# [lammps-users] Another setting up geometry question (con't)

Sorry, accidentally pushed send. Here is the message again:

Thank you all for your help with setting up my dislocation the other day. Now I am doing a different configuration (based on what some in the lit have done), and have a question. Basically, I am using a different coordinate system ([1 1 -2], [1 1 1], [1 -1 0]), and am confused how lammps maps this to xyz. Though I want to work in this coordinate system, I want the atoms to stay on normal bcc lattice positions. Currently, I am defining a lattice:

lattice custom 1.0 a1 .4082483 .4082483 -.8164966 a2 .5773503 .5773503 .5773503 a3 .7071068 -.7071068 0.0 basis 0.0 0.0 0.0

then region: region big block 0.0 1.0 0.0 1.0 0.0 1.0 units lattice

Just using 1.0 for lattice spacings to be clear what is going on... Then I use create_box with this region. I get:

Lattice spacing in x,y,z = 1.69271 1.69271 1.39385
Created orthogonal box = (0 0 0) to (1.69271 1.69271 1.39385)

This is a bit confusing to me... I can see the lattice spacings in each direction are just the sum of each of xyz components I defined in my lattice command, but this doesn't seem right. This would give a cubic volume of 3.00 Ang^3, but my orthogonal unit cube has unit length sides, and should have a volume of 1 Ang^3. Maybe I am just getting something wrong.

I also want to put atoms in traditional spots, so I was then saying 'lattice bcc 1.0', 'create_atoms' in the region, but this aligns the atoms properly with my rotated unit cube, not what is desired. This is a bit hard to explain, but any help would be appreiciated.

Erin

If you read the lattice doc page it carefully describes how the lattice
spacings are computed. They are just a scale factor, and do not
affect placement of atoms, just things like your box size. So you can
scale them however you wish when you create your box. I would use
a good viz tool and put some atoms down and see what you get.

Steve