I am interested in trying PyFr for our low Re external flow problems (Re ~ 200k and below).
How capable is PyFr for these types of problems? I saw limited information online, but the YouTube video of the “spoiler” looks promising.
What numerical output does PyFr offer? Specifically for 2D and 3D problems can it output Fx, Fy so that we can calculate the aerodynamic drag? For 3D problems do the VTK files also contain object surface pressure data? The YouTube example only showed the fluid flow features.
Thanks,
Victor
Hi Victor,
I am interested in trying PyFr for our low Re external flow problems (Re ~ 200k and below).
How capable is PyFr for these types of problems? I saw limited information online, but the YouTube video of the “spoiler” looks promising.
This should be OK.
What numerical output does PyFr offer?
Can export to .vtu or .pvtu for visualisation in Paraview.
Specifically for 2D and 3D problems can it output Fx, Fy so that we can calculate the aerodynamic drag?
Yes, see [soln-plugin-fluidforce-name] in the User Guide.
For 3D problems do the VTK files also contain object surface pressure data? The YouTube example only showed the fluid flow features.
No, not explicitly as a separate data structure, although you could extract it.
Cheers
Peter
Hi Peter,
thank you for the answers. I have since managed to set up PyFr and run the example cases.
I then attempted to run the NACA 0012 example posted by another user about a year ago, but it will not run due to significant version changes with PyFr since then.
I should be able to make it run, eventually, but perhaps you can help with a specific question when it comes to external flow simulations. GMsh seems to be rather poorly suited for airfoil mesh generation. It can be done, but takes a very long time defeating the purpose of having a fast solver - ie. if the experiment setup takes “forever” (see ANSYS), there is a limited benefit to getting a fast solution.
Thus, do you have any suggestions regarding what software or methodology can be used to create airfoil+suitable domain CGNS meshes, quickly and simply?
Thanks,
Victor
Hi Victor,
thank you for the answers. I have since managed to set up PyFr and run
the example cases.
I then attempted to run the NACA 0012 example posted by another user
about a year ago, but it will not run due to significant version changes
with PyFr since then.
I should be able to make it run, eventually, but perhaps you can help
with a specific question when it comes to external flow simulations.
GMsh seems to be rather poorly suited for airfoil mesh generation. It
can be done, but takes a very long time defeating the purpose of having
a fast solver - ie. if the experiment setup takes "forever" (see ANSYS),
there is a limited benefit to getting a fast solution.
Thus, do you have any suggestions regarding what software or methodology
can be used to create airfoil+suitable domain CGNS meshes, quickly and
simply?
"I hate meshes. I cannot believe how hard this is. Geometry is hard."
— David Baraff, Senior Research Scientist, Pixar Animation Studios
and if regular meshing is hard then generating curved meshes is at least
an order of magnitude more difficult on top of that.
In terms of format support it should not be too difficult to get grids
into PyFR. The Gmsh format is reasonably well supported and Gmsh itself
can import a wide range of mesh file formats (hence permitting them to
be converted to .msh).
Furthermore, Pointwise has native support for writing out grids in our
native .pyfrm format and are working towards high-order grid generation.
Hence, if you want to generate a grid and get it into PyFR Pointwise is
almost certainly your best bet at the moment.
Gmsh is improving...slowly. But it appears as if they are more focused
towards heat transfer type problems rather than CFD type problems (where
you need a boundary layer). As such Gmsh is extremely clunky and for
every two steps forwards they seem to move one step backwards.
Regards, Freddie.