0% found this document useful (0 votes)
21 views16 pages

Sheng - Aerodynamic Analysis of A Spinning Missile Using A High-Order Unstructured-Grid Scheme

This document summarizes research on simulating the aerodynamic performance of a spinning missile with dithering canards. Time-accurate simulations were performed using an unstructured grid scheme and dynamic moving grid technique. Both second- and third-order spatial discretization schemes were investigated. The third-order scheme improved prediction of vortices induced by the canards and their interaction with tail fin shocks and boundary layers, leading to more accurate computed forces and moments. Results were compared to experimental data and computations from NASA's OVERFLOW code.

Uploaded by

plutoni249
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
0% found this document useful (0 votes)
21 views16 pages

Sheng - Aerodynamic Analysis of A Spinning Missile Using A High-Order Unstructured-Grid Scheme

This document summarizes research on simulating the aerodynamic performance of a spinning missile with dithering canards. Time-accurate simulations were performed using an unstructured grid scheme and dynamic moving grid technique. Both second- and third-order spatial discretization schemes were investigated. The third-order scheme improved prediction of vortices induced by the canards and their interaction with tail fin shocks and boundary layers, leading to more accurate computed forces and moments. Results were compared to experimental data and computations from NASA's OVERFLOW code.

Uploaded by

plutoni249
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 16

See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/245438791

Aerodynamic Analysis of a Spinning Missile Using a High-Order Unstructured-


Grid Scheme

Article in Journal of Spacecraft and Rockets · January 2010


DOI: 10.2514/1.42988

CITATIONS READS
10 293

3 authors:

Chunhua Sheng Xiao Wang


University of Toledo Mississippi State University
86 PUBLICATIONS 757 CITATIONS 16 PUBLICATIONS 134 CITATIONS

SEE PROFILE SEE PROFILE

Qiuying Zhao
University of Toledo
37 PUBLICATIONS 189 CITATIONS

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Hover Predictions on the S-76 Rotor with Tip Shape Variation Using Helios View project

Fluid/Structure Blast Simulation View project

All content following this page was uploaded by Xiao Wang on 17 May 2016.

The user has requested enhancement of the downloaded file.


47th AIAA Aerospace Sciences Meeting Including The New Horizons Forum and Aerospace Exposition AIAA 2009-1090
5 - 8 January 2009, Orlando, Florida
AIAA-2009-1090
47th AIAA Aerospace Sciences Meeting and Exhibit
Orlando, Florida, 5-8 January 2009

Aerodynamic Analysis of a Spinning Missile with Dithering Canards Using a High


Order Unstructured Grid Scheme
Chunhua Sheng* and Xiao Wang†
Center for Advanced Vehicular Systems
Mississippi State University
Mississippi State, MS 39762, USA
[email protected]

Abstract

Time-accurate viscous flow simulations are carried out for a canard-controlled missile in which the
body spins at a constant roll rate while the canards are dithering. A Reynolds-averaged Navier-Stokes
arbitrary Mach number flow solver based on unstructured grid topology and embedded dynamic moving
grid technique is employed to solve the spinning missile with canard control surfaces. Both second and
third-order spatial discretization schemes are investigated for analyzing the unsteady aerodynamic
characteristics of forces and moments, which are compared with the benchmark solution obtained by
NASA OVERFLOW code. The computed solutions revealed the formation of strong vortices induced by
canards which interact with the tail fin shocks and boundary layers on the body. The use of third-order
discretization scheme noticeably improved the prediction of vortical flows by reducing the excessive
numerical dissipation, and thus improves the accuracy of computed forces and moments.

