> My problem is that I get a run time error for k greater than or equal to 2. > I get the exact solution not just for this simple case but for any exact > solution up to a certain value of k(upper limit of k varies with the exact > solution). > > It is not a time stepping issue : I reduced the time step by 50 when I > increased the constant from k=1 to k=5 and still get a run time error. > > It does not seem to be a solver issue either. > > This is the error message I get when I use the solver in step 21: > ---------------------------------------------------- > Exception on processing: > Iterative method reported convergence failure in step 2112 with residual > 1.55643e-07 > Aborting! > ----------------------------------------------------
Did you ever get this resolved? This message (as well as the one you show for BiCGStab) means that your iterative solver does not converge quickly enough (or, in the case of the other message, diverges). You may have to use a better solver/preconditioner pair, or reduce the accuracy you want to reach. Another option, at least in the beginning, when you solve small problems, is to just use the SparseDirectUMFPACK class which almost always works even if it isn't always the fastest option. However, it always let's you make sure that you assemble the linear system correctly and that the discretization works without having to also care about the solver. W. -- ------------------------------------------------------------------------- Wolfgang Bangerth email: [email protected] www: http://www.math.tamu.edu/~bangerth/ _______________________________________________ dealii mailing list http://poisson.dealii.org/mailman/listinfo/dealii
