My Community

Hi

I am trying to understand how to use Langevin thermostat. I am particularly concerned about LANGEVIN_GAMMA parameter as defined in https://www.vasp.at/wiki/index.php/LANGEVIN_GAMMA.

Here, the practical example fixes some atoms and make others Langevin atoms. Does this setup should be followed such that one have to assign friction coefficient to certain atoms only? In this regard, I guess the Langevin atoms act as bath.

Best Regards,
Burak

In most calculations you want to expose all atoms to a thermostat. So in that case set LANGEVIN_GAMMA for all atoms the same. In my experience a value of 5-10 works fine for most materials.
The wiki example is a special case where you want to thermostat only a few atoms. It was chosen to show that also this case works.