Temperature using the brownian dynamics

Hello all,

I am probably misunderstanding something.

I used fix langevin and its working fine and the temperature is fixed as the command states in the manual.

But I tried to use fix Brownian and the temperature (from thermo) is very large, a factor of 400!

Is there a specific way to calculate or fix the temperature when using fix Brownian?

It seems to be connected to timestep, can someone please clarify it or let me know which topic I should look into?


code used:


units lj
dimension 2
atom_style atomic
pair_style none #lj/cut 2.5
boundary p p p

System definition

region myreg block -100 100 -100 100 -0.5 0.5
create_box 2 myreg
create_atoms 1 random 2000 341341 myreg

Simulation settings

mass * 1

neigh_modify every 1 delay 1 check yes
dump mydmp all atom 1000 dump.lammpstrj


thermo 10
minimize 1.0e-4 1.0e-6 1000 10000
fix mynve all nve
#fix mylgv all langevin 1.0 1.0 0.1 1530917
fix 3 all brownian 1 12908410 gamma_t 1
fix myefn all enforce2d

timestep 0.00005
thermo 10000
run 100000

For an authoritative answer you have to contact the author of this fix.
Please note that fix brownian includes time integration and thus does not need an additional fix nve for that. LAMMPS is printing a warning about that.
Please also note that there are examples for using fix brownian provided by the author in the examples tree of the LAMMPS distribution.

Thank you very much, it helped.

Hi @coldzer,
I am encountering the same problem even with their scripts given in the example folder.
The temperature has a factor of 10ˆ6.
Could you explain to me how did you resolve it ?

In the overdamped dynamics limit you cannot talk anymore about velocities and kinetic energies because they loose their physical meaning. Then it makes no sense to talk about temperature because temperature are derived from kinetic energy.