On Thu, 27 May 2010, Andrea Hawkins wrote:

> Ah... Right! Thanks for helping me debug! ;)
>
> Yes, it is running now, and it also seems to be satisfying periodic
> boundary conditions.
>
> Thank you!

Oh, don't thank me too fast.  It's been years since I needed to use
periodic BCs myself, so the bad interaction with
more-recently-threaded code may not be the only bug lurking there.
One warning in particular: some INL folks may have found a bug when
you combine adaptivity with periodic BCs.  They saw a conservation
failure when simulating with different h levels on different sides of
a periodic BC.  Unfortunately my old simulations don't exercise that
combination well enough, they haven't cooked up a test case for me
yet, and I don't have time to fix or even verify the problem on my
own.  Work too fast and you might end up becoming the first guinea pig
instead.
---
Roy

------------------------------------------------------------------------------

_______________________________________________
Libmesh-users mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/libmesh-users

Reply via email to