[dune-pdelab] Instationary Navier-Stokes equations with Finite Volume method : Getting the velocity components at the center of the faces

[Christoph Grüninger]

Hi Samuel,
why do you write your own local elements? P1 element (Pk with
k=1) should be right for finite volumes.
If you insists on your own local element, you have do the
change in the LocalCoefficient class where you have to
calculate the LocalKey. You can find more information in
dune-localfunctions how-to.

BTW, what finite volume Stokes do you implement? Most of them
do not fullfill the inf-sub condition and are thus not stable.
Staggered grid is a well-known exception but it is difficult
to implement in PDELab. I did it last year and the code is
quite ugly and it does not include the symmetric velocity
gradient.

Bye
Christoph

-- 
Das Kommt darauf an, in welchen Zeiträumen du
rechnest. Für Geologen: ja, in naher Zukunft
[Stefan Seyfried in suse-laptop]