I tried by SSOR gives same behavior as jacobi. For mesh with 0, 1, and 2 times of refinement, it needs (jacobi: 20, 31, 67) and (SSOR: 10, 15, 31) iterations. Besides, is SSOR not applicable in matrix-free framework?
在2024年10月4日星期五 UTC+8 11:54:41<Wolfgang Bangerth> 写道： > On 10/3/24 01:01, 'yy.wayne' via deal.II User Group wrote: > > > > I hope to precondition a system similar to step-23, with implicit time > > discretization (backward euler or Crank Nicolson). For example, the > matrix has > > the form of M+dt^2K, where M is the mass matrix and K is the stiffness > matrix. > > I'm preconditioning it with Jacobi preconditioner, which is very > efficient > > because the system is dominated by M and applies to matrix-free > framework. > > However, when I increase the number of freedomes, iterations increase > with > > just Jacobi preconditioner. Is there another recommended preconditioner > or > > modification to simple jacobi preconditioners? I cannot find one since > jacobi > > preconditioner is quite efficient for this problem, the lack of > independency > > to h is its only drawback. > > Have you tried something like SSOR? > > Best > WB > > -- > ------------------------------------------------------------------------ > Wolfgang Bangerth email: [email protected] > www: http://www.math.colostate.edu/~bangerth/ > > > -- The deal.II project is located at http://www.dealii.org/ For mailing list/forum options, see https://groups.google.com/d/forum/dealii?hl=en --- You received this message because you are subscribed to the Google Groups "deal.II User Group" group. To unsubscribe from this group and stop receiving emails from it, send an email to [email protected]. To view this discussion on the web visit https://groups.google.com/d/msgid/dealii/2c773908-6be6-49e6-80e4-1d63bb561a55n%40googlegroups.com.
