overlapping atoms nanowire specific geometries?

Hi,

I am building a cylinder from within a box but am getting an error message with lost atoms that I assume is due to overlapping atoms. When I add the delete atoms it runs but the structure does not look like a perfect cylinder anymore. However, what is very odd to me is that certain geometries do not appear to have this overlap. For example, if I change my block to 0 50 0 50 0 80 and my cylinder to x 25 25 20 INF INF it will run fine? Also, if I change from box units back to lattice it works. Why are certain geometries causing overlap?

Thanks!

------------------------ INITIALIZATION ----------------------------

units metal
dimension 3
boundary p s s
atom_style atomic
variable latparam equal 2.856

----------------------- ATOM DEFINITION ----------------------------

lattice bcc {latparam} region whole block 0 100 0 100 0 120 units box create_box 1 whole region wire cylinder x 50 50 30 INF INF units box lattice bcc {latparam} orient x 1 0 0 orient y 0 1 0 orient z 0 0 1
create_atoms 1 region wire

mass 1 55.85

------------------------ FORCE FIELDS ------------------------------

pair_style eam/fs
pair_coeff * * Fe_mm.eam.fs Fe

#delete_atoms overlap 0.5 all all

------------------------- pressure calculations ---------------------------------

compute csym all centro/atom bcc
#compute peratomp all pe/atom

####################################### EQUILIBRATION

#set timestep
reset_timestep 0
timestep 0.001

#set temperature
velocity all create 300 12345 mom yes rot no
#fix 2 all npt temp 300 300 0.1 iso 0 0 1 drag 1
fix 2 all nvt temp 300 300 0.1

Set thermo output

thermo 500
thermo_style custom step lx ly lz press pxx pyy pzz pe temp

Run for at least 10 picosecond (assuming 1 fs timestep)

run 20000
unfix 2

You could be losing atoms in some cases simply
from the surface of a rough cylinder. Depending
on what atoms outside the cylinder you delete,
you might have low-coordinated atoms left.
There is no way to carve a cylinder out of a perfect
crystal and have a smooth circular surface.
If you minimize before dynamics you might get
a more stable surface. Or if you lose a few atoms
from the surface how is that different than having
removed them in the initial delete? I suggest you
visualize how you are losing atoms.

Steve