[petsc-users] difference between left and right pc

[Klaij, Christiaan]

>   What happens if you use -pc_type lu  ?

Barry,

I can't do that, it's a matshell.

Chris

>
>
> On Sep 21, 2012, at 3:29 AM, "Klaij, Christiaan" <C.Klaij at marin.nl> wrote:
>
> >
> > When I use zero initial guess, GMRES with left PC gives a huge
> > jump in true resisdual between iteration 0 and 1 and GMRES with
> > right PC is stuck, the solution remains zero, as mentioned before.
> >
> > When I use the Knoll trick, both issues are gone (!) and I do get
> > similar results for left and right preconditioning, both for the
> > iteration count and for the physics of the solution.
> >
> > I didn't expect such a difference, did you? If so, why? Somehow
> > it must be related to the rhs being quite small.


dr. ir. Christiaan Klaij
CFD Researcher
Research & Development
E mailto:C.Klaij at marin.nl
T +31 317 49 33 44

MARIN
2, Haagsteeg, P.O. Box 28, 6700 AA Wageningen, The Netherlands
T +31 317 49 39 11, F +31 317 49 32 45, I www.marin.nl