John,
> > So far I am not hearing that the problems of using CAD with MC
> > are any more difficult than the problems of using CAD with any
> > other CAE tool. In fact, the opposite may be true, in that
> > meshing may require more cleaning/healing than MC.
> Did I capture this correctly?
It’s a bit tricky to answer this in a general way. It depends on which usage scenario we’re talking about. If the idea is to have a full CAD layout of a complete facility, like the common civil engineering drawings, and use this “as is” for MC radiation modelling then it’s much more complicated to get it working. The CFD guys usually work on stripped down models already and thus, it’s much simpler to mesh and clean them. In addition the civil engineering models contain tons of info and geometry details which are not really needed for particle transport calculations but create potential problems as well as a sometimes considerable runtime overhead.
However, the actual difficulties and possible performance penalties that one might encounter highly depend on the implementation details and the flexibility of the geometry kernel & the tracker of the respective radiation transport code. One should not forget that tracking in CSG is very fast and the implementation is much simpler than in CAD kernels that should support complex shapes which do not necessarily have an analytic representation. Of course you can make tracking in CAD very very fast using kd-tree acceleration, hierarchical BB-trees, 3D digital differential analyzer etc. Yet, I have not seen a lot of those techniques in traditional radiation transport codes due to the historic tendency towards CSG.
To my mind an interesting approach would be the option to combine both worlds – use CSG for simple cases and import single but complex objects from CAD to be used inside a CSG model. That would give me more flexibility and I could restrict the level of complexity to those objects where I really need it in contrast to full blown CE drawings which might include even door handles.
Cheers
Chris
__________________________________________________________________________
You can manage unsubscription from this mailing list at
https://www.fluka.org/fluka.php?id=acc_info
Received on Thu Feb 04 2016 - 20:21:51 CET