Functionality for volumetric coupling on quadrature points of a mesh - #146
Conversation
…nd remove the old file associated with it
|
The tutorial cases to be added via precice/tutorials#653 tests this functionality. |
There was a problem hiding this comment.
Some first comments, I have not yet checked if it works with the given tutorial.
If #135 is needed for precice/tutorials#653, how can precice/tutorials#653 be a test case for this reduced version? How would the behavior change?
How easy would it be to implement another, simpler tutorial, such as https://precice.org/tutorials-volume-coupled-diffusion.html?
|
I think I did not frame my initial comment correctly. This PR contains a part of the functionality which together with the contents of #135 can be used to run the two-scale notch tutorial case. Standalone this PR cannot be tested, because we do not have a volumetric coupling case with CalculiX. We can either decide to merge this PR based on a code review, or think about adding a simple volumetric coupling tutorial case. I can look up the CalculiX tutorials to try to find a case which we can easily modify. |
|
In principle, we could merge this without a test case, hoping that someone (else?) in the future might pick into that again. But I foresee a lot of "theoretically, it supports volume coupling, but we have no example / do not really know". I would suggest to think a bit more of a simple tutorial case, ideally reusing one of the current ones, even if it does not have physical motivation. |
Actually the follow-up PR to this would be #135 which uses the functionality. I simply extracted the functionality from that PR to keep features separate and have a better git history.
Let me think a bit more about this. |
|
By the way, we should also extend the documentation of the adapter to mention the exact capabilities and limitations of this. |
Yes. Right now the documentation is in precice/precice.github.io#530. I can extract the parts relevant for this PR and prepare a new PR for the website. |
It actually looks like that PR would be ready to merge if only it shifted a bit from the multiscale to the general volume coupling perspective, showing multiscale as one example. |
I will change the language in that PR then, so we can proceed to merging it. I would also recommend merging this PR then. |
|
This pull request has been mentioned on preCICE Forum on Discourse. There might be relevant details there: |
Original PR: precice/precice.github.io#530 co-authored-by Ishaan Desai <ishaandesai@gmail.com>
|
I cleaned up the code and also removed the incorrect documentation. The functionality for extracting integration points from a provided element mesh works for C3D4 and C3D8 elements. However, this functionality is currently unusable because the data entities that the adapter can read and write are either face-based or node-based. Integration point-based data would be stresses, strains, and material stiffness. Therefore, if we merge this, we would be adding functionality that would become usable in the future. |
| ! along with this program; if not, write to the Free Software | ||
| ! Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA. | ||
| ! | ||
| ! This function is a copy of the function for calculation and printout of the lift and drag forces |
There was a problem hiding this comment.
@IshaanDesai What function does this refer to? A link to the source code of CalculiX would make it easier to port this function to a newer version of CalculiX.
There was a problem hiding this comment.
Since I have a few other PRs depending on this one, let's add the comment as soon as possible after merging.
There was a problem hiding this comment.
This function is kind of a patchwork of loops from the CalculiX source code. Previously, it was more of a direct copy, but I later removed unnecessary parts to shorten the file. I think updating this function for newer versions of CalculiX should be fairly straightforward.
3 participants
This PR adds preliminary functionality to extract integration points from a provided mesh and use them as coupling vertices. The functionality is extracted from #135 because it is deemed to be sufficient as a standalone feature. Integration points of a mesh having C3D8 (Hex) or C3D4 (Tet) elements can be extracted.