Dear members,
I am learning to use vmd/lammps to set up argon flow around a cnt.
Below is the problem set up and questions. I would appreciate your assistance.
Problem setup:
Dear members,
I am learning to use vmd/lammps to set up argon flow around a cnt.
Below is the problem set up and questions. I would appreciate your
assistance.Problem setup:
---------------------3D- domain with periodic bc.
Liquid argon at 95K with density 1342 kg/m^3. CNT is (12,0) zig-zag.
(All geometry, atom data follows a published reference that I am trying to
reproduce)Created cnt in vmd and imported into lammps script as "atomic" style via
"read_data" file.
Created argon atoms inside lammps script using "random" style.Since "random" creates overlapping atoms, "minimization" is needed before
running the simulation.However, the "minimization" command destroys the cnt structure.
Looks like "minimization" works on all atoms in the domain since it does not
have
"group-ID" or "region-ID" handles to choose atom-groups.Question:
--------------
1. Is it possible to fix the cnt atoms and do minimization only on argon
atoms ?
yes. use fix setforce to set the forces on the CNT atoms to zero. no
forces -> no motion
2. If it is not possible, any other suggestion ?
it is probably easier to just delete overlapping atoms using the delete_atoms
-------------------------------------------------------------------------------------------------------------------------------
Other procedures I tried that did not work:
------------------------------------------------------------1. Instead of "random" atom style, I tried "Lattice" style. But I am not
able to get the
exact density by arbitrarily changing the lattice size. To reproduce the
published reference,
I used the exact number of argon atoms in a given domain size.
density of what? you can always change the density of your system
after the fact by adjusting the box dimensions.
So I changed to "random style". With this random style, I ran simple case
(without cnt) of a nvt equilibrium argon
simulation using "minimization" -- simulation gives appropriate bulk
pressure at a chosen temperature.
minimization doesn't create an equilibrated system, so that pressure
is a useless number.
2. I cannot do minimization on argon atoms and then import cnt with
"read_data" since simulation
domain can be created only once.
it is pointless to do things this way around. adding the CNT will
create overlaps anyway. but you can always merge multiple systems,
remove overlaps and *then* create a combined data file. e.g. using VMD
and the topotools plugin.
there are several other ways, too, which have been discussed on this
list fairly regularly. please have a closer look at the mailing list
archives (it has an effective search engine).
axel.
Dear Axel,
Thank your much for the quick assistance.
I will try your suggestions.
regards
shankar