Dear LAMMPS users,
I've been studying grain boundaries' lowest energy configuration through LAMMPS,
but I've found few errors concerning on it. I'm not sure whether these
problems comes from my script error, or inherent bugs in lattice
command in LAMMPS.
<scripts>
variable Cu equal 3.615
lattice fcc ${Cu}
variable Xsize equal (16^2+1)^(0.5)*5
variable Zsize equal 3
variable Ysize equal 20
variable Ymsize equal -${Ysize}
region Rbox prism 0 \{Xsize\} {Ymsize} \{Ysize\} 0 {Zsize} 0.0 0.0 0.0
region Rugrain prism 0 \{Xsize\} 0 {Ysize} 0 \{Zsize\} 0\.0 0\.0 0\.0
region Rdgrain prism 0 {Xsize} \{Ymsize\} 0 0 {Zsize} 0.0 0.0 0.0
create_box 2 Rbox
lattice fcc 3.615 orient x 16 1 0 orient y -1 16 0 orient z 0 0 1
create_atoms 1 region Rugrain
lattice fcc 3.615 orient x 16 -1 0 orient y 1 16 0 orient z 0 0 1
create_atoms 2 region Rdgrain
group Gugrain region Rugrain
group Gdgrain region Rdgrain
region Rinterface block INF INF -10 10 INF INF units box
group Ginterface region Rinterface
problems are:
1. Unexpected voids are found in the boundary of a cell. To be detail,
this voids are not observable Inside the cell, but at the very
boundary between the cell and its translated image by x direction.
After few days of struggling, I've found that this problem is
sometimes solved through changing number of processors (from 16 to 1),
but on other times, it is solved by changing the x-direction size of
the cell.
(I'm running multiple simulations by changing interface misorientation
angle. in the above script, it is 7.15 deg. So, the other time means,
literally, another misorientation angle)
I cannot grasp any consistency to solve this problem.
2. While the number of atoms in the interface region has to be
constant w/ changing y-direction size of the cell, (as shown in the
script) i've found it does not.
for instance, when Ysize = 21, number of atoms in Ginterface = 1053,
even though when Ysize = 20, number of atoms in Ginterface = 1044.
it is very crucial issue in grain boundary energy because slight
change can change overall configuration of a interface.
it anyone have idea on these problems, please leave any comments.
thanks in advance.
Hayoung Chung.