correct damping parameter


I had a question related to finding the correct damping parameter for a
given system when thermostating (specifically I am using fix langevin).
The damping should be related to the effective viscosity of whatever
material you are investigating. And it seems that a correct damping
parameter should exist for a given material at a given temperature. In the
discussion on fix viscous, the damping force F is given by F = - gamma *
velocity, where gamma is a kind of friction coefficient. It is also stated
that "the gamma of fix viscous is related to the damping parameter of fix
langevin, except that the units of gamma are force/velocity and the units
of damp are time, so that it can more easily be used as a thermostat."

My question is how specifically are the two related? Since force/velocity
units end up as mass/time, and the langevin damping is in time. Assuming I
can get an accurate gamma, I would like to get an accurate langevin damping

I could use a standard damping of 10, or 100, or something that just works.
However, I should be able to better model my intended system knowing the
actual damping parameter.

Any help is most appreciated, and if this doesnt make sense please let me

Thanks all.


Benjamin Warren Beeler
Mechanical and Aerospace Engineering
University of California, Davis

They are not only related, they are the same thing, that is drag force on an isolated particle in Stokes flow(the limit of zero Reynolds number or zero fluid inertia). It only depends on geometry, slip velocity between particle and bulk fluid, and viscosity of the fluid. It should not take long to find this in a google search. The only difference, as the documentation admits, is the addition of a random force in the Langevin model.

If you want to get a handle on this, I suggest looking into the fluctuation-dissipation theorem. Having a handle on the analytic approach for treating the Langevin eq. isn’t a bad skill to have either. Hope this helps.