Introduction

The aerodynamics of a rolling airframe missile is characterized by multiple shocks and highly unsteady
and vortical flows, which interact with the tail fins and body boundary layers. Numerical simulation of a
rolling airframe missile with dithering canards poses significant challenges for any computational
approaches. Not only does the solver need to accurately predict the location and strength of shocks and
vortices generated by the missile control surfaces and sharp edges, but also needs to capture the fast
dynamic response of the missile motion associated with the dithering canards. In the past several years, a
number of researchers have attempted to use different approaches in predicting and analyzing the
aerodynamic performance of this missile configuration. Nygaard and Meakin [1,2] used an overset
structured grid approach to perform the Euler and Navier-Stokes computations with a series of refined
meshes, and assessed viscous effects on the aerodynamic forces. Their results showed that a finer mesh
would increase the drag prediction and had a great effect on the aerodynamic loading prediction. Blade and
Marcum [3,4] performed similar calculations using a second-order unstructured grid scheme for
compressible flows to predict the aerodynamic performance of the spinning missile. They used a sliding
interface [5] method to account for the motion of dithering canards, and obtained the aerodynamic forces
and moments that are compatible with the OVERFLOW results [2]. Later, Sheng, Wang, Hughson, and
Marcum [6] used an arbitrary Mach number flow solver to predict the rolling missile configuration with
fixed canards. This particular version of the U2NCLE solver [7] has showed a superior stability and
robustness comparing to its compressible flow version counterpart [3]. An extensive aerodynamic
performance database has been generated for the spinning missile with deflected canards (0, +-15, and +-30
deg.) in a range of Mach numbers (0.8 to 2.2) and angles of attack (0 to 12 deg.) [6], a first time that such a
large aerodynamic performance database was solely derived from the unsteady viscous simulations in the
missile trajectory prediction.
In the current work, a rolling airframe missile with dithering canards is numerically analyzed using
both second and third-order spatial discretization schemes of the arbitrary Mach number U2NCLE solver
[7,8]. A dynamic moving grid technique [9] was adopted to handle the relative motion between the
dithering canards and rolling airframe. It should be noted that unsteady simulations of the spinning missile
with dithering canards sweeping between two extremes still remain a technical challenge even for the
conventional second-order scheme, not to mention the higher-order scheme due to its less numerical
_____________________________________________
*
Associate Research Professor, Senior Member AIAA

Assistant Research Professor

Copyright © 2009 by the American Institute of Aeronautics and Astronautics, Inc. All rights reserved.
dissipation. A novel and efficient third-order inviscid flux reconstruction combined with a limiter detector
technique [10] is introduced for the improved prediction of vortical flows induced by the dithering canards
at a supersonic Mach number. Simulation results are compared between the second and third-order schemes
for the aerodynamic performance of forces and moments, induced vortex convection, and interaction of
vortex with the boundary layer. Predicted roll-averaged forces and moments are compared against the
experimental data and benchmark results obtained by NASA OVERFLOW code [1,2].

Missile Configuration and Computational Mesh

The missile configuration [11] considered in this study contains a hemispherical nose, cylindrical mid-
body, and a boattail forward of an additional cylindrical aftbody, two dithering canards located aft of the
nose, and four fixed tail fins located behind the boattail on the most aft cylindrical section, see Fig. 1. The
tail fins are slightly canted to impart a roll to the airframe. The two canards act as control surfaces and are
interconnected thus their motion is synchronous. In the simulations of spinning body/dithering canards, the
canards sweep between the extremes. The canard actuator flips back and forth between the two extremes
over a missile roll revolution. The missile spins at a constant roll rate. To model the complex motions of the
spinning missile and the dithering canards, an embedded dynamic moving grid technique [9] with nested
parent and child volumes are used. In this approach, the absolute motion of the canards is considered as a
combination of the dithering motion relative to the missile body and the spinning motion of the missile.
Therefore, a child moving volume grid is created around the canards as shown in Fig. 1 to model the canard
dithering, while a parent moving grid is constructed around the entire missile to handle the spinning motion.
The mixed–element type unstructured grid was generated suitable for viscous simulations. The volume grid
is constructed using the Advancing–Front Local Reconnection (AFLR) grid generation technique [12].
SolidMesh [13] was used to prepare the CAD geometry and generate the surface grids. For the viscous grid,
the normal spacing of the first grid point from the solid surface is 1.0×10-5, which leads to a y+ distribution
of less than 1.0 over all the solid surfaces, thus assures good viscous sublayer resolution. The grid
generated in this work has 9.26 million nodes, and 34.53 million mixed elements. The outer farfield
boundary of the computational domain is a sphere with a diameter of 3.67 body length of the missile.

Canards Tail fins

Missile Body

(a) Missile geometry with tail fins and movable canards (b) Missile spinning volumes

Fig. 1. Missile Geometry and Computational Model

Numerical Methods

The governing equations and numerical methods for the arbitrary Mach number solution algorithm are
described first with a brief introduction of the third-order scheme development. Details of the high order
numerical approach will be published in a separate paper.

