Partitioned heat conduction with Schwarz-type domain decomposition#381
Partitioned heat conduction with Schwarz-type domain decomposition#381BenjaminRodenberg merged 24 commits intoprecice:developfrom
Conversation
uekerman
left a comment
uekerman
left a comment
There was a problem hiding this comment.
Could we add a picture showing the overlapping domain? I think this could be very helpful to quickly grasp the difference of this tutorial to others.
How can we publish these kind of academic and advanced tutorials in a proper way?
This discussion will be part of the preECO project. Let's not preempt a decision here. For the moment, what you did looks good.
I did not run the case.
Co-authored-by: Benjamin Uekermann <[email protected]>
uekerman
left a comment
uekerman
left a comment
There was a problem hiding this comment.
Name and README look good enough for me. I did not run the case.
...ioned-heat-conduction-overlap/images/tutorials-partitioned-heat-conduction-overlap-setup.png
Show resolved
Hide resolved
MakisH
left a comment
MakisH
left a comment
There was a problem hiding this comment.
I have not tried to run this tutorial, but it anyway needs porting to preCICE v3.
I would also prefer that it already applies the tutorial structure that we have essentially discussed and converged on, which is documented in https://precice.org/community-contribute-to-precice.html
Other than that, it looks clean and clear. I particularly like that you added the TikZ sources for the figure.
partitioned-heat-conduction-overlap/fenics/precice-adapter-config-L.json
Outdated
Show resolved
Hide resolved
...ioned-heat-conduction-overlap/images/tutorials-partitioned-heat-conduction-overlap-setup.tex
Show resolved
Hide resolved
Co-authored-by: Benjamin Uekermann <[email protected]>
MakisH
left a comment
MakisH
left a comment
There was a problem hiding this comment.
The new structure is nice! 👍
I ran the code, no VTK files are produced (as expected, I supposed).
I then executed the solver-fenics/errorcomputation.py and got no output, which confused me but means that the assertion worked.
| * `python3 heat.py left` flag will run the left participant. | ||
| * `python3 heat.py right` flag will run the right participant. | ||
|
|
||
| Like for the case `partitioned-heat-conduction` (using Dirichlet-Neumann coupling), we can also expect for the overlapping domain decomposition applied here to recover the analytical solution. `errorcomputation.py` checks this explicitly, by comparing the numerical to the analytical solution and raising an error, if the approximation error is not within a given tolerance. |
There was a problem hiding this comment.
| Like for the case `partitioned-heat-conduction` (using Dirichlet-Neumann coupling), we can also expect for the overlapping domain decomposition applied here to recover the analytical solution. `errorcomputation.py` checks this explicitly, by comparing the numerical to the analytical solution and raising an error, if the approximation error is not within a given tolerance. | |
| Like for the case `partitioned-heat-conduction` (using Dirichlet-Neumann coupling), we can also expect for the overlapping domain decomposition applied here to recover the analytical solution. The script `solver-fenics/errorcomputation.py` checks this explicitly, by comparing the numerical to the analytical solution and raising an error, if the approximation error is not within a given tolerance. |
Co-authored-by: Gerasimos Chourdakis <[email protected]>
I don't completely understand this. I get VTK output...?
|
I guess I overlooked the |
3 participants
Adds a tutorial illustrating how Schwarz-type domain decomposition could be implemented with preCICE.
General question similar to #30: How can we publish these kind of academic and advanced tutorials in a proper way?