moving group in a periodic box

Hi again,

as someone else may meet similar problem, it is always nice to copy your message to lammps mail list. Another reason is that you may always receive someone else's answer who have more experience and knowledge than me.

26.04.2013, 23:45, "Mehdi Eftekhari" <[email protected]>:

Hi
thanks for your kind reply,
I know your argument was right, but I cant deform the box unless I use "fix npt" with zero pressure.

Have you tried using fix_deform or change_box with, for example, fix nve? Do try. You _can_ change box shape without fix npt, as Dr. Kohlmeyer already told you.

if you did such a simulation, can you give some advice or "a series of commands" that could do the simulation?
I want to deform matrix and CNT simultaneously, and grab the stress-strain curve.

I did not perform such a simulation myself, so the best recommendation I can give here is also searching through the mailing list. Concerning your approach, I personally think that, first, you do not need periodic boundary conditions in z direction at all, because you already simulate action of bulk atoms by applying force or displacement to your matrix atoms with fix. Second, I don't think you really need to relax x and y to zero pressure. At least you can try to keep them fixed and see how large will be the x and y pressure components. That being said, you can use periodic boundary conditions along x and y and, e.g., "s" boundary in z direction. And you can use fix nve or fix nvt for integration.

Hope this helps,
Oleg.