Hello all In this paper
http://arxiv.org/pdf/1009.4975.pdf In page 9, at the bottom, they mention this: As indicated by the circled nodes in Figure 3(a), level-one mesh incompatibility results in hanging nodes. The libMesh package handles those hanging nodes by using the projection method to enforce constraints in the stiffness matrix ... Since libMesh does not drop the constrained DOFs in the linear system, the projected system in (7) is singular when there is any hanging node. Krylov subspace methods can handle such singularities as long as the right hand side is consistent, but these singularities may cause problems for preconditioners Is this still an issue for libmesh? What preconditioners would you recommend for a linear elasticity problem with adaptive mesh refinement? I'm imposing the dirichlet conditions as a constraint in the dof_map, not with a penalty term. When using adaptive mesh refinement, is the matrix still symmetric? Can I use cg? Thanks Miguel -- *Miguel Angel Salazar de Troya* Graduate Research Assistant Department of Mechanical Science and Engineering University of Illinois at Urbana-Champaign (217) 550-2360 [email protected] ------------------------------------------------------------------------------ Slashdot TV. Video for Nerds. Stuff that matters. http://tv.slashdot.org/ _______________________________________________ Libmesh-users mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/libmesh-users
