On 3/25/19 10:46 AM, [email protected] wrote:
> 
> I have v = (0,-y^3) so just vertical, and p = x^3(1-y), easy enough.
> 
> on a rectangular domain:
> first set of BCs, no flux on sides (boundary id 0), some top rock stress 
> function (boundary id 1 for top), and Dirichlet at bottom boundary 
> (boundary id 2 for bottom).
> second set of BCs, no flux on sides, some top rock stress function, and 
> no flux on bottom boundary.
> 
> So the only difference is the bottom boundary - one has Dirichlet, the 
> other has no flux.
> 
> I do a convergence test:
> using first set of BCs, refining the grid, get correct convergence rates 
> etc.
> 
> second set of BCs, blows up - solution on bottom boundary looks like the 
> attached picture. blows up along the bottom boundary.
> 
> Why does this not work?????!!!!!

I don't know, but there are great clues that should help you find out 
what the problem is. First, what is the mesh? Do you just have a mesh 
with 4 cells along the bottom? If so, then each of the spikes would 
correspond to one node. But looking at the shape of the spikes, I think 
that you actually have many more cells along the bottom edge, so what 
distinguishes the three points where there is a spike? If you can find 
out what is special about these three points, you might be able to 
identify what is going wrong there!

Best
  W.

-- 
------------------------------------------------------------------------
Wolfgang Bangerth          email:                 [email protected]
                            www: http://www.math.colostate.edu/~bangerth/

-- 
The deal.II project is located at http://www.dealii.org/
For mailing list/forum options, see 
https://groups.google.com/d/forum/dealii?hl=en
--- 
You received this message because you are subscribed to the Google Groups 
"deal.II User Group" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to [email protected].
For more options, visit https://groups.google.com/d/optout.

Reply via email to