Add test for ellipsoidal chunk manifold conversion #2175
Conversation
gassmoeller
force-pushed
the
ellipsoidal_fixes
branch
from
77b0502
to
3ce93ee
Compare
There was a problem hiding this comment.
Hey Rene,
Thanks for looking at this! This is indeed reassuring, and addresses the part of issue #1950 which needed to be solved before the release of ASPECT2.0. But I would still like to have an documentation somewhere in the code of what this equation is based on.
So only some minor comments. When addressed, you can merge yourself!
| test_points.push_back(Point<3> (25000.0,25000.0,1000000.0)); | ||
| test_points.push_back(Point<3> (25000.0,25000.0,500000.0)); | ||
| test_points.push_back(Point<3> (25000.0,25000.0,50000.0)); | ||
|
|
There was a problem hiding this comment.
Can you for the sake of clarity in the output put a line in there which outputs something like "simple sphere test:". And also do that for the others?
tests/ellipsoidal_chunk_geometry.cc
Outdated
| test_point(ellipsoidal_manifold,test_points[i]); | ||
|
|
||
|
|
||
| std::cout << "Ok" << std::endl; |
There was a problem hiding this comment.
This line seems a bit useless, because no real check is done. Even if the output would not be oke, it would still output ok, which is a bit strange.
gassmoeller
force-pushed
the
ellipsoidal_fixes
branch
from
3ce93ee
to
5a8e038
Compare
|
Thanks, I addressed the comments. |
This closes #1950. While the manifold conversion can produce significant errors very close to the center (e.g. 10% at a distance of 50 km from the center as mentioned in #1950), it is already negligible at the inner-core boundary (1e-7 at 1000km radius), and completely insignificant at the CMB (5e-11 at 2890 km depth).