CFD For Wrap-Around Fins
CFD For Wrap-Around Fins
CFD For Wrap-Around Fins
*Corresponding author
1
Aerospace Engineering Department, Punjab Engineering College
(Deemed to be University),
Sector 12, Chandigarh, 160012-India,
[email protected]*, [email protected]
DOI: 10.13111/2066-8201.2019.11.2.13
Abstract: This paper aims to review computational fluid dynamics (CFD) analysis of wraparound fins
(WAF). An effort in the search of a benchmark computational fluid dynamics model for future studies
has been made in this paper. Iterative studies have been carried out in the past to study the cause and
effect of the anomalies in the aerodynamics of WAF missiles. The simulations that have been carried
out consist of different geometries and different conditions, each providing us a new set of data, results
and complexities. The results of the computations, their setups and the nuances of the simulations have
been presented in this paper. This paper aids in validation of new CFD simulations with the previous
CFD studies and their Experimental counterparts. Benchmarking of geometries, domain setup, grid
generation, etc. will help to obtain a higher quality of the WAF aerodynamics visualization in the future.
Key Words: Computational Fluid Dynamics, Wrap-around Fins, Aerodynamics
1. INTRODUCTION
The use of wraparound fins (WAF) in the aero-industry is becoming more relevant for the tube
launched missiles. Having superiority over the conventional planar fins in terms of storage,
these fins also show signs of reduced drag at higher angle of attacks and higher operational
stability. The desire for obtaining an optimal design and identifying the root and cause of the
discredits in WAF aerodynamics is a major concern of most researchers. These WAF show a
conventional longitudinal aerodynamic characteristic like the planar fins having identical
planform area. The major aerodynamic complication faced by the WAF is the production of
rolling moment at zero angle of attack. The design of dynamically stable WAF demands the
prediction of the roll moment coefficient for the whole flight path. The presence of inherent
rolling motion is attributed to the missile geometry, fin geometry, Mach Number and the wake
condition. It has been found that static aerodynamic characteristics of WAFs are very similar
to a set of planar fins. The computational Fluid Dynamics (CFD) technology is a tested
numerical simulation tool to solve such aerodynamic problems. The analyses have been
performed on 2-D and 3-D geometries and from inviscid to viscous turbulent flows, etc. The
advancements in the hardware and software have reduced the computational times and the
computational costs. The various turbulent CFD models and their attributes have been
discussed in this paper. These CFD numerical analyses are consistent with the experimental
data of the wind tunnel tests as well as the free flight tests, however they lack accuracy. The
dynamic instability is not caused by the unusual roll motion but instead have been reported
INCAS BULLETIN, Volume 11, Issue 2/ 2019, pp. 155 – 170 (P) ISSN 2066-8201, (E) ISSN 2247-4528
Nayhel SHARMA, Rakesh KUMAR 156
due to the presence of a side moment due to pitch, [2]. As the geometry is said to influence the
aerodynamic characteristics of the curved fins, different length to diameter (l/d) ratios have
been tested both experimentally and through CFD. The experimental missiles models
generally were bimetallic bodies (brass nose and aluminium shaft) with steel fins which can
be kept in mind while assigning material to the bodies in CFD. The center of gravity position
has also been changed to simulate the real-time burnout condition of the missile [3].
Mathematical analysis through the missile programs predicts the characteristics of WAF based
on similar projected area as that of planar fins. However, these programs are unable to analyse
the side force and moment aerodynamics of the WAF. A study consisting of supersonics flow
around a single WAF model mounted on a semi cylindrical model, has been carried out both
numerically and experimentally which provided more insights into the WAF aerodynamics,
[4], [5], [6], [7], [8]. This paper discusses the basic aerodynamics of the WAFs, its drawbacks
and the measures taken up in the past by the researchers to overcome them. Focusing on the
computational Fluid Dynamics part this paper discusses the pre- analysis methods, describing
behind the scene mathematical approaches. Along with that the wall modelling strategies,
which are of utmost importance in understanding the near wall behaviour of the fluid flow
have been discussed. Various geometries with their uniqueness have been mentioned as a
careful and accurate modelling of the missiles resulting in better aerodynamic calculations and
for comparison and validation purpose. The next part discusses the handling of domain and
the mesh for the WAFs. The physical setup and the boundary conditions are by and large the
same, the computational step is briefly defined covering all the used turbulent models, their
boundary conditions and the methods used for computing these models. In the end the
validation and verification of the CFD process has been discussed in which the convergence
parameters along with the range of the previously performed CFD analysis have been
discussed. This paper will guide systematically the researcher in each step of the CFD analysis
on WAFs.
spinning missile tests [12]. The wing curvature is found out to influence the yawing moment
whereas the pitching moment, normal force, side force and the axial force remain unaffected
by it. In comparison to the planar fins, the curved fins have significant effect in the yawing
moment, rolling moment but a negligible effect on the normal force, side force and the axial
force [13]. The supersonic flow around a WAF has been characterized near a single fin
mounted on a semi cylindrical model, both numerically and experimentally. These
experiments pave a new method for characterisation of flow field near the wrap-around fin.
