[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [ESPResSo-users] Langevin vs. Berendsen thermostat
From: |
Stefan Kesselheim |
Subject: |
Re: [ESPResSo-users] Langevin vs. Berendsen thermostat |
Date: |
Fri, 26 Apr 2013 11:17:21 +0200 |
Hi Koen,
I'm not sure what paper we mention but there is a very fundamental difference:
Using Langevin thermostat means that our particles follow the Langevin equation
of motion, which is basically Newton's plus a friction and a noise term on
every particle. Thus the dynamics become intrinsically stochastic. You find
more about the Langevin equation in most textbooks on non equilibrium stat mech.
With the Berendsen thermostat Newton's equation of motion are modified such
that in regular intervals the velocities are rescaled so that the total kinetic
energy = 3/2*n*k_B T.
Langevin is purely local (for Berendsen you need to sum up the kinetic energies
over all CPUs) and it is much more effective and stable in thermalising dilute
systems. See e.g.
http://en.wikipedia.org/wiki/Flying_ice_cube
This is why we do Langevin :-).
Cheers
Stefan
On Apr 26, 2013, at 11:09 AM, Koen Nickmans <address@hidden> wrote:
> Dear all,
>
> I'm sorry this is not so much Espresso related, but would somebody be so kind
> as to point a master student in the right direction as to what the difference
> is between the Langevin and Berendsen thermostats, if there is one? The
> Espresso user guide lists a Berendsen paper as a reference to the Langevin
> thermostat, hence my confusion.
>
> Thanks
> Koen Nickmans
>
> --
> The information contained in this message may be confidential and legally
> protected under applicable law. The message is intended solely for the
> addressee(s). If you are not the intended recipient, you are hereby notified
> that any use, forwarding, dissemination, or reproduction of this message is
> strictly prohibited and may be unlawful. If you are not the intended
> recipient, please contact the sender by return e-mail and destroy all copies
> of the original message.