Handle stateful properties in xfem moving interface #16278
Conversation
hugary1995
force-pushed
the
xfem_heal_stateful
branch
from
e31fc03
to
84af071
Compare
hugary1995
changed the title
|
Job Documentation on 56131a3 wanted to post the following: View the site here This comment will be updated on new commits. |
|
Pinging @bwspenc . |
|
Job Documentation on 84af071 : invalidated by @bwspenc |
hugary1995
force-pushed
the
xfem_heal_stateful
branch
2 times, most recently
from
7918d2a
to
73850b2
Compare
modules/tensor_mechanics/src/materials/ComputeFiniteStrainElasticStress.C
Show resolved
Hide resolved
hugary1995
force-pushed
the
xfem_heal_stateful
branch
from
73850b2
to
070052d
Compare
modules/xfem/test/tests/moving_interface/moving_ad_bimaterial_finite_strain.i
Outdated
Show resolved
Hide resolved
modules/xfem/test/tests/moving_interface/moving_bimaterial_finite_strain.i
Outdated
Show resolved
Hide resolved
hugary1995
force-pushed
the
xfem_heal_stateful
branch
from
070052d
to
95110b2
Compare
|
Can someone help me invalidate the failing tests due to conda? |
Looks good now. |
store stateful properties before healing, and set them back after re-cut. close idaholab#15817
hugary1995
force-pushed
the
xfem_heal_stateful
branch
from
7e71356
to
fd54bbe
Compare
|
Job Non unity build on e90f0db : invalidated by @bwspenc Failure seems unrelated to this PR. Just checking. |
bwspenc
changed the title
|
@rwcarlsen @bwspenc I just switched everything to use the API you suggested. But instead of changing the existing method using element references, I added a new method using pointers. If we had been doing things correctly this should not break any test. |
hugary1995
force-pushed
the
xfem_heal_stateful
branch
from
40d80ab
to
91b3d99
Compare
There was a problem hiding this comment.
All the original copy methods in MaterialData and MaterialPropertyStorage should just call the new functions you created - no need to duplicate all that code (i.e. copy(const Elem & ...) {copy(&elem_to, &elem_from...);}). And since the initProps function is just private/internal - you should just delete the old version and make all calls use the new version.
This is a reopening of stale PR #15818.
This PR handles stateful properties when an XFEM cut is healed and re-cut.
closes #15817