Primitive Variable Formulation

2
The unsteady three-dimensional compressible Reynolds-averaged Navier-Stokes equations are cast in a
Cartesian coordinate system, and are written in a finite-volume form. It represents a system of conservation
laws for a control volume that relates the rate of change of a vector of average state variables to the flux
through the face of a control volume. The conservative flux formulation is written in terms of primitive
variables for preconditioning purpose in order to compute flows at arbitrary Mach numbers. The primitive
variable formulation of the governing equation, after introducing a preconditioning matrix Γq−1 , can be
expressed as:
∂ r
MΓ q−1
∂t
∫Ω qdV + ∫ Ω F ⋅ ndA = 0

(1)

where Γ q−1 is a constant diagonal preconditioning matrix that only depends on the global reference Mach
number Mr
Γq−1 = diag [1, 1, 1, 1, 1 β ( M r )]
β ( M r ) = M rλ λ ∈ [0,2] (2)
and M is the transformation matrix from conservative variables Q=(ρ, ρu, ρv, ρw, ρet) to primitive
variables q =(ρ, u, v, w, p). F is the vector of the flux on the face of the control volume. Since the
governing equations are solved in the fixed frame of reference where the missile configuration is rolling,
there is no source term (Coriolis and centrifugal forces) being added into the above equation.

Inviscid Flux Approximation

Recall from Ref. [14] that preconditioned equations (1) can be expressed in a differential form as

∂q ∂q
MΓ q−1 + MΓ q−1 Γq M−1 AM =0 (3)
∂t ∂x
where Γ q M −1 AM is the system matrix for the preconditioned equations, and A = ∂F / ∂Q is the flux
Jacobian matrix with respect to Q. The inviscid flux approximation for the preconditioned system is
F = (F(qL ) + F(q R ))- MΓq−1 Γq M−1 AM (q R − q L )= 0
1 1 (4)
2 2
The eigensystem of the above Eq. (4) is evaluated based on the averaged variables instead of Roe variables,
which is considered to be an extension of the Roe flux approximation.
In Eq. (4) quantities qL and qR are values of the primitive variables on the left and right side of the face
of the control volume. For the first-order accurate differencing scheme, quantities qL and qR are set equal to
the values at the nodes lying on either side of the face. For the conventional second-order scheme, these
values at the face are computed with a Taylor series expansion about the central node of the control volume
r
q face = q node + ∇q node ⋅ r (5)
r
where r is the vector that extends from the central node to the face of the control volume, and ∇q is the
first derivatives of the primitive variables at the node that are evaluated with an unweighted least-squares
procedure [15].
The above second-order scheme has been used in the previous study for generating the aerodynamic
performance database [6]. For the current study of the spinning missile with dithering canards, it is used to
provide a baseline solution for comparison with the higher order discretization scheme, which will be
described next.

Third-Order Inviscid Flux Reconstruction

There is an ongoing effort in the CFD community to develop higher order discretization schemes [16-
21] because of their less numerical dissipation than the conventional second-order schemes. The use of
higher order scheme may be crucial for certain aerodynamic applications such as helicopter rotors and
missile systems with control surfaces, where the vortices induced by the helicopter rotor and missile control
surface have significant impact on the aerodynamic performance. The higher order numerical schemes may
better preserve the path and strength of vortices in the field on moderate computational grids, avoiding the
need for expensive grid adaptation and/or local grid refinement in the vicinity of vortex cores. Although
notable successes have been achieved primarily in the structured grid-based algorithms, development of an

3
efficient and robust high order discretization scheme based on unstructured grids is still a technical
challenge. This is because most high order unstructured grid schemes generally encounter a large stencil
(more computing costs), such as a finite spectral volume method [20] or discontinuous Galerkin method
[17]. Another major obstacle for the high order unstructured grid schemes is a lack of efficiency and
robustness to apply them to practical and complex aerodynamic problems, such as the case of spinning
missile with dithering canards presented in this work. It is very difficult to balance the requirements of
numerical accuracy and stability of a higher order scheme when the flow field contains both strong vortices
and shocks, as these are often two conflicting attributes. In the present study, an efficient and robust higher
order spatial discretization scheme is introduced. It uses an upwind-biased high order polynomial for the
flux reconstruction combined with a smart filter of limiter detector technique [10].
Recall in formula (5) that a second-order flux reconstruction scheme can be achieved by using a linear
representation of flow variables within the control volume. In the present study, the following quadratic
polynomial is used to replace the linear approximation (5) in order to develop a third-order discretization
scheme:
r r
q face = q node + ∇q node ⋅ r + f ( q node,∇q node )r 2 (6)
where the extra term f in Eq. (6) is to account for the high order polynomial, which is constructed based on
an upwind-biased formula using solution variables and gradients on both side of the face of control volume.
In the smooth region of the flow field, the above upwind-biased flux reconstruction provides improved
simulation accuracy while maintaining the stability of the scheme. For flows with strong discontinuity such
as shocks, the underlying numerical dissipation of Eq. (6) may not be enough to keep the solution stable,
and the use of a limiter is necessary. In the present study, a smart filter with detector technique [10] is
investigated and used to maintain the high order accuracy in the smooth region while providing enough
numerical dissipation near the flow discontinuity. This method has been successfully used in the current
simulation of the spinning missile with dithering canards, where both smooth region (vortical flow) and
discontinuous region (shocks) exist in the flow field.

