The OPAL Discussion Forum

["sherlp" <1341516422 AT qq.com>]

Dear Andreas

Thank you for your reply. It helps a lot.

Moreover, though I am working on adjustment of the whole domain because it has
to be predefined in Amrex, I don't fully understand the domain transform
part(2.3.2) in Matthias Frey's thesis. It says that:

"In order to prevent particles leaving the predefined domain of the mesh where
the AMR hierarchy is built, they are mapped into a computation space denoted
by Sc for the evaluation of Poisson’s equation, the repartition of the
particles to MPI-processes and the mesh refinement. The geometry Sc can
therefore be kept fixed."

Does it mean that there is no need to add adaptive domain function is amr
solver as the map already does it?

And in part '2.3.1  Implementation Overview' it says:

"A peculiarity of the uniform mesh PIC model in OPAL is the adjustment of the
grid Ω0 (cf. Figure 2.2) to the particle bunch."

In ParallelCyclotronTracker.cpp Line 3332 :
 itsBunch_m->boundp();
Does this create a new Ω0 containing all macroparticles?

 Best regard,
Sherlp