> Would the general point be that if one can go to higher resolutions with > distributed parallelism that the enhanced accuracy of the enrichment > functions can be compensated for? > > > Sorry, i did not get what you mean. > > Regards, > Denis. > > Hi Denis, Thanks for your prompt answers. I was shooting for a very generic question about the tradeoffs of a high-resolution parallel approach, vs the more subtle XFEM on a coarser grid. So, for example, consider an immiscible two-phase flow where the boundary between the two fluids is defined by a level set function (like in the code gallery's two_phase_flow). One way to represent the interface accurately would be to use very high-resolution so that the location of the material discontinuity, and its effect on, say, pressure, is highly localized. Another might be to essentially smooth out the pressure over the intersected cells (with material discontinuity) through the use of enrichment functions (if I understand XFEM well).
I wonder if this is a reasonable way to view the tradeoffs around interface representation (handling internal discontinuities)? Since from an implementation point of view, I currently see a choice point between pursuing XFEM(with hp::DoFHandler) or high-res (p::d::tria). Best, Sean -- 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.
