2023-11-09 Users Community Call

[gonuke]

Attending:

• UW-Madison: @gonuke, @bohmt
• Copenhagen Atomics: @ebknudsen
• MIST: @ahnaf-tahmid-chowdhury
• Naarea: @nplinden
• First Light Fusion: @shimwell
• UKAEA: @makeclean
• ANL: @pshriwise

[gonuke]

Cubit 2023.11 includes DAGMC as first class component.

Exports surface mesh generated by Cubit. More user control of mesh generation parameters using some standard mesh generation parameters, but lacking a faceting tolerance.

Metadata assignment is now more consistent with other mesh-generation metadata operations.

Builtin mesh export instead of using plugin. export cf_dagmc ...

@pshriwise will give webinar on Nov 10.

[gonuke]

@nplinden:

Challenges with imprint/merge with complex CAD-originated model.

Some discussion of best practices including experience from @bohmt

[gonuke]

@pshriwise

Python interface to interrogate DAGMC model: https://github.com/svalinn/pydagmc

Also some work on wheels for (py)MOAB by @akoen

https://github.com/Thea-Energy/moab-python/releases/tag/0.0.01-beta1

[gonuke]

Discussion of complex model comparison/benchmarks:

@shimwell - Difficult to do a direct comparison to complex CAD models because it is hard to make a CSG equivalent for benchmarking.

@makeclean - there are some examples of complex CAD models that have been simplified for conversion to CSG - those models might be good candidates for such benchmarking. However, those models often have segmenting to allow conversion that can have performance implications.

@gonuke - we could unit the segmented CAD models to have: 1) segmented CAD, 2) segmented CSG, and 3) untied CAD to compare across all

[shimwell]

Link to the Cubit Coreform talk tomorrow by Patrick https://us02web.zoom.us/webinar/register/WN_49YsNqLKSsibQVmYNGF3xQ#/registration