Langevin thermostat?

LAMMPS LAMMPS Mailing List Mirror
1

Dear all,

I want to preform langevin thermostat for part of fluid atoms in the system.

  1. Did I use the lammps commond ‘fix langevin’ correctly?

fix 1 fluid langevin 300.0 300.0 100.0 48279
fix 2 fluid nve
run 1000000

2)Here, the number of atoms(N), the volume(V) and the temperature(T) controlled by langevin thermostat are constant. How about the pressure? Is it constant or fluctuate with simulation time?

Thanks,

Hangyan

2

2011/9/12 physics.hangyan.chen <[email protected]...>:

Dear all,

I want to preform langevin thermostat for part of fluid atoms in the system.

1) Did I use the lammps commond 'fix langevin' correctly?

fix 1 fluid langevin 300.0 300.0 100.0 48279

this appears to be syntactically correct. however,
whether it is meaningful is _impossible_ to say
without any context.

fix 2 fluid nve
run 1000000

2)Here, the number of atoms(N), the volume(V) and the temperature(T)
controlled by langevin thermostat are constant. How about the pressure? Is

atoms and volume are constant by construction. well,
unless either your dynamics is bad and you "lose" atoms
unintentionally or intentionally due to choice of system
and boundary condition.

temperature is not "constant" in the strict sense.
fix langevin will apply a friction term and random
kicks to your atoms to enhance dissipation of
kinetic energy and will do it in a way that it will
maintain the desired average temperature withing
the time constant chosen. the pressure will be
whatever the physics of your system dictates.

it constant or fluctuate with simulation time?

instantaneous pressure will fluctuate and the
magnitude of the fluctuations depends on the
compressibility of your system. for condensed
systems, those fluctuations tend to be large.
same as temperature fluctuations, their total
magnitude is system size dependent.

i strongly recommend reading up on these
subjects in a statistical mechanics text book.

cheers,
    axel.