Re: RescaleFreq of temperature =1?

From: Axel Kohlmeyer (akohlmey_at_cmm.chem.upenn.edu)
Date: Thu Jul 23 2009 - 11:00:13 CDT

On Thu, 2009-07-23 at 19:48 +0400, Mikhail Suyetin wrote:
> >> Hi all.
> >> I'am developing biological system and I need to keep T=309,75 (36.6
> >> degrees centigrade) strongly.
> >
> > why??? and particularly, why so exact?
>
> Cause, 36.6 degrees centigrade are normal human body temperature

that is fine. but that is only the average over a very, very
large system. on the atomic level you will have the same kind
of thermal fluctuations that you would see in an MD simulation.
what you have to achieve is that your system is _on average_
at this temperature.

> > it is a very unrealistic setup. temperature is not a well defined
> > property for a system with a finite number of particles. we usually
> > assume equipartitioning. rescaling in every step will artificially
> > slow down all processes that would happen due to the natural
> > fluctuations.
> Wow, it's really so seriously.

just have a look at introductory text books on MD simulations,
e.g. frenkel/smit, understanding molecular simulations.

> I hope, If I take the value of rescaleFreq equals to 10, 25 or 50,
> I'll be able to keep the temperature value.

this is _very_ frequent and would have almost the same effect.
actually, you should not need to rescale at all unless your system
is far from equilibrium. perhaps you should also look into the
allen and tildesley book that explains MD more from the technical
point of view. once your system is in equilibrium, you should not
need a thermostat at all to maintain the same temperature, if it
was not for rounding errors due to the imprecision of floating
point math. attaching a thermostat algorithm is a much better choice.
that will modify the atomic velocities or accelerations in a way
as if your "small" system is coupled to a large head bath (e.g.
the body).

cheers,
   axel.

> --
> Sincerely yours,
> Mikhail Suyetin

-- 
=======================================================================
Axel Kohlmeyer   akohlmey_at_cmm.chem.upenn.edu   http://www.cmm.upenn.edu
   Center for Molecular Modeling   --   University of Pennsylvania
Department of Chemistry, 231 S.34th Street, Philadelphia, PA 19104-6323
tel: 1-215-898-1582,  fax: 1-215-573-6233,  office-tel: 1-215-898-5425
=======================================================================
If you make something idiot-proof, the universe creates a better idiot.

This archive was generated by hypermail 2.1.6 : Wed Feb 29 2012 - 05:22:15 CST