回复: Control the pressure and exclude the effects ofother boxs at the same time

first off, please *always* reply to the mailing list, not individual
people only. people on the mailing list should see how this issue is
resolved, so that it can be looked up in the mailing list archives
later. thanks.

Dear Axel,

thank you for your help first. And your comment is very valuable.

I didn't put any medium into the box. In my box there's only one MWCNT at
first and the box is very big for the purpose of "isolated MWCNT". and I
just use the command like that:
"fix 1 outer npt temp 1 1 0.1 x 1.03 10.3 1 y 1.03 10.3 1"
and I want to reach the result of this article :
Phys. Rev. B 72, 035454 (2005) - Pressure-induced structural transition of double-walled carbon nanotubes .

i don't have the time to review literature for you.

Can this make sense? Or should I change my model to one with medium? Is that

no it cannot make sense. the pressure that you compute and that goes
into fix npt is the *global* pressure, i.e. the pressure of the entire
system. but what you model is a few objects in an empty space. so you
need something that makes the CNTs "feel" the interaction with the
solvent. this can be done differently, either with explicit solvent or
some coarse grain simplified solvent approximation, some implicit
solvent approach, or some hybrid continuum/atomic modeling method.
LAMMPS has several options to offer for that.

the only way to get my purpose?

as listed above, there are several options. you will have to research
which is most suitable. just saying that you can use fix npt to
"apply" a pressure is ignoring the basics of the statistical
mechanical ensemble you are modeling.

axel.