Viscous Flux Evaluation

The viscous flux discretization is still second order in the present work, as it provides secondary
contribution to the overall numerical accuracy of the scheme. The current unstructured viscous grids are
composed of mixed types of elements including tetrahedron, quadrahedron, prism, and hexahedron. There
are two key issues in reconstructing the viscous flux. One is the positivity of the discrete operator, which
requires that all of the stencil weights be positive. Positivity is a key property for numerical stability.
Another key property is the linearity-preserving of gradients (for second-order scheme). In this work, a
positive scheme proposed in [7] is used to compute gradients of flow variables at the medial dual faces for
stable and accurate simulations of viscous flows.

Turbulence Model

Several turbulence closure models have been developed in the solver for high-Reynolds number flows.
One of them is the Spalart-Allmaras one-equation turbulence model [22], which formulates a transport
equation for the turbulence Reynolds number, and is the only one used in the present study due to its
simplicity. From the original Spalart-Allmaras formulation [22], a transport equation for the turbulence
~
Reynolds number ν (working variable) is expressed as
∂ν~ r ~ cb1 ⎤ ⎡ν~ ⎤
2
1 ⎡
+V ⋅ ∇ν = cb1 ( f r1 − f t 2 )S ν~ −
~
c f − f t2 ⎥⎢ ⎥ +
Re ⎢⎣
w1 w
∂t k2 ⎦⎣ d ⎦
1
{∇ ⋅ [(ν~ + (1 + cb 2 )ν~ )∇ν~ ]− cb 2ν~∇ ⋅ (∇ν~ )} (7)
σ Re
where the first and second terms on the right hand side of Eq. (7) are the source production, and the third
term is the turbulence dissipation. Details of the coefficients in Eq. (7) can be found in Ref. [22].

Iterative Procedures

The baseline flow solver is a node-based, finite volume implicit scheme using unstructured mixed
element grids. The time-advancement algorithm is based on the implicit Newton’s method, which yields a

4
linear system of equations for the solution at each time step. The solution of the sparse system of equations
is obtained by a symmetric Gauss-Seidel relaxation scheme. Normally 6-8 symmetric Gauss-Seidel
realizations are adequate at each time step. For unsteady time-accurate simulations, 4 Newton subiterations
are normally used to maintain the stability and temporal accuracy of the unsteady solutions.

Results

Unsteady Navier-Stokes simulations were performed to investigate the complex vortical flow field and
aerodynamic forces and moments for the spinning missile with dithering canards, using the second and the
third-order schemes. The missile roll rate is fixed at 9 Hz in the current study. The free stream Mach
number is 1.6, the Reynold’s number is 41.327×106, and the flow angle of attack is 3 degrees. The canard
dithering is at 0% command level and the time required for the canards to switch between the positive and
negative extremes is approximately 20 deg. roll angle of the body [11].
Computations of the spinning missile with dithering canards were carried out using 64 parallel
processors on a Sun X2200 M2 cluster (2048 compute cores, 2.6 GHz Opteron 2218, 4 Terabytes of RAM)
located at the High Performance Computing Collaboratory of Mississippi State University. The inter-node
communication is through the message passing interface (OpenMPI) operating on SUSE8 Linux operating
system. A converged unsteady solution was normally obtained in about 18 hours of CPU runtime for 3
missile revolutions.

Convergence History

The unsteady time-accurate simulation of the spinning missile was initiated from a uniform flow using
second-order temporal accuracy, and second and third-order spatial discretization schemes as described
before. To ensure the temporal accuracy of the unsteady solution, 6 symmetric Gauss-Seidel relaxations
and 4 Newton sub-iterations were used per time step. A minimum time step corresponding to one deg. of
roll angle per time step was used, which is significantly larger comparing to other methods [1,2] due to the
use of Newton implicit scheme. This minimum time step seems to provide a balanced need for both
temporal resolution and computational efficiency. Since the free stream flow is at supersonic Mach number,
the formation of shock waves and convection of vortices take place right quickly. The predicted
aerodynamic forces and moments were found fully converged in an half roll revolution after the simulation
was initiated. Figure 2 shows the time history of the axial and normal forces using the second-order scheme.

