Applications
- Missile Aerodynamics
- Transonic Aerodynamics
- Unsteady Aerodynamics
Models active in this Demonstration Case
- Compressible flow
- Turbulence models
- Energy models
- Ideal gasses
- Double Precision
Objective
To profile the accuracy and performance of Envenio’s EXN/Aero manycore CFD solver on a challenging missile aerodynamics test case.
Case Description
A generic missile geometry is considered in the present study with the canards at 15^{o} angle of attack. The geometry of the missile configuration is shown in Fig. 1. The dimensions of the missile are as follows:
Length 1.4 m
Missile body diameter 0.07 m
Tail body diameter 0.04 m
Canard length 0.03 m
Canard span 0.04 m
Canard thickness 0.003 m
Angle of attack for the canards 15^{o}
Tail fin length 0.05 m
Tail fin height 0.05 m
Tail fin max thickness 0.005 m
The free flow conditions employed in the present study correspond to sea level conditions and to a free stream Mach number of 0.5.
The Reynolds number with respect to the missile length is 2849872. No shock waves are developed around the missile.
Figure 1: Missile geometry.
Figure 2: Tail fins geometry.
Mesh Description
The surface mesh is shown in the following figures where it is seen that a mixed element type mesh has been generated. Specifically, the cylindrical body of the missile was meshed with triangular elements while the canards and tail fins were meshed with quadrilateral elements. The mesh was refined at the missile nose, where the flow stagnates, in order to resolve the high local flow gradients.
(a)
(b)
(c)
Figure 3. Surface mesh
The volume mesh is depicted in Fig. 4 for the configuration with the canards at 15^{o} AoA. The volume mesh is comprised of pentas, hexahedral and tetrahedral elements and has about 2.7 million elements. The distribution is as follows:
(a)
(b)
Figure 4. Hybrid volume mesh
Simulation Setup
Solver Control for the canards at 15^{ o} AoA
Steady state and Transient simulation
For the steady state we start with a CFL of 2 and progressively increased it to 50000.
For the transient simulation a constant dt of 1.0e-5 sec
Convergence criteria: 1.0e-4 RMS and 1e-2 Max
Coefficient loops 10
Advection scheme: 1^{st} run first order Upwind (steady state)
Advection scheme: 2^{nd} run second order TVD (restart from 1^{st} run - unsteady)
EXN GPU Allocation 1 Nvidia K40
EXN CPU Allocation 2 Intel Xeon 2.6GHz
X-axis orientation Longitudinal axis
Y-axis orientation Yaw axis
Z-axis orientation Pitch axis
Boundary Conditions
Turbulence Kinetic Energy 0.0001 m^{2}/s^{2}
Turbulence K Dissipation 0.0003 m^{2}/s^{3}
Wall model Smooth wall
Outlet Constant Specified Pressure
Initial Velocity [170.1, 0, 0] m/s
Angle of Attack 0.0
Mach Number 0.5
Temperature 300 Kelvin
Pressure 101.325 kPa
Fluid Settings
Initial Velocity [170.1, 0, 0] m/s
Initial Turbulence Kinetic Energy 0.0001 m^{2}/s^{2}
Initial Turbulence K Dissipation 0.0003 m^{2}/s^{3}
Turbulence Model RANS SST k-ω
Flow type Compressible air, Ideal Gas (R = 286.9 J/kg K)
Constant Viscosity 1.715 x 10^{-5} kg/m s
Precision Double precision
Simulation Outcomes, Timing, and External Factors
The Mach number distribution is shown in Fig. 5. The pressure distribution on the missile surface is shown in Fig. 6. A vortex ring is formed at the tail of the missile and it is resolved quote well by EXN/Aero as it is shown in Fig. 7. Moreover, flow separation takes place over the canards and the separation bubble is resolved quite well also, as it is seen in Fig. 8.
(a)
(b)
(c)
Figure 5. Mach number distribution
Figure 6. Pressure distribution over the missile
Figure 7. Vortex ring at the tail of the missile
Figure 8. Separation bubble over the canards.
Table 1: Simulation performance outcomes
Reporting Item |
EXN/Aero |
Simulation starting conditions |
Transient simulation initial conditions corresponded to steady state calculation. |
Time to 1st wash-through |
Simulated time of 0.0082 sec requires 820 time steps. This is equivalent to a wall clock-time of ~8.9hrs. |
Real time per time step |
39sec |
CPU type |
Intel Xeon @ 2.6GHz |
CPU cores |
2 |
GPU type |
NVIDIA Tesla K40 |
Keywords
- External Unsteady Flow
- RANS SST k-ω
- Compressibility
- Energy Models
- Double Precision