Rostislav, > 1) Is there any particular reason why there are no implementations for > calculating barycenter/measure for faces in 3d?
Not really -- I was just too lazy to compute the formulas for that back when I wrote most of these functions, and nobody has needed them since. I think it should be pretty clear how to implement them if you look at how the measure and barycenter functions are computed for 3d cells, for example. If you happen to implement the missing functions, let us know and we'd be happy to add them to the library. > 2) When I generated a mesh in CUBIT deal.ii sometimes is not able to > calculate the max_coupling_between_dofs - it throws an exception. I have > looked inside the code and the problem is that calculated > max_adjacent_cells is equal to 10. Is there a simple way to make CUBIT to > generate the meshes that will be always successfully read into deal.ii? I don't know CUBIT well enough. But I think there are better ways anyway: when you call max_coupling_between_dofs, you probably do that because you want to create a SparsityPattern, but since the estimate returned by max_coupling_between_dofs is going to be a gross overestimation for cases where many 3d cells meet in one place, you'll have to allocate a ton of memory almost all of which you'll release again later on in the SP::compress() function. The better way is to generate first a CompressedSparsityPattern and the copy the result into a SparsityPattern -- see several of the tutorial programs, as well as I believe the documentation module on sparsity patterns. Best Wolfgang ------------------------------------------------------------------------- Wolfgang Bangerth email: [email protected] www: http://www.math.tamu.edu/~bangerth/ _______________________________________________ dealii mailing list http://poisson.dealii.org/mailman/listinfo/dealii
