Am Sonntag, den 06.11.2011, 14:28 +0100 schrieb robert: > > > > > > Diffusion is as "nice" as possible in terms of stability and > > well-established theory, multigrid will normally work very well. In > > the latest release of PETSc, you could try -pc_type gamg. > > Otherwise/alternatively, configure using --download-ml or > > --download-hypre, then run with -pc_type ml or -pc_type hypre. > > Thanks for your reply. > > Do you have experience with lumping of the mass matrix? > > The following options semm to work fine if I don't use lumping: > > mpiexec -np 4 ./ThermoPaine3d-opt -ksp_type fgmres -pc_type ksp > -ksp_ksp_type cg -ksp_pc_type jacobi -ksp_monitor_true_residual > -ksp_converged_reason > > (any comments to the options???) > > However, for heat diffusion, especially for non-steady starting > conditions, oscillations seem to be a common problem (AFAIK). > If I apply lumping of the mass matrix - which means I put the row-sum of > the mass matrix on the main diagonal - the solver doesn't converge any > more. Before it converged after 2 to 5 iterations, now I was running > several hundreds. > > > Maybe some of you - definitely more experienced users - have some > ideas/suggestions? > > P.S.: I have tried a lot of different combinations of solvers and > preconditioners (including those of the last reply).. > > Thank you, > Robert
I have to correct my last entry: I get the following output: 0 KSP preconditioned resid norm 9.794488229448e+17 true resid norm 9.794488229448e+17 ||Ae||/||Ax|| 9.424443120337e-03 1 KSP preconditioned resid norm 0.000000000000e+00 true resid norm 9.794488229448e+17 ||Ae||/||Ax|| 9.424443120337e-03 Linear solve converged due to CONVERGED_ATOL iterations 1 Isn't it strange to have a preconditioned residual norm of exactly 0 (that's why I thought there might be a problem somewhere)??? Thank you, Robert
