James, > I have a very coarse mesh that I wish to refine, but splitting the > hexahedra unevenly. Specifically, what I want to do is to make sure that > I get mesh nodes at some specified points in space; but doing this > through successive refinement creates a huge number of cells, without > ensuring that I get nodes exactly at the desired points. > > Basically, the setup consists of an enourmous space, somewhere in which > sits a molecule. I want to start out by creating a coarse mesh that has > nodes at the centers of each of the nuclei, but with big cells, i.e. > without having to refine much of the empty space. In theory, I could > define the structure cells by hand and use > Triangulation::create_triangulation to create the mesh. However, > creating the CellData and SubCellData seems a little out of my league. > > Is there an easier way?
Not really. deal.II really only subdivides cells by regular bisection, i.e. the mesh sizes of the children is half that of the parent (assuming the parent is box shaped). You can do anisotropic refinement (where a cell is only refined in one or two directions, not all) but even then we only halve cells. In your case, the way through CellData and SubCellData would indeed sound like the way to go. You may wish to look at step-14 to see how these classes can be used -- it's really not very complicated. Cheers W. ------------------------------------------------------------------------- Wolfgang Bangerth email: [email protected] www: http://www.math.tamu.edu/~bangerth/ _______________________________________________ dealii mailing list http://poisson.dealii.org/mailman/listinfo/dealii
