Re: [deal.II] Re: Nedelec Elements for unstructured adaptively refined meshes

The problem in the thread you cited is for 3D. I am happy to update our disclaimer for the class. I am just saying that our statement: "For the moment, it is safe to use the element on globally refined meshes with consistent orientation of faces." with the addition "in 2D" and for "lowe

Re: [deal.II] Re: Nedelec Elements for unstructured adaptively refined meshes

2018-05-10 16:52 GMT-04:00 Matthias Maier : > with the addition "in 2D" and for "lowest and second order" is true :-) > Yeah you are right, it's always in 3D that people have problems. Bruno -- The deal.II project is located at http://www.dealii.org/ For mailing list/forum options, see https:/

Re: [deal.II] Re: Nedelec Elements for unstructured adaptively refined meshes

On Thu, May 10, 2018, at 15:41 CDT, Bruno Turcksin wrote: > Many people have reported problems see this thread for example > https://groups.google.com/forum/#!searchin/dealii/nedelec%7Csort:date/dealii/ao-bi_Nh7gc/jHq1qMJlAgAJ > If I remember correctly, J-P couldn't make Nedelec work for his pr

Re: [deal.II] Re: Nedelec Elements for unstructured adaptively refined meshes

2018-05-10 16:35 GMT-04:00 Matthias Maier : > > We keep repeating that the current element only works on Cartesian > meshes, but I am not sure whether this is actually true. > Many people have reported problems see this thread for example https://groups.google.com/forum/#!searchin/dealii/nedelec%7

Re: [deal.II] Re: Nedelec Elements for unstructured adaptively refined meshes

On Thu, May 10, 2018, at 08:10 CDT, Bruno Turcksin wrote: > The current version of FE_Nedelec should only be used for cartesian mesh. I have had no problems on unstructured, locally refined 2D meshes with our FE_Nedelec element so far. (With verified correct results when comparing against ana

[deal.II] Re: Nedelec Elements for unstructured adaptively refined meshes

Artur, On Wednesday, May 9, 2018 at 8:51:37 PM UTC-4, Artur Safin wrote: > 1) Can FE_Nedelec > work > for unstructured adaptively refined meshes (for linear polynomials)? > 2) There is a new Nedelec FE class (pull 2240 >