Hi Georg,

On Mon, Feb 29, 2016 at 09:00:16AM +0000, [email protected] wrote:
> I tried running my multidomain model with the AMG-Solver (not yet in
> parallel). First error that popped up was about AmgTraits not being
> set. As a first try I used the AmgTraits defined in the
> el2ppropertydefaults. However, it does not converge. For larger time
> steps I get warnings from UMFPACK that the matrix is singular. For
> smaller timesteps the solver does not converge.

Indeed, having a singular coarse level matrix is suspicous. AMG is
very sensible in the way the linear system is setup (in contrast to
direct solvers).

A good test would be to setup the AMG such that it does not do any
coarsening. Just set the first constructor parameter
CoarsenCriterion to 0. Then all AMG will do is gather the matrix on
one process and solve it directly. If that does not converge in one
step then there is definitely a problem with the parallel setup.

Cheers,

Markus
P.S.: dune-project.org is down until tonight.
-- 
Dr. Markus Blatt - HPC-Simulation-Software & Services http://www.dr-blatt.de
Hans-Bunte-Str. 8-10, 69123 Heidelberg, Germany,  USt-Id: DE279960836
Tel.: +49 (0) 160 97590858

Attachment: signature.asc
Description: Digital signature

_______________________________________________
Dumux mailing list
[email protected]
https://listserv.uni-stuttgart.de/mailman/listinfo/dumux

Reply via email to