Sorry for double posting, forgot to change the subject.
I wondered about ReaxFF memory management in LAMMPS. I know there are
problems and I've been following all the complaints/solution offers on
the mailing list, however, errors like 'not enough space for bonds''
or 'ran out of space on far_nbrs' are still very case dependent: for
example, my calculation runs on certain number of cpu-s and with
certain size of the system and with certain skin size and timestep,
otherwise giving the first error. One can find 'golden settings' with
trial and error but of course this limits the freedom of setting up
calculations. In my opinion at least the size of the system and the
number of cpu-s shouldn't matter. I've troubled with those a lot and
just wanted to ask if there is any work going on or planned in the
nearest future to fix these issues with reax/c.
Thanks in advance,