[lammps-users] Barostatting

Hi,

this is probably an easy question, but I couldn’t figure out the answer.
The lammps documentation says for the fix nvt/npt/nph commands at the very end:
“Non-periodic dimensions cannot be barostatted.”

Why is this?
I commented the corresponding lines in the constructor of fix_nh.cpp and started the simulation with non-periodic and shrink-wrapped BC. The simulation worked, but the results were useless, because the cell was squeezed too much even with small pressure.

My question: is there a physical reason for this limitation or is it just due to stability purposes?

Thanks for help.
Best, Manfred

Hi,

hi manfred,

this is probably an easy question, but I couldn't figure out the answer.
The lammps documentation says for the fix nvt/npt/nph commands at the very
end:
"Non-periodic dimensions cannot be barostatted."

Why is this?

because it doesn't make any sense.

npt/nph ensembles are supposed to represent a bulk system.

without periodic boundaries, you have a cluster

any external pressure on that cluster would have to be
generated by similar size objects and those you can
model explicitly (either through addtional atoms and
plus fix setforce or fix spring, or one of the fix wall options.
strictly speaking, in those cases only nve integration makes
sense, because all coupling of that cluster to its environment
would have to be explicit through particles.

I commented the corresponding lines in the constructor of fix_nh.cpp and
started the simulation with non-periodic and shrink-wrapped BC. The
simulation worked, but the results were useless, because the cell was
squeezed too much even with small pressure.

My question: is there a physical reason for this limitation or is it just
due to stability purposes?

there are very valid physical reasons. you have to carefully
consider what your simulation represents.

axel.