| Commit message (Collapse) | Author | Age |
|---|
| |
|
|
| |
git-svn-id: http://svn.einsteintoolkit.org/cactus/EinsteinInitialData/NoExcision/trunk@24 4ec1db94-0e4f-0410-ada3-8bed251432c9
|
| |
|
|
| |
git-svn-id: http://svn.einsteintoolkit.org/cactus/EinsteinInitialData/NoExcision/trunk@20 4ec1db94-0e4f-0410-ada3-8bed251432c9
|
| |
|
|
|
|
|
|
|
|
| |
grid functions using CCTK_Reduce. Now, reductions are done alongside the
computations into local arrays and then these are reduced using
CCTK_ReduceLocArrayToArray1D. This is much more efficient. In addition
OpenMP directives are added to allow for mixed MPI/OpenMP usage.
git-svn-id: http://svn.einsteintoolkit.org/cactus/EinsteinInitialData/NoExcision/trunk@15 4ec1db94-0e4f-0410-ada3-8bed251432c9
|
| |
|
|
|
|
|
|
| |
procedure. Works by setting the admbase variables to zero in these
regions before doing the smoothing. Needs some testing.
git-svn-id: http://svn.einsteintoolkit.org/cactus/EinsteinInitialData/NoExcision/trunk@12 4ec1db94-0e4f-0410-ada3-8bed251432c9
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
| |
used in the "turducken" paper. This includes some spline blending from
Erik and the most recent method of filling in the interior using an elliptic
equation with the "good" points as boundary values. This method uses a
conjugate gradient solver (built in), that should be robust (i.e. it should
always work) but may not be the fastest. However, since this is only done on
initial data this shouldn't be an issue. The elliptic method currently support
a second order, fourth order and sixth order equation, that will give C0, C1
and C2 solutions, respectively, across the boundary of the fill in region.
git-svn-id: http://svn.einsteintoolkit.org/cactus/EinsteinInitialData/NoExcision/trunk@7 4ec1db94-0e4f-0410-ada3-8bed251432c9
|
|
|
git-svn-id: http://svn.einsteintoolkit.org/cactus/EinsteinInitialData/NoExcision/trunk@2 4ec1db94-0e4f-0410-ada3-8bed251432c9
|