Fig. 2 Time History of Aerodynamic Forces of the Spinning Missile with Dithering Canards

The third-order flux approximation scheme was restarted from the converged second-order solution
using the same solver input parameters. It was found that after only one third roll revolution (120 deg.) the
third-order solution was fully converged, although the simulation has been carried out for another full 3
body revolutions. The aforementioned higher order numerical flux approximation is proven to be rather
successful in predicting such a complicated and challenging aerodynamic problem. It provides improved

5
numerical accuracy in both smooth and discontinuous regions, which noticeably affected the aerodynamic
forces and moments acting on the missile body.

Unsteady Forces and Moments

One goal of the present study is to investigate the unsteady aerodynamic forces and moments on the
spinning missile with dithering canards. The dynamic pressure, the cross-sectional area of the missile body
and the body length are used to normalize the force and moment coefficients presented here:
Force Moment (8)
C =F C = M
Q∞ S ref Q∞ S ref Lref
where Q = 1 ρ U 2 is the free stream dynamic pressure, Lref is maximum body diameter, and
∞ ∞ ∞
Sref is the cross
2
sectional area at maximum body diameter.
Figure 3 (a)-(f) show the predicted force and moment coefficients using the second and the third-order
discretization methods, along with the comparison with OVERFLOW results [1]. Note that all force and
moment coefficients are plotted on a relative scale, which have been normalized by their maximum values.
All simulation results were based on the same flow condition and canard moving schemes. OVERFLOW
code uses a Chimera overset structured grid methodology. The OVERFLOW results were generated based
on an extremely fine spatial and temporal resolution using 41 million grid points and 12,000 time steps per
revolution, vs. 9.26 million grid points and 360 time steps per revolution for the current simulations.
It is seen from Fig. 3 that major aerodynamic characteristics associated with the dithering canards have
been captured in the simulations. The force and moment coefficients predicted by the third-order scheme
are very close to the second-order results, except for the axial force and rolling moment coefficients. In fact
the largest difference also appears in these two quantities when compared with the OVERFLOW results.
As to be discussed in the following, both axial force and rolling moment are mostly affected by the canard
deflected angle, where the vortices induced by canards may have a strong impact on the boundary layer
development on the missile body. This implies that the axial force and rolling moment are most sensitive to
the numerical dissipation of the scheme, since the higher-order discretization method may better preserve
the strength and path of the canard-induced vortices. Another major factor is owing to the grid resolution.
Previous grid refinement study performed by Nygaard, et al. [1] and Blade, et al. [3] about the same missile
configuration showed that the axial force is also affected by the grid resolution. On a relative basis, the
maximum difference in the axial force coefficients between the current simulation (3rd-order) and the
OVERFLOW result is 2.46%. Comparing to the second-order simulation, results obtained by the third-
order solution seems to be closer to the OVERFLOW results.
Table 1 summaries the roll-averaged mean values of the force and moment coefficients obtained in the
current simulations, and the comparison with the experimental data [11,2]. Note that the reference point
used for all moment computations in this table matches with the experimental condition, which is different
from the moment calculations based on the center of gravity shown in Fig. 3. It is seen that predicted axial
force, normal force, and pitching moment coefficients by the current simulations match well with the
experimental data. In particular for the roll-averaged axial force coefficient, which presents the drag acting
on the missile body, the difference between the current simulation (3rd order) and the measured value is
1.17%. The difference in the side force and yawing moment coefficients between the computation and the
experiment may be due to a varying experimental canard pitching sequence, which is different from the
simulation at 0% command level.

Axial Normal Side Rolling Pitching Yawing


force coef. force coef. Force Coef. moment moment moment
(Cx) (Cy) (Cz) coef. (Cmx) coef. (Cmy) Coef. (Cmz)
2nd-order 1.195 0.475 0.00366 -0.01009 -1.226 -0.000144
3rd-order 1.193 0.468 0.00273 -0.00935 -1.236 -0.005513
Experiment* 1.155–1.176 0.45–0.61 0.15–0.20 0.019–0.036 -1.5 – -0.40 -1.5 – -0.93

Table 1. Comparison of Roll-Averaged Force and Moment Coefficients


*
Experimental data were averaged over ten body revolutions during a varying canard pitching sequence