Flow
Domain
compressible incompressible
Space Space
Time Viiscous Inviscid Time Viiscous Inviscid
Dynamics Dynamics
where 𝝆𝝆 is the fluid density, V is the fluid velocity vector, 𝝉𝝉𝒊𝒊𝒊𝒊 is the viscous stress tensor, p is
pressure, F is the body forces, e is the internal energy, Q is the heat source term, t is time, 𝜱𝜱
is the dissipation term, and 𝜵𝜵. 𝒒𝒒 is the heat loss by conduction. Fourier’s law for heat transfer
by conduction can be used to describe q as:
𝒒𝒒 = −𝑘𝑘∇𝑇𝑇 (5)
where k is the coefficient of thermal conductivity, and T is the temperature.
The energy equation (Equation 3) is a requirement of the compressible high-speed flow
as it captures the shock pattern effectively. The majority of the WAF missiles flow in the
turbulent conditions. The turbulence modelling involves selecting the near wall modelling
approach and intrinsically defining the inlet and outlet conditions. Reynold number is a
determining factor which defines the flow of the modelling.
𝜌𝜌. 𝑈𝑈. 𝐿𝐿
𝑹𝑹𝑹𝑹𝑳𝑳 = (6)
𝜇𝜇
where L is the length scale and 𝜇𝜇 is the kinematic viscosity. The transition to turbulence varies
depending upon the type of flow;
External flow
• Along the surface : 𝑅𝑅𝑅𝑅𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆 > 50000
• Around on obstacle: 𝑅𝑅𝑅𝑅𝑂𝑂𝑂𝑂𝑂𝑂𝑂𝑂𝑂𝑂𝑂𝑂𝑂𝑂𝑂𝑂 > 20000
Internal flow
• Internal : 𝑅𝑅𝑅𝑅𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖 > 2300
The basic computational approaches for turbulent flows are Direct Numerical Simulation
(DNS), Large Eddy Simulations (LES) and the Reynolds Averaged Navier-Stokes Equation
(RANS). The RANS approach is the most widely used approach in the wraparound missile
simulations, best suited for complex geometries. Some of the early researches have also used
an inviscid code for their simulations; however due to the new advancements and increasing
computational capabilities, finding better possibilities using viscous and turbulent codes are in
trend. The RANS based models are: the Spalart-Allmaras model, the Κ − 𝜀𝜀 family and the
𝜅𝜅 − 𝜔𝜔 family models, the Reynolds Stress Models and the Transition Models (i.e. the 𝜅𝜅 −
𝜅𝜅𝜅𝜅 − 𝜔𝜔, transition shear stress transport (SST) models). These models are placed in order of
cost increase calculated on iteration. The SST 𝜅𝜅 − 𝜔𝜔 and the Realizable Κ − 𝜀𝜀 are the
recommended choices for the standard wraparound simulations. The shear stress transport
(SST) is preferred where the boundary layers are critical and fine distinct resolved heat profiles
are required. The near-wall treatment for the boundary layer profile prediction is done by
making the velocity and the wall distance dimensionless. The velocity is made dimensionless,
by dividing the velocity with shear velocity near the wall of the turbulent flow 𝑼𝑼⁄𝑼𝑼𝝉𝝉 where
𝑻𝑻𝒘𝒘𝒘𝒘𝒘𝒘𝒘𝒘 𝒚𝒚𝑼𝑼𝝉𝝉
𝑼𝑼𝝉𝝉 = � 𝝆𝝆
. The wall distance is made dimensionless 𝒚𝒚+ = 𝝊𝝊
where y is the distance from
the wall and 𝝊𝝊 is the dynamic viscosity of the fluid. A predictable boundary profile is obtained
using these dimensionless quantities. The wall modelling strategies for the near wall treatment
use the wall function approach in which a typical 𝑦𝑦 + value is such that 30 < 𝑦𝑦 + < 300 and
where resolving of viscous sub layer is required 𝑦𝑦 + ≈ 1 is set with the mesh growth rate not
greater than ≈1.2, which is related directly to the inflation layers. In the wraparound fins the
forces and the flow around the wall are important hence the recommended turbulence model
in most of the cases is the SST 𝜅𝜅 − 𝜔𝜔. In various modified methods, SST 𝜅𝜅 − 𝜔𝜔 is used near
the wall region and Κ − 𝜀𝜀 is used away from the walls. Assessment of various turbulent models
have been done in the past to improve the accuracy of the turbulent simulations. These models
are Tam-Ganesan Κ − 𝜀𝜀 model in which modification is done to accommodate heat effects in
the turbulent jets, Tam-Thies Κ − 𝜀𝜀, the standard Chien Κ − 𝜀𝜀 and the Menter SST
formulations; the modified RANS models represent the potential towards the accuracy of the
turbulent methods and an approach towards experimentally observed data mean trends [14].
