On Mon, May 19, 2008 at 1:55 PM, Nasser Mohieddin Abukhdeir <[EMAIL PROTECTED]> wrote: > Hello: > A while ago I mentioned something about the time stepping algorithm > employed by AdaptiveTimeSolver as being non-ideal for my PDE system. I > think I have implemented a very simple adaptive time solver based on the > local truncation error: > > core_time_solver->du() / calculate_norm(_system, *_system.solution)
As it's currently implemented, du() is the difference between two successive timesteps, not a measure of the truncation error. While this may be working for your particular application, I don't know of a theoretical justification for this particular method of adaptive timestep selection. -J > where it simply increases the time step up to a default of 2 > (max_growth=2). The only advantage of this approach for me application > is that the default algorithm is expensive (three nonlinear steps for > one effective time step) and, while locally it gives a good estimate of > the error, is too optimistic with its adaptive time stepping (my PDEs > are highly nonlinear). > If this makes any sense for inclusion into LibMesh's > AdaptiveTimeSolver I can clean-up and provide the code for review. > > -- > Nasser Mohieddin Abukhdeir > Graduate Student (A.D. Rey Research Group) > McGill University - Department of Chemical Engineering > http://webpages.mcgill.ca/students/nabukh/web/ > http://people.mcgill.ca/alejandro.rey/ > > > ------------------------------------------------------------------------- > This SF.net email is sponsored by: Microsoft > Defy all challenges. Microsoft(R) Visual Studio 2008. > http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/ > _______________________________________________ > Libmesh-users mailing list > [email protected] > https://lists.sourceforge.net/lists/listinfo/libmesh-users > ------------------------------------------------------------------------- This SF.net email is sponsored by: Microsoft Defy all challenges. Microsoft(R) Visual Studio 2008. http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/ _______________________________________________ Libmesh-users mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/libmesh-users
