I have two cells each cubed shaped and 40A per side. They were created
using NPT ensemble and the final atom positions depend on coordinates and image flags. One cell contains a long chain polymer and the other a nitrogen gas. The purpose is to simulate a free surface in the z direction as a “poor man’s” non-periodic boundary.
If one stacks the two cells to create a new cell for a new simulation, do the image flags for both cells need to be recalculated since the new cell length, in the z direction, is now twice as long as it was?
I’m sure there are better ways to do such a follow-on simulation, but I’m attempting to have a better understanding of the image flags and how to build up simulation geometries from previous simulation results.
Thanks in advance,
-bruce
I have two cells each cubed shaped and 40A per side. They were created
using NPT ensemble and the final atom positions depend on coordinates and
image flags. One cell contains a long chain polymer and the other a nitrogen
gas. The purpose is to simulate a free surface in the z direction as a "poor
man's" non-periodic boundary.
If one stacks the two cells to create a new cell for a new simulation, do
the image flags for both cells need to be recalculated since the new cell
length, in the z direction, is now twice as long as it was?
that depends on what kind of tool you are using to merge the cells.
e.g. if you use read_data twice, you only need to shift the cell (in z
by 40 \AA) and offset the types.
if you use VMD's topotools plugin, coordinates will be unwrapped upon
reading, then you can displace system and then merge the cells and
output the combined system with image flags again.
I'm sure there are better ways to do such a follow-on simulation, but I'm
attempting to have a better understanding of the image flags and how to
build up simulation geometries from previous simulation results.
in general the "unwrap-displace-merge-wrap" sequence is what i would
recommend to follow.
axel.