[lammps-users] one question about periodic condition

Dear LAMMPS users:

What I’m studying is to simulate the crystallization of Lennard-Jones system on one patterned substrate with LAMMPS. The boundary condition I set is P,P,F for X,Y,Z axises respectively. However, the region where L-J atoms crystallized is weird. Two snapshots are added to this mail. one grain crystal is divided into four parts in the view of one box but if periodic condition in x-y plan is applied, the four parts become one whole nucleus again. The size of the nucleus is smaller than the half value of the x-axis in box. Thus, the question confusing me is that is it one normal physical process for the L-J particles to form in this way? What role does the periodic condition play in this simulation?

I will be very grateful for your reply!

Warm regards!

Xiaodong Su

fig1-1.png

Dear LAMMPS users:

What I’m studying is to simulate the crystallization of Lennard-Jones system on one patterned substrate with LAMMPS. The boundary condition I set is P,P,F for X,Y,Z axises respectively. However, the region where L-J atoms crystallized is weird. Two snapshots are added to this mail. one grain crystal is divided into four parts in the view of one box but if periodic condition in x-y plan is applied, the four parts become one whole nucleus again. The size of the nucleus is smaller than the half value of the x-axis in box. Thus, the question confusing me is that is it one normal physical process for the L-J particles to form in this way? What role does the periodic condition play in this simulation?

this is not really a LAMMPS question, but a question about basics of MD simulation methodology. when you apply periodic boundary conditions, the origin of your simulation box and the entire view of the system becomes invariant to translation, i.e. the physics of the interactions doesn’t "know where a box starts or where it ends.
if you need a more detailed explanation, please discuss with your adviser/supervisor and consult the relevant MD textbooks.

axel.