configuration along with individual

15
American Institute of Aeronautics and Astronautics
components. The entire UAV is assumed to be made of Aluminum. The skin thickness is assumed to be
approximately 1 mm. The estimated static margin of -26.23 mm indicates the stable nature of UAV.
Downloaded by ROKETSAN MISSLES INC. on January 15, 2015 | http://arc.aiaa.org | DOI: 10.2514/6.2015-0771

Figure 25: Engineering Views of UAV Configuration

Table 1: Location of Center of Gravity for UAV


Part Volume Area Mass CG
Fuselage 1.952e-4m3 0.374m2 0.531kg [458.237, 0, 0] mm

Left Wing 2.92e-4m3 0.294m2 0.791kg [406.829 , 200.00, 3.767] mm

Right Wing 2.92e-4m3 0.294m2 0.791kg [406.829, -200.00, 3.767] mm

V tail 3.637e-4m3 0.076m2 0.986kg [1003.084, 0.0, 47.136] mm

Propeller 6.351e-5m3 0.032m2 0.172kg [-18.168, 0, 0] mm

Motor & Battery - - 1.5 kg [-92.00, 0.0, 0.0] mm


3 2
Total 0.002m 0.888m 4.923kg [422.846, 0.0, 11.086] mm

2. Mesh Dependency Study

Mesh convergence studies for the flow field computation around the UAV model with AW2 annular wing of AR
of 2, inclined at 00 angle of attack with flow velocity of 60 m/s is shown in Figure 26 which the variation of the
computed drag coefficient with the number of mesh cells N used in the computation. It can be seen that for a mesh
consisting of 8 million cells, the drag coefficient appears to attain a constant value asymptotically.

Apart from mesh dependency study, it is also necessary to examine the spatial (grid) convergence using
Richardson extrapolation and hence finding the Grid convergence index. The calculation for the same has been
carried out in similar fashion as outlined for the earlier annular wing AW1, AW2 and the solution is well within the
asymptotic range of convergence which for this case is estimated at 1.08, indicating that the computed solution is
within the range of convergence.

16
American Institute of Aeronautics and Astronautics
0.1
0.09
0.08
0.07

Drag Coefficient
0.06
0.05
0.04
0.03
0.02
0.01
0
Downloaded by ROKETSAN MISSLES INC. on January 15, 2015 | http://arc.aiaa.org | DOI: 10.2514/6.2015-0771

0.00E+00 2.00E+06 4.00E+06 6.00E+06 8.00E+06


N (Mesh Size)
Figure 26 : Mesh convergence studies: Variation of drag coefficients vs. number of mesh points N

3. Computed Aerodynamic Coefficients of UAV in an uniform flow

The aerodynamic characteristics of the UAV are estimated computationally by varying the angle of attack at
interval of 10 until stall is observed. The lift, drag and moment coefficients are monitored to assess the convergence
before accepting the steady value. The moment coefficient is computed around center of gravity of the UAV
configuration. Figure 27 shows the convergence of computed lift coefficient as the angle of attack is varied to stall
angle with number of iterations. The angle of attack is varied statically keeping using the previously calculated flow
field as initial condition to achieve faster convergence and more accurate value.

Figure 27: Lift coefficient convergence w.r.t. No. of iterations

The computed aerodynamic lift, drag and moment characteristics shown in Fig. 28 which shows a stall angle at
around 180 angle of attack from Fig. 28(a). The drag coefficient shown in Fig. 28(b) appears to conform to the
standard drag coefficient used in flight performance studies. Figure 28(c) shows the computed variation of the
coefficient of moment about the UAV center of gravity vs. the angle of attack. Some discrepancies in the vicinity of
the post stall angle region is observed resulting in an unusual nature of moment curve which motivates further
investigation. Figure 28(d) shows that the maximum lift to drag ratio occurs at 50 angle of attack.

