[aspect-devel] Failure to converge

[Timo Heister]

Hi Eric,

I think I fixed it. For some reason it was using prescribed boundaries
which were zero before, but are now set for the inclusion benchmark.

On Fri, Apr 20, 2012 at 4:44 PM, Eric Heien <emheien at ucdavis.edu> wrote:
> I just tried running the latest version of Aspect (with the latest deal.ii repository code) and I get the following error when using box.prm:
>
> login1$ ./lib/aspect box.prm
> Number of active cells: 256 (on 5 levels)
> Number of degrees of freedom: 3,556 (2,178+289+1,089)
>
> *** Timestep 0:  t=0 seconds
>   Solving temperature system... 1 iterations.
>   Rebuilding Stokes preconditioner...
>   Solving Stokes system...
>
> +---------------------------------------------+------------+------------+
> | Total wallclock time elapsed since start    |      10.7s |            |
> |                                             |            |            |
> | Section                         | no. calls |  wall time | % of total |
> +---------------------------------+-----------+------------+------------+
> | Assemble Stokes system          |         1 |    0.0248s |      0.23% |
> | Assemble temperature system     |         1 |    0.0464s |      0.43% |
> | Build Stokes preconditioner     |         1 |    0.0247s |      0.23% |
> | Build temperature preconditioner|         1 |   0.00457s |     0.043% |
> | Solve Stokes system             |         1 |      10.6s |        99% |
> | Solve temperature system        |         1 |  0.000835s |    0.0078% |
> | Setup dof systems               |         1 |    0.0265s |      0.25% |
> +---------------------------------+-----------+------------+------------+
>
>
>
> ----------------------------------------------------
> Exception on processing:
> Iterative method reported convergence failure in step 2467 with residual 5.67724
> Aborting!
> ----------------------------------------------------
>
> This same error occurs on 3 different platforms so I don't think it's a problem with my setup.  Has something changed lately in Aspect that would cause this?
>
> Thanks,
>
> -Eric
>
> _______________________________________________
> Aspect-devel mailing list
> Aspect-devel at geodynamics.org
> http://geodynamics.org/cgi-bin/mailman/listinfo/aspect-devel



-- 
Timo Heister
http://www.math.tamu.edu/~heister/