This looks like another good application of FENothing... With this other option, you use the hp framework and associate with your void elements a FENothing class (Which means that there are no degrees of freedom associated with those elements). When you need them back again, you just associate again to them your favorite FE.
Luca. On 03/set/2010, at 06.48, Guido Kanschat wrote: > Julian, > > you can use the properties subdomain_id() or material_id() to mark cells that > are inside or outside the 'solid' region. > > Best, > Guido > > On 9/2/2010 11:30 AM, Julian wrote: >> Markus, >> >> maybe I should provide a bit more of context: I am working on topology >> optimization, and the idea is to remove 'void' elements altogether from the >> analysis. This would imply that the element gets created when the >> triangulation is created, but it does not contribute to the global stiffness >> matrix--and importantly, degrees of freedom that completely lie on the >> 'void' region (i.e. have no connection to 'solid' or 'gray' elements) are >> also taken out of the stiffness matrix. If in a subsequent iteration of >> the optimization the element becomes 'gray' or 'solid', then it should be >> included again in the analysis. >> >> -Julian >> >> --- El *jue, 9/2/10, Markus Bürg /<[email protected]>/* escribió: >> >> >> De: Markus Bürg <[email protected]> >> Asunto: Re: [deal.II] Is there a way to inactivate elements? >> A: [email protected] >> Fecha: jueves, 2 de septiembre de 2010, 11:00 am >> >> Hello Julian, >> >> could you please describe in more detail, what you mean by >> "removed from analysis"? What do you want to do? >> >> Best Regards, >> Markus >> >> >> >> Am 02.09.10 17:48, schrieb Julian: >>> Is there a way that, after a triangulation has already been >>> created, I can flag an element to be removed from the analysis >>> (possibly re-activating it afterward)? >>> >>> Thanks! >>> >>> -Julian >>> > > _______________________________________________ > dealii mailing list http://poisson.dealii.org/mailman/listinfo/dealii _______________________________________________ dealii mailing list http://poisson.dealii.org/mailman/listinfo/dealii
