[gmx-users] T-coupling/freeze groups

[Rainer Boeckmann]

>Check a few things please:
>1. Are you using a cut-off? Then you will get large temperature increase
>  if you switch off t coupling (also without freezing atoms)

PME

>2. Have you defined center-of-mass motion removal group for the frozen
>  atoms? Otherwise they will still move.

I tried that - but with com-group specified to the freeze atom group I get NAN
for the temperature...

>3. Is you time step too long (particularly since you run at 1000 bar)?

0.5fs - I think that should do it.


>Have you discussed this with Bert de Groot?
>When using pressure coupling you could get some nasty effects
>with freeze groups as the restrained particles are not scaled,
>while the rest of the system is scaled. This was discussed about
>a month ago in this list (with Bert)

yes. I know about that discussion. Thought or better hoped that there might be a
solution now....

--
Rainer Böckmann
Theoretical Molecular Biophysics Group
Max-Planck-Institute for Biophysical Chemistry
Am Faßberg 11
D-37077 Göttingen

Tel.: +49(551)201-1141, Fax.: +49(551)201-1089
email: rboeckm at gwdg.de