[tbraun84]

Whenever I use Workbench's ANSYS meshing to create a CFD mesh...

(as much as I prefer Gambit, it is getting harder to patch the installation to work whenever we get OS updates... but I digress...)

I get long "hierarchical" names for all of my zones. For example "wall-casename-zonename-casename (again!)-volume1"


Is there a method for deleting the redundant naming convention or scripting to map a text file list of names onto the zones? Typing out the list is still better than zone-rename and "OK" for a hundred or so zones.

For reference, I know the [next] best option is being diligent in ANSYS DesignModeler and Meshing with naming every important surface and avoiding the automatic naming convention. Still, it would be nice to have a cleanup after import for something that didn't translate.

I could not find an existing answer in the forum, but if I have unknowingly duplicated another thread, kindly direct me to the right place!

[RaiderDoctor]

Hi, I actually have never seen that happen before. Could you post a couple images showing the issue?

[`e`]

Naming the boundaries in ANSYS Meshing would be the normal approach, but you could use the TUI to rename the boundaries and cell zones in bulk, for example:

/mesh/modify-zones/zone-name 1 inlet
/mesh/modify-zones/zone-name 2 outlet
/mesh/modify-zones/zone-name 3 wall

Note: a list of the zones are available with /mesh/modify-zones/list-zones, and then an external script could be developed to automate creating the code above (based on your own naming convention).

[Micael]

As `e` wrote, naming your boundary in ANSYS MESHING is the neat way. Similarly as you would do in gambit.

From gambit to ansys meshing, there is some paradigm to overcome, but, IMO, that's totally worth the move.

[tbraun84]

Thanks for the feedback, guys.
In Workbench/Mesh I really haven't been able to replicate the "Cooper" mesh propagation which is one of the best features of GAMBIT. Otherwise, I can get to where I want to be with enough decomposition and edge/face mapping. I'd be interested in what specific paradigms you are speaking of though. Is there a "Once I started to use ____ it was much easier." ?

[Micael]

Cooper equivalent in ansys meshing is 'MultiZone Method'

I guess paradigm was a bit exaggerated, it not that bad to do the switch from gambit to ansys meshing. I did use gambit for several years than ansys meshing for years as well and would say the latter is better.

[tbraun84]

Following `e`'s advice, I exported the names using /define/bc/list-zones and copied the text to a file. Then, I did a find and replace for redundant words replacing <redundant name> with <>. I used a python script to recursively do this through all 500 names that I had! You could do the same thing in any text editor. My script then took the new lines with the zone numbers and stored them in an array. I then went through the array and recursively added "/mesh/modify-zones/zone-name" to each member of the new array. Finally I wrote that result out to another text file. It gives a list like stated above and which can be read in as a journal file.

/mesh/modify-zones/zone-name 1 inlet
/mesh/modify-zones/zone-name 2 outlet
/mesh/modify-zones/zone-name 3 wall

You could do all of this in vi or excel or many other tools I'm sure. I'm documenting this in case it helps anyone in the future.