6
(a) Axial force coefficient

(b) Normal force coefficient

(c) Side force coefficient

7
(d) Rolling moment coefficient

(e) Pitching moment coefficient

(f) Yawing moment coefficient

Fig. 3 Comparison of Predicted Normalized Forces and Moments with OVERFLOW

8
Pressure and Shock Structures

The supersonic flow around the spinning missile is characterized by a bow shock in front of the missile
nose, and multiple oblique shocks and expansive waves generated by canards, boattails, and body fins.
These shocks directly interact with the body boundary layers, and most importantly, the vortices induced by
the canards. Figure 4 shows an overall flow structure of pressure contours for this missile configuration
obtained by the third-order scheme, which illustrates various shocks and expansive waves. There is no
doubt that numerical simulation of this unsteady flow field poses significant challenge, in particular to the
higher-order methods. As described before, the third-order flux discretization method presented in this
work is intended to capture the strong discontinuity (shocks) while preserve the higher-order accuracy in
the smooth region. Figure 5 compares the pressure contours near the missile nose region obtained by both
second and third-order methods. There is only slight improvement to the predicted bow shock thickness
using the current third-order flux formula comparing to the second-order method, which is expected since
the use of the smart limiter in this work is intended to maintain the numerical stability at discontinuity and
only improve the numerical accuracy in the smooth region. In general, the pressure contours obtained by
the third-order method shows more details comparing to the second-order results.

Fig. 4 Predicted Instantaneous Pressure on Missile Surface and x-y Plane

(2nd Order) (3rd Order)

Fig. 5 Predicted Bow Shocks and Pressure Contours on the Nose and Canard

9
Canard Vortex Convection

In addition to the shock waves discussed above, another dominant flow feature for this spinning
missile is the canard-induced vortices, which introduce very complicated and unsteady interactions with
shocks, expansion waves, and body boundary layers. Each canard induces two counter-rotating vortices, the
one at the tip of the canard called outboard vortex, and the one at the root of the canard called inboard
vortex [1]. Since the inboard vortex is near the missile body, its development and convection to
downstream has a significant impact on the missile aerodynamic performance. This is the reason that
accurate prediction of the vortices generated in the flow field is crucial in the simulation and analysis of the
missile performance and trajectory. However, most conventional Eulerian-based numerical methods suffer
excessive numerical dissipation, resulting in quick smearing of the vortex strength predicted in the flow
field. The higher order flux scheme developed in the present study is intended to reduce the excessive
numerical dissipation to better preserve the vortex strength and convection in the flow field. It avoids the
need for grid adaptation and/or local grid refinement, which is very expensive and difficult to implement
for large-scale unsteady and parallel computations, in particular on unstructured grids owing to the dynamic
memory allocation and load balancing issues.
In the context of fluid dynamics, a measure to the strength of vortices is called helicity, which defines
the helical motion of the fluid. When a parcel of fluid is undergoing a rotating motion about an axis parallel
to the direction of motion, it will create helicity. Helicity is simply the dot product of the velocity and the
vorticity vector:
H = u ⋅ (∇ × u )d 3 r
∫ (9)
where u is the local flow velocity and r is the local distance for the volume integration. An advantage of
using the helicity vs. other quantities (such as vorticity) to present vortical flows is that it shows not only
the strength of vortices but also the direction of the vortex rotation. As shown in the following Fig. 6, the
red color indicates a positive helicity, which is rotating clockwise when viewed from the missile nose to
tail. The blue color indicates negative helicity whose rotation is counter-clockwise. The missile body is also
rotating clockwise when viewed from the nose to tail.
Figure 6 shows the instantaneous helicities at three different missile roll angles (120, 240, and 360 deg.)
predicted by the second and third-order schemes, where the 360 deg. roll angle represents the missile initial
position with two canards positioned horizontally and the free stream approaching at 3 deg. angle of
incidence. These helicity contours are calculated at various cutting planes in the free stream direction
starting right behind the canard surfaces up to the end of the missile body. The figure reveals a dynamic
vortical flow motion associated with the spinning missile. Both inboard and outboard canard-vortices move
along with the dithering canards and rotate with the spinning missile. It is seen that canard-induced vortices
remain strong during most of the missile roll motion. The weakest vortices occur as the canard flaps its
pitch angle from positive to negative, or vice versa, in which the induced vortices change the direction of
rotation. At all three missile orientations, the solution predicted by the third-order method shows less
numerical dissipation and better vortex convection to the downstream comparing to the original second-
order solution. Figure 7 shows predicted induced velocity within the vortex core along the missile body.
The inboard vortices are decayed faster than the outboard vortices because of their interference with the
missile body. However, predicted velocities using the third-order approach show less numerical diffusion
within the vortex core along the missile body comparing to the second-order results, in particular for the
outboard vortices (Fig. 7 (b)). This clearly demonstrates the efficacy and improvement of the new third-
order method in predicting the complex, viscous, and vortex-dominant flows, even in the presence of strong
shocks. The successful simulation of this flow field involving both strong vorticies and multiple shocks all
together implies the applicability and robustness of this new method in predicting other complex and
challenging aerodynamic flow problems. In the following section, a refined flow detail is revealed using the
third-order approach, which provides an improved prediction and understanding of the vortex-boundary
layer interaction for the spinning missile.

