[Getdp] Issue with cohomology solver

[Frederic Trillaud Pighi]

Dear Guillaume,

Thanks for the quick answer, it is then an actual bug. IN my understanding, the current orientation should provide the orientation of the cut. I have already implemented the proposed solution, it is not so ideal but it leads to a temporary fix.

Best,

Frederic


> On Feb 13, 2017, at 01:40, DILASSER Guillaume <Guillaume.DILASSER at cea.fr> wrote:
> 
> Dear Frederic,
>  
> Your issue is caused by inconsistent orientation of cohomology cuts. If you mesh and generate cohomology cuts for your model within the Gmsh GUI, you will see that 3 of your cuts are oriented in one direction and the one corresponding to the first turn is in the opposite direction. Therefore the current flows in the opposite direction in the first turn. A quick fix is to set up your current constraint in the pro file on a per turn basis (put +I on each turn with the correct orientation and -I on the other) but this is not practical when the number of turns grows. Another solution is to post-process the msh file and reverse the orientation of each line element within the cuts that have bad orientation (this is presently what I do). There may be a way to handle the issue directly in Gmsh but unfortunately I was not able to find it.
>  
> Sincerely Yours,
>  
> Guillaume DILASSER
> Doctorant SACM / LEAS
> CEA - Centre de Saclay - Bât.123 - PC 319c
> 91191 Gif sur Yvette Cedex - France -
>  
> guillaume.dilasser at cea.fr <mailto:guillaume.dilasser at cea.fr>
>  
>  
> -----Message d'origine-----
> De : getdp [mailto:getdp-bounces at ace20.montefiore.ulg.ac.be] De la part de Frederic Trillaud Pighi
> Envoyé : samedi 11 février 2017 02:25
> À : getdp at geuz.org
> Cc : Cesar Simón Lopez Monsalvo <cslopezmo at conacyt.mx>
> Objet : [Getdp] Issue with cohomology solver
>  
> Dear all,
>  
> I do not know if I am doing wrong here. I have an axisymmetric model of superconducting tapes in parallel carrying the same impressed current.
>  
> In the case of triangular mesh, all the tapes carry the current in the same direction leading to the expected total current integrated over all the tapes. However, when I use quadrangles (exact same code), the first tape does not have the same direction of current as the remaining tape. It seems that there is an issue with the association of the current with the cut. I may think that it is some kind of a bug resulting from the choice of quadrangle meshing.
>  
> I have joined the code based and resulting results.
>  
> Best,
>  
> Frederic

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://onelab.info/pipermail/getdp/attachments/20170213/388a91a5/attachment.html>