Re: size of integration step

From: Axel Kohlmeyer (akohlmey_at_gmail.com)
Date: Tue Oct 13 2009 - 11:24:57 CDT

On Tue, Oct 13, 2009 at 11:44 AM, Jorgen Simonsen <jorgen589_at_gmail.com> wrote:
> Hi all,
> I am a little puzzled - is it reasonable to do a integration step of 1.5
> within using rigid bonds none - I have seen that most people using 2.0 fs
> are using the rigid bond all and at 1.0 fs people use rigid bond none. The
> frequency of hydrogen is 7-8 fs - so is it possible to use 1.5 without rigid
> bond constraints?

in my experience from ab initio simulations, you cannot use a timestep larger
than 0.5fs when you let hydrogens move and vibrate freely; a choice of 0.25fs
is a safe choice. however, if you use langevin dynamis with a
significant friction
coefficient and/or use weight redistribution (i.e. you make your
hydrogens heavier
and your heavy atoms correspondingly lighter), then you may be able to crank
up the time step. rule of the thumb: 2x mass => sqrt(2) x time step.

cheers,
   axel.
> Any comments and suggestions appreciated
> Thanks in advance
> Jorgen

-- 
Dr. Axel Kohlmeyer    akohlmey_at_gmail.com
Institute for Computational Molecular Science
College of Science and Technology
Temple University, Philadelphia PA, USA.

This archive was generated by hypermail 2.1.6 : Wed Feb 29 2012 - 15:51:34 CST