Re: Constraint failure in RATTLE algorithm - shouldn't a rerun fail at the same step?

From: Bjoern Olausson (namdlist_at_googlemail.com)
Date: Thu Sep 10 2009 - 05:26:00 CDT

On Wednesday 09 September 2009 17:40:29 Elia Zumot wrote:
> Hi Bjoern,
>
> I work with membrane proteins and run into the same problem when I'm using
> a 2 fs timestep (even though I set the rigidbonds to all). Have you tried
> running at 1 fs?
>
> Best
>
> Elia
>

This would have been my first try, but thought the crashes happen seldom and I
automated the restarting process, I will stick to 2fs.
A python script compares the current steps with the total expected steps and
restarts the simulation from the beginning if the current and expected number
of steps don't match. One run is just a round 12h so it does not hurt that
much if the simulations gets restarted here and then.

Thanks and greetings
Bjoern


This archive was generated by hypermail 2.1.6 : Wed Feb 29 2012 - 15:51:28 CST