[lammps-users] questions about box size and boundary conditions

Hi, I am a new users of Lammps, and I am working on MD simulation of CNTs. I am quite confused about the relations between box size and boundary conditions. How to decide the the box size (xlo, xhi, ylo, yhi, zlo, zhi) based on the dimensions of CNTs.

  1. for fixed boundary, the box size should be large than the dimensions of CNTs. How larger should it be?
  2. if I have PBC at the axial boundaries (z directions), what should zlo and zhi be if the CNTs is from -50 to 50 in z direction?

These questions should be really simple, so I cannot find any answers… Many thanks.

Please read the boundary doc page. Once you've figured
out what kind of boudaries you want (e.g. periodic or not),
then you can set the box size accordingly.

Steve

Hi, I am a new users of Lammps, and I am working on MD simulation of CNTs. I am quite confused about the relations between box size and boundary conditions. How to decide the the box size (xlo, xhi, ylo, yhi, zlo, zhi) based on the dimensions of CNTs.

  1. for fixed boundary, the box size should be large than the dimensions of CNTs. How larger should it be?

For fixed BC, just make sure the simulation box size is larger than the actual length of CNTs. However, if you do like tension, you’d need to make sure this during the entire tensile process.

  1. if I have PBC at the axial boundaries (z directions), what should zlo and zhi be if the CNTs is from -50 to 50 in z direction?

In this case, you’d know what the unit cell length of the CNTs is. Then you’d figure out what the real length of a repeated unit cell is by specifying how many units in your case. -50 to 50 is just a rough number.

Ajing

Hi, Thank you very much.
I think I have understood your explanation about the box size of PBC. The z dimension of box size is half cell length longer than the CNT at each end. Is it right?
I have another question about the energy minimization of CNT. I want to study the thermal conductivity of CNT. Before this, I perform the energy minimization first. But after the minimization, the length of CNT will change, so the box size will not be half cell length longer than the CNT at each end. How to address this problem?

Again, thank you for your kind reply.

Liang Chen

Hi, Thank you very much.
I think I have understood your explanation about the box size of PBC. The z
dimension of box size is half cell length longer than the CNT at each end.
Is it right?
I have another question about the energy minimization of CNT. I want to
study the thermal conductivity of CNT. Before this, I perform the energy
minimization first. But after the minimization, the length of CNT will
change, so the box size will not be half cell length longer than the CNT at
each end. How to address this problem?

there is no problem.

you can experiment with computing single energies (run 0)
for various cell dimensions and you should see.

axel.

Dear Axel,

Thank you for your explanation. I have tried cases with different cell dimensions (zlo and zhi). One case has a z dimenstion with half cell length longer than the CNT at each end. The other cases have random larger z dimension. Then I run 0. All the cases have the periodic boundary condition in z direction and free conditions in x and y direction. The results (potential energy, total energy and coordinates) after minimization are are quite different. I thought the box size for PBC should be carefully set. Is this all right?
BTW, what do you mean by cell dimension? Is it the box size (xlo xhi ylo yhi zlo zhi)? Many thanks.

Regards,
Liang Chen