10
(a) Roll angle = 120 deg. (2nd order) (b) Roll angle = 120 deg. (3rd order)

(c) Roll angle = 240 deg. (2nd order) (d) Roll angle = 240 deg. (3rd order)

(e) Roll angle = 360 deg. (2nd order) (f) Roll angle = 360 deg. (3rd order)

Fig. 6 Canard-Induced Vortex Structures at 120, 240, and 360 deg. Roll Angles

11
0.9
0.8
0.7
0.6
0.5
0.4
0.3
X=0.2

(a) Inboard vortex

0.9
0.8
0.7
0.6
0.5
0.4
0.3
X=0.2

(b) Outboard vortex

Fig. 7 Predicted Induced Velocity in Vortex Cores by 2nd and 3rd Order Schemes

Vortex and Boundary Layer Interaction

As discussed above, the canard induced vortices strongly interact with the missile airframe surface and
tail fins, and inboard vortex directly interacts with the body boundary layers. The use of third-order flux
scheme not only improves the prediction of the vortex strength and convection, but also helps capture the
fine details of the flow field that was omitted or dissipated in the second-order simulation results. Figure 8
shows the instantaneous cross-flow velocity (velocity in y-z plane) contours at a section of x=0.2 measured
from the nose of the missile body, or a 0.05 body distance behind the canard trailing edge. The missile
airframe orientation is at 360 deg. roll angle. It is seen that both inboard and outboard vortices induced by
each canard have been predicted by the second and third-order simulation results. However, the third-order
result reveals stronger induced velocities by canard-vortices and thus stronger vortex-boundary layer
interaction near the body surface, as shown in Fig. 9. This indicates that accurate prediction of vortical
flows is very important for the prediction of aerodynamic forces and moments of the missile system, as
these canard-induced vortices may penetrate into the boundary layers and change the velocity profile. As

12
already demonstrated in this study, the third-order simulation has noticeably improved the prediction of the
axial force (drag), because it is more sensitive to the velocity profile change in the boundary layer. It is
expected that more refined flow details would be captured when an even higher order discretization method
is used.

(2nd order) (3rd order)

Fig. 8 Cross-Flow Velocity Contours at Section x=0.2

Weaker Stronger
induced induced
velocity velocity

(2nd order) (3rd order)

Fig. 9 Cross-Flow Velocity Vectors at Section x=0.2

Conclusions

Unsteady viscous simulations of a spinning missile with dithering canards were carried using second
and third-order flux reconstruction schemes. A novel approach is developed and presented for a third-order
inviscid flux formulation that provides improved numerical accuracy in the smooth region while maintain
the stability in the discontinuity such as shocks. This new method has been successfully applied and
demonstrated in the simulation of a practical and challenging supersonic flow that involves fast missile
motion with dithering canards, strong vortex convection, multiple shocks and expansive waves, and their
interacting among vortices, shocks, and boundary layers. The following concluding remarks are given in
the present study:

13
1. A new flux reconstruction method has been demonstrated for improved numerical accuracy and
less dissipation in the smooth region while maintaining the stability of the solver at discontinuities,
a key attribute to be applied to complex and challenging aerodynamic problems.
2. Noticeable improvement of the axial force has been obtained by the third-order flux reconstruction
method as compared with the original second-order scheme, owing to the better preservation of
canard-induced vortices using the new technique.
3. Each dithering canard introduced a pair of inboard and outboard vortices, which move along with
the canard and change the direction of rotation from back and forth as the dithering canard flaps
the pitch angle. These canard-induced vortices have a strong impact on the aerodynamic
performance of the spinning missile.
4. Improved prediction of induced vortices, shocks, and their interaction with the boundary layer
using the high order approach helps better understand the complex aerodynamic characteristics
associated with the spinning missile with dithering canards.