The basic two equation models fail to accurately predict the onset and amount of
separation in adverse pressure gradient flows. Homogeneous turbulence, small Reynolds
numbers, small pressure gradients etc. sometimes fail to correctly predict the aerodynamic
flows. The 𝜅𝜅 − 𝜔𝜔 model in which the damping functions are not considered, allows simple
Drichlet boundary conditions to be specified. It is unable to predict the asymptotic behaviour
of the turbulence as it approaches the wall. In the 𝜅𝜅 − 𝜔𝜔 model the Κ and 𝜀𝜀 model-based
distribution is inconsistent with the DNS data. The 𝜅𝜅 − 𝜔𝜔 model in the logarithmic region is
superior to the Κ − 𝜀𝜀 model. The logarithmic regions are important in the flows where we have
moderate pressure gradients. The SST model accounts for the transport of the principal
turbulent shear stress which clearly helps in accurately predicting the adverse pressure gradient
conditions [15].
4. GEOMETRY
The Technical Cooperation Program (TTCP) model (Figure 1) has been widely used and
accepted geometrical model for the WAF missile CFD analysis. All the dimensions are in
respect of the radius or the diameters and in uniform calibres. A salient feature of this model
which is adapted in most of the geometries is that the fins are pinched at 45°, both at the leading
and trailing edges. The sharp edges provide a stark advantage over the blunt fins in terms of
aerodynamics and the post processing visualisations. The edge shape has a significant effect
on the roll moment of the wraparound fins. All dimensions are expressed in terms of calibres
(1 calibre = 0.01524m) where as in scaled down models used in Jet Propulsion laboratory,
California Institute of Technology, (JPL) dimensions were again expressed in terms of calibres
(1 calibre = 0.1016m). The Aerial Delivery Research and Development Establishment [1] took
(1 calibre = 0.2m) fins with symmetric pinched leading and trailing edges at 45˚. In the free
flight tests [16] the scaled down TTCP model was adopted (1 calibre = 0.0191m). In the wind
tunnel, water table and free flight tests [17] a rocket with the cavity housing was studied
(1caliber = 0.122m). The symmetry of the body helps in testing only one fourth of the body
with fin inclined at 45° from both the planes [18]. The thickness of the fin and the fin curvature
are important parameters and should be considered in geometry modelling. The dimensions of
the various previously used geometries with their unique features has been summarised in
Table I. Based on 1 calibre = 0.01524m a missile model is shown in Figure 1, with the fin
alone schematic of same calibre dimensions; however, the dimensions have been stated in
mms so that accurate geometries can be designed in future.
