Thanks for the hint. I have now the following:
for (unsigned int t_step=0; t_step != n_timesteps; ++t_step) {
system.time += deltat;
equation_systems.reinit();
system.solve();
}
But in my FEMSystem::element_time_derivative(...) I have to do the
following to get the correct displacements for the new timestep:
for (unsigned int i = 0; i < 3; ++i) {
const std::vector<dof_id_type>& u_idx =
context.get_dof_indices(i);
DenseVector<Number> Uc(u_idx.size());
context.get_system().get_dof_map().extract_local_vector(*context.get_system().current_local_solution,u_idx,Uc);
for (unsigned int j = 0; j < u_idx.size(); ++j) {
U[j][i] = Uc.el(j);
}
}
instead of using just context.get_elem_solution(i).el(j);
Is there an easier way for doing this?
Robert
On Fri, May 30, 2014 at 08:47:28AM -0500, Roy Stogner wrote:
>
> On Fri, 30 May 2014, Robert wrote:
>
> > Are there any additional step necessary in each timestep to apply
> > the boundary conditions or what else am I missing?
>
> One little step, one big one:
>
> To apply the boundary conditions, you'll want to run
> EquationSystems::reinit() before each time step.
>
> Don't trust the accuracy without testing. If you're not updating
> System::time to match the *end* of your time step rather than the
> beginning when reinit gets called, then your functions will get called
> with the wrong time and you'll have an O(deltat) error from that.
> ---
> Roy
------------------------------------------------------------------------------
Learn Graph Databases - Download FREE O'Reilly Book
"Graph Databases" is the definitive new guide to graph databases and their
applications. Written by three acclaimed leaders in the field,
this first edition is now available. Download your free book today!
http://p.sf.net/sfu/NeoTech
_______________________________________________
Libmesh-users mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/libmesh-users
