PBC

Dear all,
I have a question related to periodic boundary condition. The system what I am studying is as follows: Liquids are confined between two rectangular plates and the upper plate is misaligned with respect to the lower one. I do not understand how to implement periodic boundary conditions along X and Y directions. Is it possible to do in LAMMPS? If you make some comments it would be helpful.
Best
Pritam

Dear Pritam,
I think this link will be helpful to you.
http://lammps.sandia.gov/doc/boundary.html

Also see this link:

http://lammps.sandia.gov/doc/read_data.html

It is a long web page so search for the words “triclinic” and “boundary”.

Although I mentioned triclinic boundary conditions, I don’t see why it would be necessary to use triclinic boundaries for your simulation. There is no reason that both plates need to lie entirely within the simulation boundary. You could probably use ordinary rectangular boundaries. (recommended unless there’s a good reason to do otherwise)

cheers
-andrew

Thank you Andrew. Is it possible to simulate when a part of the surface is outside the box? Because it will immediately show that atoms are outside the box error. Here is my system schematic. Or somehow I have to take the same area of upper and lower plate?
Thank you Keshab. I looked at them but somehow I can not connect with my problem.
Simulation_new.png

The picture helps.

I thought you just meant that one of the plates was shifted relative to the other. Now I see your problem. This is not a limitation of LAMMPS. Simulating the system you drew is harder because for most angles, your system is not periodic. You will have to approximate your system by a (preferably rectangular) periodic one, for example by making one of the plates slightly larger or smaller. (For example, by making the diagonal square larger by a factor of 1.5/sqrt(2))

I don’t know if my reply helped.

Cheers

Andrew

Simulation_new.png

Thank you very much. I imagine a problem if the upper plate is bigger as compared to lower plate. In that case, I have to put PBC according the size of the upper plate and in that case the liquid will go out of the confined space. Is it possible to use different xlo (lower boundary of X) and xhi (xlo (higher boundary of X) for the plate and the liquids?
Best
Pritam

Thank you very much. I imagine a problem if the upper plate is bigger as
compared to lower plate. In that case, I have to put PBC according the size
of the upper plate and in that case the liquid will go out of the confined
space. Is it possible to use different xlo (lower boundary of X) and xhi
(xlo (higher boundary of X) for the plate and the liquids?

​you are not making much sense here. *any* system with multiple periodic(!)
lattices can be represented with a supercell, regardless of differences in
lattice constant, rotation, or tilt.
the major issue is that you have to find the multipliers for each lattice
that result in commensurate supercell dimensions.

axel.

Thank you Axel. I agree with the point you mentioned.
I was using mica as surface and unit cell is constructed with several atoms.
So if I want to have the same occupation in the XY plane by both surfaces I have to delete the atoms from the lower plate (which is rotated by some angle) which are outside i.e., xhigh; yhigh of the upper plate where I made the rotated plate (here lower plate) bigger than the upper one. Then one can simply use the periodic boundary condition. Does it make sense?
Best
Pritam

Thank you Axel. I agree with the point you mentioned.
I was using mica as surface and unit cell is constructed with several
atoms.
So if I want to have the same occupation in the XY plane by both surfaces
I have to delete the atoms from the lower plate (which is rotated by some
angle) which are outside i.e., <xlow and >xhigh; <ylow and >yhigh of the
upper plate where I made the rotated plate (here lower plate) bigger than
the upper one. Then one can simply use the periodic boundary
condition. Does it make sense?

​no.​