Figure 1: A standard model geometry with 1 caliber = 0.01524m, and detailed wraparound fin schematic, the
dimensions are however shown in mm
Table I: Summarised geometry details
Reference Geometry Type Analysis type Calibers features
[1] TTCP CFD 1 Caliber = Both leading and trailing edges
0.2m pinched at 45°
[4] l/d of 20 & l/d of 17, Both Experimental- 1 Caliber = Short brass nose & long brass
cylindrical bodies with Von Free flight tests and 0.0143m nose to simulate center of
karman ogive noses CFD gravity in burnout condition
[5-9] Single Fin test model Both Experimental— 1 Caliber = Single fin, second-order
Wind tunnel and 0.0159m continuity in the longitudinal
CFD direction ensured
[11] US, Airforce institute, Ohio Experimental/ Wind 1 Caliber = slotted
Tunnel 0.0191m
[17] Airforce Armament Experimental/ free 1 Caliber = Slotted
laboratory flight tests 0.0191m
[18] Half scale of 122mm field Experimental/ Wind 1 Caliber = Slender missile body with
rocket Tunnel, Water Table, 0.122m cavity housing
Free Flight data
[19] TTCP CFD 1 Caliber = Root & the Tip chord are
0.1016m parallel
[20] TTCP CFD 1 Caliber = Symmetric leading and trailing
0.01524m edges at 45°
[20] TTCP CFD 1 Caliber = Both leading and trailing edges
0.01524m pinched at 45°
[21] Ogival nose Experimental/ Wind - Canted fins, in non zero roll
Tunnel orientation
[22] ARF Experimental/ free 1 Caliber = Beveled Fins, with swept fins
flight tests 0.01905m
[23] Airforce Research Model Experimental/ free 1 Caliber = Missile body with base cavity
flight tests 0.01905m
[24] 2.1 caliber Von Karman Wind Tunnel - Body with High wing (concave
nose side windward) and Low wing
(convex side windward)
[25] Infinitely long body CFD - A 10% bi-convex airfoil shape,
pinched at the end
Fin aspect ratio of 2
In an interesting case a test model (Figure 2) was designed to represent a single fin of a typical
WAF configuration. The model fin has the same proportions as free-flight models. It has a
total length of 10.92r, where r = 0.0159m is the radius of curvature of the missile fin. Within
this 5r length, a flat sur face is blended to a 1/3-cylinder section in such a way as to ensure
second-order continuity in the longitudinal direction. This semi-cylindrical body has a length
of 5.12r and a maximum height of 0.5r from the tunnel ceiling. The fin has a span of √2r, a
chord length of 1.28r, a thickness of 0.2r. The fin is attached at the aft-most part of the cylinder
section, having the same radius of curvature as cylindrical body. Fin dimensions, as well as a
bevelled leading edge, were chosen to match the free-flight models [5-9]. A schematic which
can be used for future simulations is shown in Figure 3.
(a)
(b)
(c)
Figure 3: Single fin WAF test model (r = 0.0159m) dimensions shown are in mm
Figure 4: Using of symmetric one fourth body and non-reflecting boundary condition can help in reducing the
grid points significantly
Figure 6: Grid boundaries and zonal Structures for single WAF configuration for a single fin based on references
[5-9]
Order
Upwind
Scheme
[1] Pressure Pressure Model Air as Ideal 1bar for - Steady First order
inlet outlet Surface as gas M > 0.2 state up
walls solution winding
Absolute
scheme
Velocity Coupled
used implicit Κ − 𝜀𝜀
formulatio model
n
Green-
involving
Gauss cell
energy
gradient
equation.
method
varying
Reynolds
number
from
27Million
to
57Million
at Mach
1.2 to 2.5
[4] Velocity Velocity Inviscid - - Implicit, Inviscid
inlet outlet flow steady flow
conditions state Euler conditions
flow
solver
[19] - - No Slip, Freestream 101.3 294.75K Explicit Viscous
condition condition kPa time flow
marching
Density: Non-
method
1.198 reflecting
kg/𝑚𝑚3 boundary
conditions
Reynolds
Number:
30-69
million for
sea level &
17-69
million for
wind-
tunnel
[26] - - Wall Inviscid - - Euler -
surface flow based flow
condition conditions solver
[5] Velocity Wall 0.2039 217kPa 294K at Mach Explicit Inviscid
inlet surface kg/𝑚𝑚3 2.9 solver
First order
condition
accuracy
with a full
flux
method
The viscous and inviscid computations show different visualizations and different
aerodynamics. Their roll- reversal Mach numbers also show variations. In case of a pinched
fin missile, a conical shock at the projectile nose followed by an expansion wave and a
relatively weaker shock is formed in front of the fins as compared to those having blunt fins.
[1] To understand the flow between two fins (fins passage) Figure 8. Axial visualizations along
the fins also portray the behaviour of the flow on the missile due to curved fins; this can be
seen in Figure 9 at various fin lengths. Also, the CFD results of the single fin test model and
their comparisons with the experimental results near single WAF at various Mach numbers
are excellent visualizations for the understanding of flow on the curved fin surfaces. An
attempt was made to analyse flow around a single fin both planar and curved, at Mach 2.5; the
results of the pressure contours are shown in Figure 10.
(a)
(b)
Figure 8: Mach number (a) and static pressure variations (b) in the fin passage at different span wise locations for
free stream Mach number 1.4 [1]
A Chord wise
(a) (b)
(c) (d)
Figure 9: Normalised pressure contours at various axial positions on the fin lengths at 2.5 Mach [19]
Figure 10: (A) & (B) Computed Pressure contours at the leading edge of single Wrap around Fin, (C) Computed
Pressure Contours at the trailing edge of the curved Fin, (P) Computed pressure contours around a Single planar
fin
10. CONCLUSIONS
The WAF aerodynamics characterised in the previous studies has been reviewed in this paper
with the focus on the CFD of WAF’s. CFD is an effective tool in understanding the WAF
aerodynamic anomalies. The new computational advancements allow us to use turbulent
models economically. The modification of RANS turbulent methods leads to the corrections
of the turbulent kinetic energy fields, the turbulent viscosity and thus should further be
explored. In representation of the results the pressure contours help significantly in
understanding the aerodynamic characteristics of the WAF model. The pressure on the
concave surface is higher than the pressure on the convex surface which thereby produces the
rolling moment, also termed as the induced rolling moment in some literature. The increasing
Mach number contours show that the shock waves at the nose become stronger and the shock
wave angle becomes smaller gradually [21].
