[lammps-users] freeze part of the system

hi all
I have a system with three types of particles that interact with the potential buck / coul / long.
I already realize the stabilization of the system, and got the restart file. What I would do is take this system and “freeze” two of the three types of particles.

My general idea is to restart the run, with the command read_restart, but modify some values.

I thought :
-to put a high value in the masses of the particles that I want to freeze:
mass 1 999999
mass 2 999999
mass 3 16

  • to put in command of the velocities
    velocity TYPE1 create 0.0 8779

velocity TYPE2 create 0.0 8779

velocity TYPE3 create 1000.0 8779 dist gaussian

  • to put in the fix of the integration of motion only the type of atom that can move
    fix 1 TYPE3 npt 1000.0 1000.0 100 xyz 1.0 1.0 20.0 drag 1.5

My question is if this is the best way ! I appreciates any opinion / suggestion.

Cristian

hi all
I have a system with three types of particles that interact with the
potential buck / coul / long.
I already realize the stabilization of the system, and got the restart file.
What I would do is take this system and "freeze" two of the three types of
particles.
My general idea is to restart the run, with the command read_restart, but
modify some values.
I thought :
-to put a high value in the masses of the particles that I want to freeze:
mass 1 999999
mass 2 999999
mass 3 16

- to put in command of the velocities
velocity TYPE1 create 0.0 8779

velocity TYPE2 create 0.0 8779

velocity TYPE3 create 1000.0 8779 dist gaussian
- to put in the fix of the integration of motion only the type of atom that
can move
fix 1 TYPE3 npt 1000.0 1000.0 100 xyz 1.0 1.0 20.0 drag 1.5

in most cases that i came across, this - the third option -
was the best way to go about it.

cheers,
   axel.

thanks for the suggestion axel.
Anyway, I could implement all three simultaneously?

best regards, Cristian

thanks for the suggestion axel.
Anyway, I could implement all three simultaneously?

if you do the third, the other two changes have no effect.
you'll have to make up your mind, though what to do with
the npt integration. if the box is expanded or shrunken, the
particle positions get typically rescaled with it. that may not
be what you want. have a close look a the docs on that
subject.

cheers,
   axel.

ok!
I’m thinking of avoiding this problem by using an ensemble NVT or NVE, no problem there! Thank you very much for the suggestion axel.