[fkonias]

So far in V11 I could use “Meshing” in a workbench project and create a mesh file in a *.gtm or *.cmdb form and use it separately in other projects or standalone CFX-Pre mode. Now in V12 I can not save separately the mesh files or any other file type. I can only save the whole project in workbench, which connects the various steps from geometry to post, automatically without intermediate save files from every step. How can I extract a mesh file from “Meshing” in *.gtm or *.cmdb form to use separately in CFX-pre standalone mode?

Thank you for your time

[fkonias]

anyone???

[PSYMN]

There is an export option...

In ANSYS Meshing, go to File => Export...

(sorry I don't have it up in front of me so I forget the exact wording of the option, but I recall there were lots of options in there including fluent.msh files, ICEMCFD.uns file, CGNS files, etc. I don't remember if CFX Mesh was an option, but it probably was...)

Simon

[fkonias]

I know about the export tool, but there is no CFX-Meshing option. Never mind though, I found a way to get a *.cmdb file and do my job. I am sure there is a more sophisticated way than mine, but I don’t mind at the moment because I am on hurry. Thanks

[saravana]

We can export mesh from workbench as Fluent msh file and can use the same in CFX stand alone.
That is how i have been doing so for.

[kjetil]

The Fluent mesh exporter in Ansys 12.1 is terribly bad, and cannot be read by "other" software as i.e OpenFOAM. Importing it into Gambit always means a whole lot of "auto repair", as it is stated from the command field - adding points, lines and volumes for the mesh to be valid. As for the cgns exporter in workbench - it doe not work either, so Ansys has made a pretty poor job in making clients able of testing simulations in other cfd environments.

[PSYMN]

Kjetil,

That is the first complaint I have ever heard about the Fluent mesh output (and you can imagine how many thousands of customers use it regularly). Certainly, I have had no problem getting the meshes to load in Fluent, TGrid or ICEM CFD.

I have heard a minor complaint about the CGNS format (that was easily worked around and has since been fixed).

I would like to get to the bottom of this. Can you send me more specific messages or perhaps the actual files?

I would like the Workbench Files as well as the fluent.msh files so I could see if the problem is in your setup.

As far as I know, it creates a perfect standard Fluent File. If there is a problem importing a Fluent mesh file into OpenFoam, I would assume that OpenFoam has trouble with the Fluent format rather than assuming the Fluent OEM doesn't know how to output their own format. On the other hand, if you know of other "Fluent" files that load properly into OpenFoam or if you know what is different in the file formats, I could take that to development and sort this out. Perhaps the Fluent format has been enhanced since the last version that OpenFoam was setup for?

[customer]

PSYMN,

I am a newer about the ansys workbench 12.1 and in a hurry for my project, may i ask some questions about how to make the workbench to read the mesh file from other thrid party ( i.e. the mesh files are generated by fortran). i mean i only want to use the mesh , boundarys ,loading and material information form this fotran generated files and want to convert this files to the workbech mesh file which the workbench can tell.

due i have a mesh generator code, it can generate the shell mesh, so i want to transfer this mesh file to workbench and use simulation to solve it. would you have some idea about it, i am in a very very hurry about it ,hope i can get some responses soon. anyway, thanks advance.

[kjetil]

Simon,

I acknowledge the number of Ansys clients, and I'm equally surprised that this problem has not been addressed earlier.

To me, it seems like regardless the simplicity of the geometry, the Fluent file export from Ansys is possible to import into OpenFoam - at least for the past 1.5 years.

The .msh-files from Ansys are always significantly smaller than after being imported and re-saved in Gambit. Please see http://www.tadaa.no/pipetest.zip , which is a rather simple geometry - where the .msh-file ending in "_g" is the one repaired in Gambit. Do not misunderstand the 'helix' prefix of the files - this has to do with the flow characterstics.

When it comes to exporting to cgns, I simply get an error message saying that I should "contact vendor/supplier", can't quite recall the exact message. Though I have not tried to save this specific mesh to cgns.

[PSYMN]

To Customer,

WB needs to have a geometry to apply bocos, etc.

If you only have a mesh, you must first load it into FE Modeler (a free app on the workbench platform) and "skin" it. This creates a geometry from the surface of the mesh model. Then you can proceed to WB Simulation with the original mesh and the skinned geometry, apply boco's, etc.

Best regards,

Simon

[PSYMN]

Kjetil,

It took looking at your files to remind me... The file that WB outputs by default is in a binary format. OpenFoam needs the older ASCII format (hasn't caught up in the past few years).