To understand the phenomenon behind this roll reversal, most of the CFD studies have
correlated it with the pressure profiles, the shock structures, along with the increasing Mach
numbers. Static pressure profiles against the Mach numbers have been studied both axially as
well as parallel to the missile body at various chord lengths. The viscous flow over the curved
fin surface opens a new set of flow dynamics which needs to be explored.
The SST based Κ − 𝜔𝜔 model should be a preferred turbulence model for the complex geometry
simulations which shows superior results as compared to the other turbulent models. It is
observed that at lower Mach numbers there is a negative value trend and at the higher Mach
numbers there is a positive value trend for the roll moment coefficients. Many investigations
have used inviscid flow solutions to save time in computations. These inviscid flow solutions
may vary the aerodynamic results, therefore use of turbulent flow conditions is advised for
accurate results. Even in the supersonic regime the aerodynamic characteristics such as the
drag, roll show Mach number dependencies. Flights above Mach 3 show turbulent flow over
the major parts of the missile body, imbedded in the laminar boundary flow; these hinder the
accuracy of the Drag results in the CFD as compared to the flight tests. In this case choice of
Reynolds number again plays an influential role [4], [19]. A multiple-zone strategy for
meshing of WAF has shown results having better agreement with the experimental results.
Applying of non-reflecting boundary layers and using one fourth domain help in reducing the
grid elements and ultimately the computational time with no compromise in the accuracy of
the results. A converging and diverging nozzle analogy after examining the chord-wise
pressure have been suggested [25]. Rolling moment is a function of Mach number and the side
force is linked with the roll angle of the missile. The Single fin semi-cylindrical model
interprets the WAF aerodynamics excellently and show remarkable CFD as well as
experimental results. The single fin semi cylindrical model can be utilized in both inviscid,
viscous & turbulent CFD analysis. This paper can be used as a benchmark for future
wraparound fin, CFD simulations which may help in better understanding of the wraparound
fin aerodynamics.
REFERENCES
[1] A. Seginer, B. Bar-Haim, Aerodynamics of wraparound fins, J. Spacecr. Rockets, 20, 339–345,
doi:10.2514/3.25603, 1983.
[2] E. Air, F. Base, Roll Motion of a Wraparound Fin Configuration at Subsonic and Transonic Mach Numbers,
253–255, doi:doi:10.2514/6.1985-1777, 1986.
[3] G. L. Winchenbach, A. Branch, W. Riner, Aerodynamic test and analysis of a missile configuration with curved
fins, AIAA Paper 92-4495, 1992.
[4] C. P. Tilmann, J. R. Huffman, T. A. Buter, R. D. W. Bowersox, Characterization of the flow structure in the
vicinity of a wrap-around fin at supersonic speeds, 34th Aerospace Sciences Meeting and Exhibit, Reno,
NV, U.S.A., 1996.
[5] J. R. Huffman, C. P. Tilmann, T. A. Buter, R. D. W. Bowersox, Experimental Investigation of the Flow Structure
in the Vicinity of a Single Wrap-Around Fin at Mach 2.9, 1996.
[6] C. P. Tilmann, Numerical and Experimental Investigation of the Flowfield Near a Wrap-Around Fin,
Dissertation, USAF/Wright Laboratory/Flight Dynamics Directorate, AFIT/DS/ENY/97-1.
[7] R. D. W. Tilmann, C. P., Huffman, Jr R. Buter, T. A. Bowersox, Experimental Investigation of the Flow
Structure Near a Single Wraparound Fin, Journal of Spacecraft and Rockets, Vol. 34, No. 6, pp. 729-736,
https://doi.org/10.2514/2.3303, 1997.
[8] C. P. Tilmann, T. A. Buter, R. D. W. Bowersox, Characterization of the Flow field near a Wrap-Around Fin at
Mach 2.8, J. Aircr. 35, 868–875, doi:10.2514/2.2406, 1998.
[9] C. W. Dahike, The Aerodynamic Characteristics of Wrap-Around Fins at Mach Numbers of 0.3 to 3.0, Rep.