(no subject)

Hi all,

Supposedly, I have non-periodic polymer system (boundary s s s). Now I want to deform it

Could you please tell me which command I can use to tether polymer to the wall such that when I move the wall, my tethered polymer chains deform together and are not slipped out wall in tensile progress.
I also wonder whether I can relax the system in 2 y, z directions when I only deform system in x direction e.g.
Thanks for your time
Ha.

Hi all,
Supposedly, I have non-periodic polymer system (boundary s s s). Now I want
to deform it
Could you please tell me which command I can use to tether polymer to the
wall such that when I move the wall, my tethered polymer chains deform
together and are not slipped out wall in tensile progress.

for a non-parallel system, the only want you can induce
deformation in a (to me) meaningful way, is to use fix move
on those atoms and not include them in the time integration.
similarly if you have atoms that you don't want to move, you can
also not include them in the time integration or set its
velocities and forces to zero.

I also wonder whether I can relax the system in 2 y, z directions when I
only deform system in x direction e.g.

sure. you are the master of your system, you can push
and pull at particles any which way you like.

axel.

All the variants of the fix wall command in LAMMPS are for
an idealized surface that particles interact with. There is no
way to "tether" a polymer chain to such a wall. Tethering implies
that you have a frozen monomer that the rest of the chain
is attached to via bonds. The typical way to do this is
to build a "wall" of frozen particles, which are your tethering
points. You can then move the frozen particles as you wish
via something like the fix move command. There is no
reason you couldn't use both a fix wall command, and also
have frozen tethering points (near or on the idealized wall)
that are moving in tandem, but they are 2 different things
from the LAMMPS perspective.

Steve

Dear Axel, Steve
Thanks for your suggestions.
In running process, however, I realized some problems that I may have misunderstood.

  1. I have frozen monomers in determined group (named rightgroup of system e.g.). Afterwards, I applied fix move to rightgroup to deform the system towards x.
    fix 1 rightgroup setforce 0.0 0.0 0.0
    velocity rightgroup set 0.0 0.0 0.0
    fix 2 rightgroup move linear ${vx} 0.0 0.0 in which vx is a constant
    According to doc page, we cannot use any of fix nve, nvt to relax non-periodic system once I used fix move
    Is there any ways to relax the system in y, z direction (apply time integration in y, z direction) for the rest of system (except rightgroup) such that the temperature is controlled
    Thanks a lot
    Ha.

I don't understand your question. There is no
problem using fix nve or fix nvt with non-periodic
systems. You can use them on a group of atoms
that doesn't include the atoms you are moving
via fix move, and thermostat the other atoms.

Steve

Dear Steve,
I’m sorry because of not making my question clear.
I have polymer system. I divide it into 2 parts: left (1) and right (2), say
Now, I move part (2) , whose monomers are frozen. As a result, temperature will be decreased suddenly.
How can I append temperature such that the system’s temperature is as many as its before deforming.
Special thanks for you patient helps
Ha.

Dear Steve,
I'm sorry because of not making my question clear.
I have polymer system. I divide it into 2 parts: left (1) and right (2), say
Now, I move part (2) , whose monomers are frozen. As a result, temperature
will be decreased suddenly.
How can I append temperature such that the system's temperature is as many
as its before deforming.

check out the commands "compute temp" and "thermo_modify"

axel.

And you can thermostat the non-frozen part however you
wish to maintain its temperature.

Steve

Yes, I also was thinking how to thermostat for y, z for non-frozen part.
In other words, I don’t want time integration in x direction for monomers in non-frozen part.
I used fix temp/rescale + compute temp/partial the same to shear examples, but just for rescalling temperature (velocity) in y, z directions
And is it possible time integration on y,z also.
Thanks so much
Ha.

See the doc page for the fix move command. You can move
in some dimesnsions and do NVE integration in others, by
using the NULL keyword.

Steve