You can get ANSYS Meshing to output in the FLuent ASCII format with an environment variable (as shown in this thread)...

AWP_WRITE_FLUENT_MESH_ASCII=1

I am not sure what the Gridgen autorepair is about, probably just a conversion process, but I will pass your nicely bundled file set to development and ask them about it. It is most likely just a newer Fluent format than the Gambit reader.

Best regards,

Simon

[PSYMN]

Kjetil,

It just occurred to me what you were seeing with CGNS export. You are not complaining about the format, but rather that it didn't let you export at all...

There was a licensing bug where the CGNS export, which was originally part of the ICEM CFD code, was still looking for an ICEM CFD output license (aioutcfd). If you didn't have this key, it may have asked you to contact your ANSYS vendor. This was not intentional and has been fixed with 13.0, which is due out in November.

In the mean time, if you had contacted your vendor or tech support, they could have sent you the extra licensing increment.

Best regards,

Simon

[customer]

PSYMN,

Thank you very much. may i ask another question, what kind of mesh file format can be read by ANSYS WORKBENCH .my mesh generator can generate three type of mesh file, one includes the nodes coordinate, one includes the elements information by nodes and material information for elements, last includes the loading information for nodes. how to conver these files to workbench, i know it need code a convertor, bur i don't know which file format can be read by workbench and how to do. sorry for a lot of questions, hope someone can response me early. thanks again.

[PSYMN]

Customer,

The mesh formats for ANSYS (.in), Nastran (.nas) or Fluent (.msh) are all published and all can be read into FE modeler. You should look at these formats and decide which best suits your needs. There are others also, please check the full list of imports to FE modeler and then decide.

It could be that your mesh is already in what could be considered a standard format or at least once you identify the name of the format you currently have, you could look for a translator to one of the formats FE modeler can import.

Do some research before coding.

Best regards,

Simon

[PSYMN]

kjetil,

Acording to the headder at the top of your fluent.msh file, it was output by WB version 11.0.7 (a version from 3 years ago).

Since version 12.0, we have used a different output interface. (Not sure why you thought you were using 12)

Please try a more recent version and let us know if it works any better for you.

[siri]

Hi Simon.

I have read thru your replies in this post and thought you would be of help in this case ?!?

I am having problem in setting up the 2D axisymmetric analysis in "stand alone" Ansys Fluent after importing a pure quad mesh generated using Ansys Mesh (Workbench 12).

I am suspecting a conflict going on between the two with respect to coordinate system orientation here - while workbench export requires the axis along Y; Fluent searches for axial along X. Am I right to say so?

Now the issue I face is that Fluent is unable to initialize the solution using velocity inlet boundary condition (i want to specify axial velocity and i tried doing it along X or Y, neither is working). In the velocity and density boxes on the Solution initialization tab, it computes them from Inlet as #.IND
What does it mean?

Lastly, I couldn't find any better than this post on the whole webpages about this ..... this post was the closest I could read on such an axisymmetric model.

Has anyone done anything like it before, plz respond if anyone did.


P.S: Don't have the license to run Fluent in workbench, hence the situation


Thanks to everyone here.

[kjetil]

Originally Posted by PSYMN

kjetil,

Acording to the headder at the top of your fluent.msh file, it was output by WB version 11.0.7 (a version from 3 years ago).

Since version 12.0, we have used a different output interface. (Not sure why you thought you were using 12)

Please try a more recent version and let us know if it works any better for you.

Simon,
thanks for your effort. I have not spent awfully lot of time on trying to export to the cgns format - it just occured to me that OpenFoam hadde import capabilites from it, as I at that point still couldn't get the Ansys .msh exporter to work (as we all know now - to get it to output to ascii). As you can see the content of "pipetest" zip file originates 9 months back in time, which happens to be the time I was working on my thesis. I picked this bundle due to the specific geometry simplicity, and I have, nor had, no control of what licenses or software versions the university department could access. As of today, I do have access to a 12.1 workbench - but I didn't want to publish here my most recent works

[PSYMN]

Siri, sorry I don't really know about that issue. I would suggest starting a new thread for it.

[alberto]

Hi PSYMN ,

you said formats are published, but where could we find the binary format specifications of .msh? FLUENT documentation seems to include only the "old" ASCII format.

Thanks,

[natantyt]

Hi

I have two .msh files for Fluent one is a random packing of particles and one is the pore network between them. These are created and meshed with OpenFoam separately. Is it possible to merge them together?I would like to analyze the whole pore/solid geometry in Fluent. Is it possible to merge the original stl. files?