Damp parameter of fix langevin || Viscosity requiring a different fix

Hello all,

I have been trying to incorporate langevin dynamics into my system, however I am really confused about the exact role of the damping parameter in the fix langevin command.

I have come across many systems where they claim to have damping in the range of 100-200. However when I set damp to 100 in fix langevin, the temperature of the system starts to blow up. The thermostatitng only works effectively at much lower damping values such as 1. I have not seen my system exhibit any bad dynamics as I have also run nvt simulations on the same system and the thermostatting was effectively able to maintain a temperature of 1.

In order to set the viscous drag experienced by the particle proportional to its velocity, does fix langevin require a prior setting of the viscosity through the fix viscous command, since the damping parameter is only meant to be used for controlling thermostatting?


1 Like

This means that either your system is not yet equilibrated or that your choice of time step is too large for your system.

You can run for a while with the shorter damping and then switch to a longer damping parameter for a “production” run. For the latter you should also do a test running without fix langevin and it should preserve the temperature and total energy, or else there are issues with your force field or time step.

Please consult your MD text book for more information about equilibration and energy conservation and choosing a proper time step.

Thanks for your help!

I would like just to seek your clarification on the same topic
Suppose I have two fixes for viscous and langevin as follows:

fix 1 all viscous eta
fix 2 all nve
fix 3 all langevin 1.0 1.0 damp 88694

Now if the mass of my particle is m, would it imply that the effective drag coefficient in the langevin equation would be (m/damp+eta) ?

The two fixes are independent and know nothing about each other, thus the net effect is the sum of what those two fixes do.

Thanks for clarifying!