Thank you so much Guido! In effect, I was using a relaxation parameter of 1.2. When I make it 1.0, the run converges up to the tolerance I had specified initially. And yes, it is PreconditionSSOR<SparseMatrix<double> >.
Regards, -Julian --- El mié, 12/29/10, Guido Kanschat <[email protected]> escribió: De: Guido Kanschat <[email protected]> Asunto: Re: [deal.II] Problem with CG A: "Julian" <[email protected]> Cc: [email protected] Fecha: miércoles, 29 de diciembre de 2010, 12:42 pm Julian, a remark in advance: the SSOR implementation was wrong in 6.3.1 most likely 6.2.1. It was only correct if the relaxation parameter equals to 1. It will be fixed in the next release. Did you use a relaxation parameter different from 1? And, we are talking about PreconditionSSOR<SparseMatrix<double> >? Best, Guido Julian wrote: > Hello, > > I have a code using preconditioned (SSOR) CG that converges (in <500 > iterations) in one system (32 bit, Ubuntu 9.10, gcc 4.4.1, no threading, > deal.II 6.2.1), but fails to converge (in 10,000 iterations) in another > system (4 bit, RedHat 4, gcc 4.4.2, no threading, deal.II 6.3.1). Has anyone > run into a similar problem or might otherwise know what the problem could be? > > In a nutshell my code: > - is a derivation of the Step 8 tutorial for linear elastic analysis > - uses the cell user_index to point to a vector of scalars that weight each > cell stiffness matrix (I am doing topology optimization) > - uses the periodic-constraints.h module provided in a previous posting of > this list to impose periodic boundary conditions. > > Thanks in advance for any help you could > provide! > > -Julian > > > > > > > ------------------------------------------------------------------------ > > _______________________________________________ > dealii mailing list http://poisson.dealii.org/mailman/listinfo/dealii
_______________________________________________ dealii mailing list http://poisson.dealii.org/mailman/listinfo/dealii
