simulation cell box size

Dear Lammps users,

Does anyone have a clue on how to decide on the appropriate box size for a rotated unit cell?

I am applying the rotation of

orient x 1 0 - 1 orient y 1 1 - 1 orient z 1 -2 1

to a periodic system containing Fcc-Fe with the lattice parameter of 3.65 Angstroms.

Finding the lattice spacing calculated by either LAMMPS or hand, is a piece of cake. When I, however, use a multiplicative of the lattice spacing as the box size, I still don’t get a correct arrangement of atoms and I can see lots of internal interfaces and over-lapping of atoms. Using the minimization of energy also proves that I don’t have a fcc at hand.

I would be appreciative if someone could give some advices.


– Hamid

You need to read a solid-state physics book on lattices and figure out
the periodic replication distance for your orientation. In 2d you can
generally draw it on a piece of paper. For orientaion [abc[, the inter-plane
distance is something like 1/(a^2 + b^2 + c^2), and you need to
figure out that every Nth plane is identical. But I’m not looking
at a text book …



Your periodic length for this particular orientation in X would be 2^1/2 similar Y would be 3^1/2 and Z would be 6^1/2.

The unit cell volume would now be 6a^3 , ‘a’ is lattice parameter.