[aspect-devel] free surf & non linear iterations

[Ian Rose]

Hi Cedric,

The main issue is that I have mostly been developing with basic IMPES, and
I have not tried to generalize the free surface computations to the
nonlinear solvers.  There should not be any difficulty in principle with
using it with other solvers, so long as you only advect the free surface
once per timestep (rather than once per stokes solve).  This is certainly
something we could look at during the hackathon, though you could try to
call free_surface_execute() at the beginning of the iterated Stokes block
and see if it works.

Also, I see you are using  3D model there.  I have not done a lot of
calculations in 3D with the free surface, and do not know how well it
works.  It is on the todo list.

Best,
Ian

On Tue, Mar 17, 2015 at 11:03 AM, Thieulot, C. (Cedric) <c.thieulot at uu.nl>
wrote:

>  Hi all,
>
> I tried this afternoon to use the free surface with the iterated Stokes
> but the code crashed because
> as you can see on the attached pic, the free surface scheme is only
> implemented for the IMPES solver.
> Why is this ? I fail to see why/how the geometry of the grid influences
> the iteration scheme scheme.
> More importantly, can it be fixed in a simple way or will this be a
> hackathon-type of undertaking ?
>
>  Cheers,
>
>  C.
>
>
>
> _______________________________________________
> Aspect-devel mailing list
> Aspect-devel at geodynamics.org
> http://lists.geodynamics.org/cgi-bin/mailman/listinfo/aspect-devel
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.geodynamics.org/pipermail/aspect-devel/attachments/20150317/030251c6/attachment-0001.html>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: Screen Shot 2015-03-17 at 18.31.38.png
Type: image/png
Size: 76868 bytes
Desc: not available
URL: <http://lists.geodynamics.org/pipermail/aspect-devel/attachments/20150317/030251c6/attachment-0001.png>