Re: [deal.II] Re: Anisotropic fe space
Well, when I asked the question, I did not know how to do it. I then saw from my own answer to Jean-Paul's question that I could use a lower degree FE_RaviartThomasNodal itself to get test functions for moments. Thanks praveen On Mon, Aug 28, 2017 at 9:51 PM, Wolfgang Bangerthwrote: > On 08/28/2017 09:42 AM, Praveen C wrote: > >> Yes, I could use FE_RaviartThomasNodal(k-1) to get the test >> functions, though I would have liked to choose other nodes. >> > > Then I don't think I understand what the original question was. Do you > need an element, or just the polynomials? For the latter, you can still see > how FE_RTNodal constructs the polynomials it then uses as its shape > functions. > > > Best > W. > > -- > > Wolfgang Bangerth email: bange...@colostate.edu >www: http://www.math.colostate.edu/~bangerth/ > > -- > The deal.II project is located at http://www.dealii.org/ > For mailing list/forum options, see https://groups.google.com/d/fo > rum/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. > -- 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.
Re: [deal.II] Re: Anisotropic fe space
On 08/28/2017 09:42 AM, Praveen C wrote: Yes, I could use FE_RaviartThomasNodal(k-1) to get the test functions, though I would have liked to choose other nodes. Then I don't think I understand what the original question was. Do you need an element, or just the polynomials? For the latter, you can still see how FE_RTNodal constructs the polynomials it then uses as its shape functions. Best W. -- Wolfgang Bangerth email: bange...@colostate.edu www: http://www.math.colostate.edu/~bangerth/ -- 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.
Re: [deal.II] Re: Anisotropic fe space
On 08/28/2017 06:51 AM, Praveen C wrote: I am working with Raviart-Thomas spaces. For degree k, the x component of the vector field would have degree (k+1) in x variable and degree k in y variable, i.e., it is in Q_{k+1,k}. Similarly the y-component is in Q_{k,k+1}. This is already provided in FE_RaviartThomasNodal. I need the test functions for the moments, and these test functions which live in Q_{k-1,k} for x-component and in Q_{k,k-1} for y-component. Since you already found how the FE_RaviartThomas does it, can you not use the same approach for your element? Best W. -- Wolfgang Bangerth email: bange...@colostate.edu www: http://www.math.colostate.edu/~bangerth/ -- 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.
Re: [deal.II] Re: Anisotropic fe space
Hello Jean-Paul I am working with Raviart-Thomas spaces. For degree k, the x component of the vector field would have degree (k+1) in x variable and degree k in y variable, i.e., it is in Q_{k+1,k}. Similarly the y-component is in Q_{k,k+1}. This is already provided in FE_RaviartThomasNodal. I need the test functions for the moments, and these test functions which live in Q_{k-1,k} for x-component and in Q_{k,k-1} for y-component. Best praveen On Mon, Aug 28, 2017 at 6:09 PM, Jean-Paul Pelteretwrote: > Hi Praveen, > > I'm a little bit naïve when it comes to these things, but can you not > achieve this by using an FESystem with each component given by a different > FE and ensuring that all components are fully coupled? > > Regards, > Jean-Paul > > > On Monday, August 28, 2017 at 12:32:33 PM UTC+2, Praveen C wrote: >> >> Hello >> >> I want a space of polynomials like Q_{r,s} where the degree is r in one >> direction and s in another direction. Discontinuous is fine. I know one can >> construct anisotropic quadrature rules, but I dont see a way to get an fe >> space. FE_DGQArbitraryNodes seems to need same degree in all directions. >> >> Thanks >> praveen >> > -- > 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. > -- 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.
[deal.II] Re: Anisotropic fe space
Hi Praveen, I'm a little bit naïve when it comes to these things, but can you not achieve this by using an FESystem with each component given by a different FE and ensuring that all components are fully coupled? Regards, Jean-Paul On Monday, August 28, 2017 at 12:32:33 PM UTC+2, Praveen C wrote: > > Hello > > I want a space of polynomials like Q_{r,s} where the degree is r in one > direction and s in another direction. Discontinuous is fine. I know one can > construct anisotropic quadrature rules, but I dont see a way to get an fe > space. FE_DGQArbitraryNodes seems to need same degree in all directions. > > Thanks > praveen > -- 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.