Tutorial: transonic flow over NACA 0012
The tutorial employs a mesh generated by FreeCase group and developed for inviscid flow over the NACA0012. Please kindly let me know if it is correctly to use this mesh for the turbulent flow, in which a considerable clustering is required in the boundary layer on the airfoil ?
[Edit: Based on an email exchange]
- Login to post comments
Select Language
Free 30-Day Trials
Feedback
Questions? Ideas? Problems?
Mesh isn't Idea
You are correct that the mesh isn't idea for this transonic turbulent case. The y+ value varies between 1000-7000, so clearly there is little boundary layer resolution. However, the tutorial is really an exercise for importing a mesh and setting the flow conditions for a compressible 2D case. I'll add a note to the tutorial to clarify these points.
Thanks for the feedback.
Computing time increases awfully
Many thanks for your response. It would be great if you not only add a note to the tutorial, but provide an example of calculation of turbulent transonic on a mesh that provides y+ , say, about 10. My own tests with OpenFOAM has shown that in such a case the computing time increases awfully, so that the employment of OpenFOAM becomes senseless.
If your calculation with y+ about 10 confirms this conclusion, then the advertising of caedium for real transonic flows turns out to be misleading.
Wall Functions Help
Maybe if you are using low-Re number turbulence models (i.e., no wall functions, e.g., Launder-Sharma K-E) you might want to get down to y+ <= 10.
However, in my EngD (PhD equivalent) work with FLUENT I found reasonable results (cp and cf distributions) with wall functions (200 < y+ < 350) for RAE 2822 @ M=0.73. I think it depends on what level of accuracy you are looking for and what computing resources you have available.
Admittedly pressure-based solvers are more optimal for incompressible flows, but I think with wall functions they can also cope with transonic flow enough to make the claim.