Use of Berendsen thermostat in Langevin Dynamics simulation

I am trying to simulate a system of 50 charged particles with Debye-Huckel interaction among them. I am using Langevin Dynamics method. At a very small value of friction coefficient the the system doesn’t thermalize to the desired value of temperature within reasonable computer time. Will it be correct to use a Berendsen thermostat in that case along with Langevin thermostat to bring the system to thermal equilibrium ?

I would say no, this would not be correct. Having 2 thermostats acting on the same atoms is certainly a bad idea. What you could try is to bring the system to the right temperature using a stronger coupling between your system and the bath. Once your system is at the right temperature, change the coupling and run a long simulation while ensuring that the temperature does not deviate too much from the desired value.

1 Like

Thanks for the reply. Can you comment on how much temperature deviation is allowed ? I mean, can I calculate a range of temperature variation over which I can consider the system to be reasonably in thermal equilibrium ?