17
American Institute of Aeronautics and Astronautics
2.5 0.6

0.5
2
Lift Coefficient

Drag Coefficient
0.4
1.5
0.3
1
0.2
0.5
0.1

0 0
-6 -4 -2 0 2 4 6 8 10 12 14 16 18 20 22 24 0 2 4 6 8 10 12 14 16 18 20 22 24
Downloaded by ROKETSAN MISSLES INC. on January 15, 2015 | http://arc.aiaa.org | DOI: 10.2514/6.2015-0771

-0.5 Angle of Attack Angle of Attack


(a) (b)
0.02 9
0 8
-0.02
7
Moment Coefficient

Lift to Drag Ratio


-0.04
6
-0.06
5
-0.08
4
-0.1
-0.12 3

-0.14 2
-0.16 1
-0.18 0
0 2 4 6 8 10 12 14 16 18 20 22 24 0 2 4 6 8 10 12 14 16 18 20 22 24
Angle of Attack Angle of Attack
(c) (d)
Figure 28. Aerodynamic Coefficients with varying Angle of Attack (a) Lift (b) Drag (c) Moment (d) Lift to Drag Ratio

4. Computed Flowfield Contours in the Vicinity of the UAV at Various Angles of attack

The computed skin friction lines on the UAV configuration and vorticity contours selected Trefftz planes across
the UAV configuration and in the wake at four different angles of attack, i.e., 50, 100, 150 and 200 and free stream
velocity of 60 m/s are shown Fig. 29. Corresponding to each angle of attack shown adjacently in Fig. 29, a zoomed
view of the vorticity contours on a Trefftz plane cutting the mid-span of the annular wing at that angle of attack is
also shown. Figure 30 shows computed vorticity contours on chordwise planes along the annular wing. These
figures collectively provide an insight on the topology of three-dimensional flow separation from the surfaces of the
UAV and the wake vorticity.

18
American Institute of Aeronautics and Astronautics
(a) 50 AOA
Downloaded by ROKETSAN MISSLES INC. on January 15, 2015 | http://arc.aiaa.org | DOI: 10.2514/6.2015-0771

(b) 100 AOA

(c) 150 AOA

(d) 200 AOA


Figure 29: Computed Skin Friction lines on UAV surface and Vorticity contours on Treftzz planes

19
American Institute of Aeronautics and Astronautics
(a) 50 AOA (b) 100 AOA
Downloaded by ROKETSAN MISSLES INC. on January 15, 2015 | http://arc.aiaa.org | DOI: 10.2514/6.2015-0771

(c) 150 AOA (d) 200 AOA


Figure 30: Computed Vorticity contours on selected planes across the UAV Configuration

Figure 31 shows the computed surface pressure coefficient contours on the UAV surface and streamlines around
the UAV for all the four angles of attack. This figure also shows the interaction of wing downwash with leading
edges of the empennage and also shows the flow separation from the annular wing surfaces as the angle of attack
increases and this can be seen clearly from the skin friction lines on UAV surface shown in Fig. 29. Also, the same
can be observed from the vorticity contours on the mid-plane Trefftz plane across the UAV wing in Fig. 29.
A reduction in lift is observed post stall angle (i.e. 180), as flow separation occurs on both the upper and lower
portions of the wing at 200 angle of attack. Also, the vorticity contours on the Trefftz planes show the strength of
vortices being shed from the trailing edge of the annular wing and distance over which it tends to diminish. The
same can be seen in the computed vorticity contours shown in Fig. 30. In Fig. 31, the CP value of +1 at the leading
edge of the wing defines the stagnation point. The tendency for the stagnation point to shift as the angle of attack
increases and the occurrence of a negative pressure coefficient value is observed at higher angles of attack on the
leading edges of the upper portion of the annular wing resulting in lower values of lift coefficient. Figure 32 shows
the Line Integral Convolution (LIC) contours at the mid-plane of the UAV configuration showing the flow pattern in
the vicinity of the UAV at various angles of attack. The maximum flow velocity is observed near the leading edge of
the upper portion of the annular wing.

