frist off, please *always* reply to the mailing list, not only
individual people. thanks.
Thank you for your reply.
In my example, a box with 's s s' boundary conditions has been defined, and
relaxed via the 'minimize command'. However, after such a relaxation the
pressure in x, y and z direction is not zero, in other words, Pxx, Pyy, Pzz
= ~0.8-1.0. As I known, the equilibrium state should be obtained under no
pressure conditions.
please note that these are likely to be relatively small numbers with
respect to your system size and choice of units.
but more importantly, if you have an "open" system, you do not have
something against which a pressure can form. you effectively put your
atoms into an infinite vacuum. because of that the pressure you obtain
is fairly meaningless and mostly indicative of some residual internal
interactions.
In other examples, for instance, free surface energy calculation, an
equilibrium state is required.
And, as you said, there is no point to relaxing a box with no-period
boundaries, but what commands can be used to get the equilibrium state of
these boxes having no periodicity?
equilibrium and zero pressure are two different things. i suggest you
review what an equilibrium state requires in your preferred text books
on statistical mechanics and MD simulations.
axel.
axel.