Dear lammps users,
I have to do a long simulation which I’ll have to chop into several consecutive jobs.
I also need to output the trajectory with unwrapped coordinates:
dump_modify 0 unwrap yes
Consecutive jobs will read the restart file from the previous job. However, I read in the documentation (read_restart.html) that write_restart will wrap the atomic positions onto the primitive cell, so that later read_restart will read the atomic positions with no trouble. If this is the case, will the unwrapped positions from the previous simulation be lost when I start the following simulation? How then can I obtain an unwrapped trajectory for the whole simulation?
Thanks in advance,
Jose Borreguero
Dear lammps users,
I have to do a long simulation which I'll have to chop into several
consecutive jobs.
I also need to output the trajectory with unwrapped coordinates:
dump_modify 0 unwrap yes
Consecutive jobs will read the restart file from the previous job. However,
I read in the documentation (read_restart.html) that write_restart will wrap
the atomic positions onto the primitive cell, so that later read_restart
will read the atomic positions with no trouble. If this is the case, will
the unwrapped positions from the previous simulation be lost when I start
the following simulation? How then can I obtain an unwrapped trajectory for
the whole simulation?
well, first of all. this kind of question *begs* for making a simple
test and verifying what is going to happen rather than asking.
that being said, you misinterpreted the manual. yes, coordinates will
be wrapped back, but that does happen anyway at every rebuild of the
neighbor list. the unwrapped coordinates are reconstructed from the
wrapped coordinates and their corresponding image flags. those image
flags are updated during reneighboring and are also stored into
restart files, so there should be no worries about losing the unwrap
information.
HTH,
axel.