20
American Institute of Aeronautics and Astronautics
(a) 50 AOA (b) 100 AOA
Downloaded by ROKETSAN MISSLES INC. on January 15, 2015 | http://arc.aiaa.org | DOI: 10.2514/6.2015-0771

(c) 150 AOA (d) 200 AOA


Figure 31: Computed UAV Surface Pressure Coefficient contours and Streamlines around the UAV

(a) 50 AOA (b) 100 AOA

(c) 150 AOA (d) 200 AOA


Figure 32: Velocity LIC contour at a mid-plane of UAV body

21
American Institute of Aeronautics and Astronautics
IV. Conclusion and Future Work
The present computational study which is in support of recent revival of interest in annular wing applications has
shown a potential for using annular or ring wing concept in unmanned aerial vehicles. The comparison of the
computed aerodynamics of annular wings composed of symmetrical and cambered airfoils has demonstrated
the advantages of using annular wings with cambered airfoils. The effect of aspect ratio of the annular wing on the
computed aerodynamic characteristics has shown expected aerodynamic behavior which is in general agreement
with available experimental data. A brief comparison of the aerodynamic characteristics of the cambered annular
wing and the equivalent rectangular planar wing consisting of the same airfoil has shown the potential advantages of
the annular wing. A possible application of the AW2 annular wing to build a UAV design configuration for which a
static stability analysis has shown the static margin within acceptable limits for static stability and for which
the aerodynamic characteristics have been computed. While the computed aerodynamic characteristics are all in
agreement with expected flow physics, discrepancies observed in moment characteristics in the post stall region has
set the base for further investigation into the aerodynamics of the UAV which will be addressed in future work
which will also include a flight dynamics coupling to define the unsteady aerodynamic characteristics of a propeller
Downloaded by ROKETSAN MISSLES INC. on January 15, 2015 | http://arc.aiaa.org | DOI: 10.2514/6.2015-0771

powered UAV.

Acknowledgments
The authors acknowledge a research grant-in-aid from the High Performance Computing Division of the
Department of Electronics and Information Technology (DEITY), Government of India for setting up the High
Performance Computing Laboratory (HPCLab@IITGN) at IIT Gandhinagar where most of the computational work
is done. The first author also acknowledges a Junior Research Fellowship supported by DIETY.

References
1
Fle , H. S., Experimental Investigation of Lift, Drag and Pitching Moment of Five Annular Airfoil ,
NACA TN 4117, Oct. 1957
2
K , A., G a , P., a O a a , O., D Fa UAV M a S a P a D ,
AIAA Paper 2007-6375 presented at the AIAA Modeling and Simulation Technologies Conference and Exhibit,
Hilton Head, SC, USA, 2007
3
Tra b, L. W., Experimental Investigation of Annular Wing Aerodynamics, Journal of Aircraft, Vol. 46,
No. 3, pp. 988-996, May-June 2009.
4
Prandtl, L., and Tietjens, O. G., Applied Hydro and Aeromechanics, Dover Publications, New York, 1934
5
M , M. M., G a B a T , NACA R . 151, 1923.
6
D a , L., I a C M I Da Closed Wing Systems and C-Wings,
AIAA Journal of Aircraft, Vol. 44, No. 1, pp. 81 99, Jan-Feb 2007
7
R b , H. S., T R A N a a F , Journal of Aeronautical Sciences, Vol. 14, No. 9, pp. 529
530, 1947
8
S a , H. J., T A a aR A , Quarterly of Applied Mathematics 2, Vol. 2, p. 136-141,
July 1944
9
Ma , A., a G , T. H., A a E a A a W Ba L a E S
A a , AIAA Journal, Vol. 51, No. 2, pp. 529-534, 2012.
10
Computational Fluid Dynamics and Multi-physics Engineering Software-STAR-CCM+ Version: 9.04.2014,
User Manual, CD-Adapco, Lebanon, NH, USA. URL (http://www.cd-adapco.com/products/star-ccm)

22
American Institute of Aeronautics and Astronautics
11
Roache, P. J., Verification and Validation in Computational Science and Engineering, Hermosa
Publishers, Albuquerque, NM, USA, pp. 107-136,1998.
12
P a , E. C., A C V L S a -Edge Delta Wings Based on a Leading-Edge-Suction
Analogy , NASATN D-3767, 1966.
13
Tulapurkara, E.G., NPTEL (E-learning courses from IITs and IISc, India) Course Notes, Airplane Design
(Aerodynamic), Chapter 6, Lecture 26,. URL: http://nptel.ac.in/courses/101106035/, Maintained by NPTEL Web
Studio, IIT Madras, Chennai, India, since 1999-2014.
Downloaded by ROKETSAN MISSLES INC. on January 15, 2015 | http://arc.aiaa.org | DOI: 10.2514/6.2015-0771

23
American Institute of Aeronautics and Astronautics

Potrebbero piacerti anche