How can I realize Langevin Dynamics without inertia (Brownian Dynamics) by fix langevin or other command?

Dear LAMMPS users,

I am trying to realize overdamped Langevin Dynamics in my system. The Wikipedia told me that in the case of overdamped Langevin Dynamics, the mass would be ignored, and then the particle will not accelerate. In LAMMPS, I tried fix langevin, but based on the documentation, there is still the variable m ( the mass of the particle) in the Ff and Fr term. So I am not sure by fix langevin command, whether the particle can accelerate or not. If the particle still can accelerate, is there any other command can realize Brownian Dynamics and the particle will not accelerate? Thank you very much.

Best regards,

These is not. You’d have to write your own fix. You can get close to the overdamped limit by making the time constant of your fix langevin (much) smaller than all other relevant timescales.

Note that, even if you are interested in the overdamped limit, using Langevin dynamics is a much better idea from a numerical stability point of view, and the larger time step it allows might even give you better performance.

Pair style lubricateU coupled with pair style Brownian does accomplish this. However, there are additional terms that account for 2-body and averaged multibody interactions via zero Reynolds number hydrodynamics. If you’re just looking to supply some random perturbations to your system as Stefan suggests, you still might look at those as starting points.