Pressure calculations

Dear Lammpsusers,

I am simulating a crystal by using periodic boundary conditions in x and y directions and free surface with a vacuum above and below in z direction. I am using NVT for equilibrating the crystal at 300 K. As there is vacuum above and below the crystal (in Z direction), I thought the crystal will expand (to achieve mechanical equilibration) and pressure will be zero during equilibration. But I get a very high value of pressure.

When I decompose the pressure into Pxx, Pyy, Pzz, I observe that Pzz flucutates around zero which is expected as there is vacuum above and below the crystal in z direction. But Pxx and Pyy are as high as 20000 bars when NVT ensemble is used.

. Does Pxx and Pyy really have meaning as I have used fixed boundary condition in z direction.

Thanks

NIkhil

Dear Lammpsusers,

I am simulating a crystal by using periodic boundary conditions in x and
y directions and free surface with a vacuum above and below in z direction.
I am using NVT for equilibrating the crystal at 300 K. As there is vacuum
above and below the crystal (in Z direction), I thought the crystal will
expand (to achieve mechanical equilibration) and pressure will be zero
during equilibration. But I get a very high value of pressure.

When I decompose the pressure into Pxx, Pyy, Pzz, I observe that Pzz
flucutates around zero which is expected as there is vacuum above and below
the crystal in z direction. But Pxx and Pyy are as high as 20000 bars when
NVT ensemble is used.

. Does Pxx and Pyy really have meaning as I have used fixed boundary
condition in z direction.

​yes, it has. please keep in mind, that crystals i.e. solids have in
general very low​ compressibility, so pressure fluctuations will be large
and even a small deviation from the equilibrated/relaxed box length in x or
y will result in a large pressure.

axel.

Thanks for your reply I still have few concerns:

As I have fixed boundary condition in z direction, I am wondering if we should really worry about Pressure values as given by Lammps. Because LAMMPS will use whatever volume is defined by the simulation cell boundaries i.e. lxlylz. I am wondering if the volume used by lammps under fixed boundary conditions is correct for pressure calculations. I understand that Lammps would accept the most general volume for pressure calulations that is the volume of simulation box, which makes sense for periodic boundaries. But does it make some sense for non-periodic boundary conditions?

I would be grateful if you can help me for the case of non periodic boundary conditions.
Thanks

Sent with Mailtrack

Thanks for your reply I still have few concerns:

As I have fixed boundary condition in z direction, I am wondering if we
should really worry about Pressure values as given by Lammps. Because LAMMPS
will use whatever volume is defined by the simulation cell boundaries i.e.
lx*ly*lz. I am wondering if the volume used by lammps under fixed boundary
conditions is correct for pressure calculations. I understand that Lammps
would accept the most general volume for pressure calulations that is the
volume of simulation box, which makes sense for periodic boundaries. But
does it make some sense for non-periodic boundary conditions?

I would be grateful if you can help me for the case of non
periodic boundary conditions.

​while total pressure and P_zz are affected, P_xx, P_yy and P_xy should not
(beyond finite size effects). it is embarrassingly simple to make some
tests to verify that.​