Re: About timestep again. -- Fastest motion O-H streching period 10 fs.

From: Peter Freddolino (petefred_at_ks.uiuc.edu)
Date: Mon May 08 2006 - 16:09:25 CDT

Hi Boyang,
you are correct that the general rule given is that to avoid having
atoms artificially travel through barriers or otherwise behave
unphysically, you ought to set a timestep at least one order of
magnitude below the fastest oscillation in the system; if you follow the
rule "at least" instead of "roughly" you get 1 fs timesteps, which is
what is customarily used for all atom MD. If you use rigidbonds, then a
2 fs timestep is appropriate because you have fixed these degrees of
freedom. If you're interested in more discussion on this, I highly
recommend reading up on MD timesteps in Leach's /Molecular Modelling:
Principles and Applications/ or a similar textbook. There's no guarantee
that you'll face catastrophe by using 2 fs timesteps in all atom MD, but
it is usually best to be safe as any errors that do appear may not be
obvious.

Peter

Boyang Wang wrote:

> Hi all. I am again asking something related to the timesteps in NAMD.
>
> As stated in many molecular dynamics methods, the time step should be
> about one order of magnitude smaller than the fastest motion in the
> system we simulate.
>
> In a typical chemical or biological system, the fastest motion, O ---
> H bond streching, or N -- H bond streching, has a period of about 12
> fs, according to the force constant of OT--HT bond in the Charmm force
> field.
>
> Therefore, it seems OK to me if I apply a timestep of 2 fs in the
> simulation of a chemical system.
>
> Could anybody who is familiar in this topic give me some hint? Thanks
> for your time.
>
> Sincerely,
>
> Boyang.
>
>
> Boyang Wang,
> Email - bwang9_at_uic.edu
> Major - Phys. Chem.
> dept Chemistry,
> Univ Illinois, Chicago,
> 60607 IL US.
>
> ------------------------------------------------------------------------
> ÑÅ»¢1GÃâ·ÑÓÊÏä°Ù·Ö°Ù·ÀÀ¬»øÐÅ <http://cn.mail.yahoo.com>
> ÑÅ»¢ÖúÊÖ-ËÑË÷¡¢É±¶¾¡¢·ÀɧÈÅ <http://cn.zs.yahoo.com>

This archive was generated by hypermail 2.1.6 : Wed Feb 29 2012 - 15:43:36 CST