Hi Bruno, I definitely agree - we should mention this in step-37. I will open an issue since I will not get to do it today.
Best, Martin On 27.10.2017 14:44, Bruno Turcksin wrote: > Martin, > > Could you add this explanation to the documentation and/or to a > tutorial. I think mentioning it step-37 would be great. > > Best, > > Bruno > > On Friday, October 27, 2017 at 3:57:09 AM UTC-4, Martin Kronbichler > wrote: > > Dear Michal, > > This is expected: the matrix-free operator evaluation cannot apply > non-homogeneous boundary conditions while solving (at least I have > never figured out how to do that). To solve such a problem, you > need to bring the non-homogeneous part on the right hand side > first. I often solve this by first creating a vector that is > filled with the Dirichlet values and apply the operator on that > (without setting the zero constraints) and then solve for an > increment that has zero boundary conditions. I attach an example > which is an extension of step-37 (and uses coefficients described > here, if you want some more background: > http://www.it.uu.se/research/publications/reports/2017-006/2017-006-nc.pdf > > <http://www.it.uu.se/research/publications/reports/2017-006/2017-006-nc.pdf> > > In the program, you will find that LaplaceOperator not only > contains a vmult() function, but also a compute_residual() > function that reads the data in from an FEEvaluation object > without Dirichlet conditions, due the operation, and assembles the > right hand side into an FEEvaluation object with Dirichlet > conditions (-> local_compute_residual()). > > Please let me know in case you have questions. > > Best, > Martin > > On 26.10.2017 22:07, Michał Wichrowski wrote: >> Dear deal.II developers, >> I think I found problem with non-homogeneus boundary conditions. >> By changing: >> VectorTools::interpolate_boundary_values (dof_handler, >> 0, >> Functions::ZeroFunction<dim>(), >> constraints); >> to >> VectorTools::interpolate_boundary_values (dof_handler, >> 0, >> Functions::ConstantFunction<dim>(5), >> constraints); >> The results shown on included picture were outputted. I've tried >> to figure out what is going on, and it looks like matrix-free >> framework is solving problem with homogeneous constrains and then >> Dirichlet nodes are overwritten. >> -- 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 >> <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 >> dealii+unsubscr...@googlegroups.com >> <mailto:dealii+unsubscr...@googlegroups.com>. For more options, >> visit https://groups.google.com/d/optout >> <https://groups.google.com/d/optout>. > > -- 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 dealii+unsubscr...@googlegroups.com > <mailto:dealii+unsubscr...@googlegroups.com>. For more options, visit > https://groups.google.com/d/optout. -- 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 dealii+unsubscr...@googlegroups.com. For more options, visit https://groups.google.com/d/optout.
