B Y : E V A N N E Y 1 2 / 1 6 / 1 4
CFD Simulation B Y : E V A N N E Y 1 2 / 1 6 / 1 4 Converging - - PowerPoint PPT Presentation
CFD Simulation B Y : E V A N N E Y 1 2 / 1 6 / 1 4 Converging - - PowerPoint PPT Presentation
Converging Diverging Nozzle CFD Simulation B Y : E V A N N E Y 1 2 / 1 6 / 1 4 Converging Diverging Nozzle - Background Hour glass shaped nozzle HIGH Fully subsonic isentropic flow Velocity increases in converging section and
Converging Diverging Nozzle - Background
Hour glass shaped nozzle
Inlet of decreasing cross sectional area
Outlet of increasing cross sectional area
Used to accelerate pressurized fluid to supersonic speeds
Useful in many turbine and jet engine applications.
Reference 1
Outlet (Back) Pressure HIGH LOW
Fully subsonic isentropic flow
Velocity increases in converging section and decreases in diverging section
Flow becomes choked (Mach 1 at throat)
Upper limit of back pressure range for shocks
Converging section becomes time independent
Normal shock wave is formed near throat
Instantaneous property changes across shock
Pressure increase
Temperature increases
Velocity decreases
Shock wave location moves towards outlet
Shock is just at outlet. More complex shock patterns begin to occur
Flow pressure at exit is less than ambient pressure – Overexpanded nozzle
Fully supersonic isentropic flow (design condition)
Lower limit of back pressure for shocks in
- verexpanded flow
Back pressure low enough that it equals the nozzle exit pressure
Shocks disappear
𝒆𝑩 𝑩 = 𝒆𝑾 𝑾 (𝑵𝟑 − 𝟐)
Project Aims
2)
This project aims to examine viscous boundary layer separation with shock waves in an over expanded supersonic nozzle.
1)
This project aims to examine inviscid and real flow in a 3D supersonic converging diverging nozzle and the development of normal shocks in the nozzle.
Geometry and Mesh
Geometry
The nozzle profile will aligned with that tested by Craig Hunter, NASA (profile depicted below)
Throat area At =4.317 in^2
Expansion ratio Ae/At = 1.797
Width 3.99in
Best mesh:
Profile:
Edge sizing with number of divisions per section
Bias towards wall
Sweep:
Swept through the width with bias towards edges
Mirrored over symmetry plane
ANSYS Fluent Setup
INVISCID MODEL
Solver Settings:
Steady solver
Density based (as the flow is compressible)
Inviscid model
2nd order upwind
Boundary conditions:
Pressure inlet at inlet
Pressure outlet at outlet
No slip wall on sides and top plane
Symmetry on bottom plane of model
Material Properties:
Working fluid was standard air
Air as ideal gas model TURBULENT MODEL
Solver Settings:
Steady solver
Density based (as the flow is compressible)
K-Epsilon/K-Omega
2nd order upwind/1st order upwind
Boundary conditions:
Pressure inlet at inlet
Pressure outlet at outlet
No slip wall on sides and top plane
Symmetry on bottom plane of model
Material Properties:
Working fluid was standard air
Air as real gas (Soave Redlich Kwong) NPR = inlet total pressure / outlet pressure
Inviscid Results
Mach: NPR 8.78 (isentropic) Mach: NPR 1.4 (shock)
Turbulent Results
Mesh 3 vs Empirical
NPR 1.4 NPR 2.0 NPR 3.0
Pressure Comparison Solver Study NPR 2.0 Velocity
References
1.
http://www.ivorbittle.co.uk/Books/Fluids%20boo k/Chap[6pter%2013%20%20web%20docs/Chapte r%2013%20Part%203%20Complete%20doc.htm
- 2. http://www.engapplets.vt.edu/fluids/CDnozzle/cdi