References

1. Nygaard, T.A. and Meakin, R.L., “An Aerodynamic Analysis of a Spinning Missile with Dithering
Canards,” AIAA-2002-2799, 20th AIAA Applied Aerodynamics Conference, June 24-26, 2002, St.
Louis, MO.
2. Nygaard, T.A. and Meakin, R.L., “Aerodynamic Analysis of a Spinning Missile with Dithering
Canards,” Journal of Spacecraft and Rockets, Vol. 41. No. 5. September-October 2004.
3. Blades, E.L., Marcum, D.L., and Mitchell, B., “Simulation of Spinning Missile Flow Fields using
U2NCLE,” AIAA-2002-2797, 20th AIAA Applied Aerodynamics Conference, St. Louis, MO, 24-26 Jun.
2002.
4. Blades E.L., and Marcum D.L., “Numerical Simulation of A Spinning Missile with Dithering Canards
Using Unstructured Grids,” Journal of Spacecraft and Rockets, Vol. 41, (2), pp. 248-256, March-April
2004.
5. Blades, E. L., and Marcum, D. L., "A Sliding Interface Method for Unsteady Unstructured Flow
Simulations," International Journal for Numerical Methods in Fluids, Vol. 53, No. 3, pp. 507-529,
2007.
6. Sheng, C., Wang, X., Hughson, M., and Marcum, D., “Unsteady Navier-Stokes Simulations of a
Canard Controlled Missile Configuration,” AIAA-2008-7324, 26th AIAA Applied Aerodynamics
Conference, 18-21 August, 2008, Honolulu, Hawaii.
7. Sheng, C., Newman III, J., Remotigue, M., Chen, J.P., Marcum, D, and Whitfield, D., “Development
of Unstructured Computational Capabilities Applicable to MSU TURBO with an Arbitrary Mach
Number Algorithm,” Final Report for NASA Grant NAG 3-1699, MSSU-COE-ERC-02-16,
Mississippi State University, October 2002.
8. Sheng, C., and Wang, X., “Characteristic Variable Boundary Conditions for Arbitrary Mach Number
Algorithm in Rotating Frame,” AIAA-2003-3976, Proceedings of the 16th AIAA Computational Fluid
Dynamics Conference, June 23-26, 2003, Orlando, FL.
9. Sheng, C., and Narramore, J., “Unsteady Simulations of Bell-Agusta 609 Rotor with Higher Harmonic
Oscillation,” Journal of Aircraft, Vol. 25, No. 3, May-June, 2008.
10. Amir Nejat, Carl Ollivier-Gooch, “A high-Orde Accurate Unstructured Finite Volume Newton-Krylov
Algorithm of Inviscid Compressible Flows,” Journal of Computational Physics, 227 (2008) 2582-2609.
11. “Defensive Missile Wind Tunnel Test for the Validation and Verification of CFD codes,” Dynetics TR
PO 40725, Huntsville, AL. Jan. 2002.
12. Marcum, D.L. and Gaither, J.A., “Mixed Element Type Unstructured Grid Generation for Viscous
Flow,” 14th AIAA Computational Fluid Dynamics Conference, AIAA Paper 99-325A, Norfolk, VA.
13. Gaither, J.A., Marcum, D.L., and Mitchell, B., “SolidMesh: A Solid Modeling Approach to
Unstructured Grid Generation,” 7th International conference on Numerical Grid Generation in
Computational Field Simulations, Whistler, BC, Sept. 2000.
14. Briley, W.R., Tayler, L.K., and Whitfield, D.L., “High–Resolution Viscous Flow Simulations at
Arbitrary Mach Number,” Journal of Computational Physics, 184(1), 79-105, 2003.
15. Anderson, W.K., Rausch, R.D., and Bonhaus, D.L., “Implicit/Multigrid Algorithms for Incompressible
Turbulent Flows on Unstructured Grids,” AIAA–95–1740, 1995.

14
16. Mahesh, K., “A Family of High Order Finite Difference Schemes with Good Spectral Resolution,”
Journal of Computational Physics 145, 332-358 (1998).
17. Cockburn, B., and Shu, C., “The Runge-Kutta Discontinuous Galerkin Method for Conservation Laws
V: Multidimensional System,” Journal of Computational Physics, Vol. 141, pp. 199-224, 1998.
18. Tang, T. and Li M., “A Compact Fourth-Order Finite Difference Scheme for Unsteady Viscous
Incompressible Flows,” Journal of Scientific Computing, Vol. 16, No. 1, March 2001.
19. Jiang, G. and Shu, C., “Efficient Implementation of Weighted ENO Schemes,” Journal of
Computational Physics 126, 202-228 (1996).
20. Wang, Z. and Zhang, L., “High-Order Spectral Volume Method for 2D Euler Equations,” 16th AIAA
Computational Fluid Dynamics Conference, 23-26 June 2003, Orlando, Florida.
21. Luo, H., Baum, J., and Lohner, R., “A Hermite WENO-based Limiter for Discontinuous Galerkin
Method on Unstructured Grids,” 45 AIAA Aerospace Sciences Meeting and Exhibit, 8-11 January
2007, Reno, Nevada.
22. Spalart P., and Allmaras, S., “A One-equation Turbulence Model for Aerodynamic Flows,” AIAA
Paper 92-0439, January 1991.

15

View publication stats

You might also like