CFD Online Discussion Forums

CFD Online Discussion Forums (https://www.cfd-online.com/Forums/)
-   OpenFOAM Native Meshers: snappyHexMesh and Others (https://www.cfd-online.com/Forums/openfoam-meshing-snappyhexmesh/)
-   -   Helyx-OS Replicating Model - help (https://www.cfd-online.com/Forums/openfoam-meshing-snappyhexmesh/113788-helyx-os-replicating-model-help.html)

JR22 February 26, 2013 14:50

Helyx-OS Replicating Model - help
 
Helyx-OS works very well to get a quickstart on OpenFoam models. Using the GUI + a few steps, a model is built. However, trying to copy and modify a model that already ran has proven a bit frustrating. Any body has experience with this? What am I missing?

This are the steps I take:
1. Copy the entire folder into a new one (new folder = new_name)
2. I rename any file name with the old_name to the new_name
3. I try to change any text within the old files to the new_name
Code:

find ./ -type f -exec sed -i 's/singleSnozzle-1m/singleSnozzle-2m/g' {} \;
4. The I remove the previous processors data from the run directory
Code:

rm -rf proc*
5. I go back to Helyx-OS, change the mesh parameters, and re-run it.
6. After blockMesh+snappyHexMesh do their job, you reload the case from within Helyx-OS, and it is supposed to let you move into the case parameters themselves (boundaries, etc). However, this does not happen, and it never lets me set the boundaries. Has anybody tried this? It sure would be nice a "Save-as" capability to do precisely this.

Thanks very much in advance,

Jose

chegdan February 26, 2013 17:25

If you upload an example case I will try to recreate your steps.

JR22 February 26, 2013 21:10

Hi Dan,

The description of the steps is here (I answered my own question and applied it to this problem).
http://www.cfd-online.com/Forums/ope...stl-files.html

Please find the case attached to this thread (I meshed it, ran it, and then delted the results by deleting the proc* folders and the 1000 folder (results at step 1000). My problem is in trying to re-establish a Helyx-OS case after erasing the results.

Quote:

Originally Posted by chegdan (Post 410276)
If you upload an example case I will try to recreate your steps.


elvis February 27, 2013 03:00

Hi,

if you have PyFoam installed what about using pyFoamCloneCase.py
"Creates a copy of a case with only the most essential files" http://openfoamwiki.net/index.php/Co...ting_case_data

I call PyFoam a must have for any serious OF user ;-).

chegdan February 27, 2013 13:44

Jose,

So I looked at the case and was able to start from scratch, mesh, run, reconstruct*, delete proc* and 1000, and then load the case back in Helys-OS and define BCs again and run again.

if you want to remesh the domain, you need to go in and rename/remove the polyMesh folder and when you load the case. helyx-OS will read the existing settings from the snappyHexMeshDict. You will then need to go in and redefine the base mesh accordingly in helyx-os. This will be improved in future releases.


Quote:

Originally Posted by JR22 (Post 410325)
Hi Dan,

The description of the steps is here (I answered my own question and applied it to this problem).
http://www.cfd-online.com/Forums/ope...stl-files.html

Please find the case attached to this thread (I meshed it, ran it, and then delted the results by deleting the proc* folders and the 1000 folder (results at step 1000). My problem is in trying to re-establish a Helyx-OS case after erasing the results.


JR22 February 27, 2013 18:36

1 Attachment(s)
Hi Dan,

I followed your directions by renaming the main directory and the .foam file to "gasExpansion2", erased the polymesh and the proces*, then open with HelyxOS, and clicked on the Generate Mesh button. It went to the terminal and went about its business well. However, when I reload the file into HelyxOS to show the meshing results, it did not do it. The meshing did not load into HellyxOS. I added a screenshot of the HelyxOS messages.

Thanks

Quote:

Originally Posted by chegdan (Post 410488)
if you want to remesh the domain, you need to go in and rename/remove the polyMesh folder and when you load the case. helyx-OS will read the existing settings from the snappyHexMeshDict. You will then need to go in and redefine the base mesh accordingly in helyx-os. This will be improved in future releases.


chegdan February 27, 2013 18:45

Ah ok.

When you delete the polyMesh it deletes the mesh. So a simultaneous look at the mesh and the action to allow a remesh in the mesh tab (that may be confusing) is not possible. Also, it is not necessary to change the name of the *.foam file.

Edit: Also, when I loaded your case that you provided without doing anything...I get the same error in your screenshot. When cloning, there is no need to find and replace things.

JR22 February 27, 2013 19:05

Quote:

Originally Posted by chegdan (Post 410552)
When you delete the polyMesh it deletes the mesh. So a simultaneous look at the mesh and the action to allow a remesh in the mesh tab (that may be confusing) is not possible.

I only tried to reload after generating the mesh in the same way I would with a fresh case. It normaly lets me move to the options in the "Case" tab. In this case, it didn't.

Thanks

chegdan February 27, 2013 19:19

Looks like a possible bug :) . In controlDict, you have changed the valid and supported option

Code:

writeFormat ascii;
to a currently incompatible

Code:

writeFormat binary;
when you ran your case. If you go into the controlDict and change that back to ascii, then it should read into HELYX-OS fine. This will get fixed in an upcoming release if its accepted :D I recorded a ticket on sourceforge

JR22 February 27, 2013 20:14

It worked!!!. I introduced that bug. When I ran it the first time, I set the "Write Format" in the "Run Controls" to Binary.

When the meshing ran, I did the reload, and it did update the mesh. When I went to the "Case Setup" mesh. It picked up my previous "Solution Modeling" options. It did not pick up the "Boundary Conditions". But that would be too much to ask. It works.

Thanks.


Quote:

Originally Posted by chegdan (Post 410557)
Looks like a possible bug :) . In controlDict, you have changed the valid and supported option



All times are GMT -4. The time now is 16:15.