On 05/14/2017 01:57 AM, hanks0...@gmail.com wrote:
I guess this problem comes from the second derivatives of solution.
I don't know whether the problem really comes from that, but you can't take
the second derivative of a finite element solution without problem. The issue
is that it exists, of course, in the interior of elements. But it is a
delta-function like thing at the interfaces between cells because the finite
element solution is continuous but not continuously differentiable across the
interface (it has a "kink"), and consequently the second derivative consists
of delta functions. I suppose you are not taking them into account in your
formulation.
If you do need second derivatives, you need to either use a C^1 element (quite
difficult to do), or use a mixed formulation in which you rewrite the equation
-Delta q = f into
u + nabla q = 0
-div u = -f
and then use a continuous element for the variable u. This way u=-grad q,
and Delta q = -div u, which you can compute because u is continuous.
Best
W.
--
------------------------------------------------------------------------
Wolfgang Bangerth email: bange...@colostate.edu
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 dealii+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.