Hi Konstantin, Thanks a lot for the contribution! I'll be taking a look at it later in the day.
Best, J-P On Wednesday, May 3, 2017 at 6:39:58 AM UTC+2, Konstantin Ladutenko wrote: > > Dear J-P and Wolfgang, > > I`ve tried to incorporate your answers to the documentation, see > https://github.com/dealii/dealii/pull/4339 > > Best regards, > Kostya > > вторник, 2 мая 2017 г., 20:05:28 UTC+3 пользователь Jean-Paul Pelteret > написал: >> >> Hi Konstantin, >> >> Thank you for your answers! >>> >> >> You're welcome! >> >> >>> As I see the actual problem is to define a step function for material >>> parameters. In FEM simulations mesh is often aligned to the problem`s inner >>> interfaces (like in a "Better mesh" extension of step-6). Is it possible to >>> define for such a case with a discontinuous coefficient using continuous >>> element? E.g. is it valid to take coefficient evaluation out of the >>> q_index cycle and use >>> >>> current_coefficient = coefficient<dim> (cell->center()) >>> >>> in step-6 code for the case of mesh aligned to the interface of >>> coefficient drop? >>> >> >> Sure, it obviously depends on what you're trying to model but in many >> cases I would say that this is legitimate. The underlying material >> description is, for most problems (e.g. those with no discontinuities >> internal to a cell), completely independent of the chosen basis functions. >> If your mesh exactly discretises your geometry into material patches, then >> could divide up your triangulation (note: nothing to do with DoFs) into >> these different material zones using cell->set_material_id(). You could >> then choose your material parameters based on a cell->material_id(), which >> is exactly the same as what you had proposed. You are, in essence, assuming >> that the material parameters are piecewise constant over each cell. Whether >> or not this is a good approximation of how your material parameters vary in >> space is entirely up to you to decide, but I use this assumption (and the >> approach that I laid out) almost exclusively in my own work. A >> counter-example for which you'd have to be more careful are anisotropic >> (e.g. fibrous) media, like is shown here >> <https://github.com/dealii/code-gallery/tree/master/Linear_Elastic_Active_Skeletal_Muscle_Model> >> . >> >> Best regards, >> J-P >> > -- The deal.II project is located at http://www.dealii.org/ For mailing list/forum options, see https://groups.google.com/d/forum/dealii?hl=en --- You received this message because you are subscribed to the Google Groups "deal.II User Group" group. To unsubscribe from this group and stop receiving emails from it, send an email to [email protected]. For more options, visit https://groups.google.com/d/optout.
