[DFTB-Plus-User] DFTB-Plus-User Digest, Vol 40, Issue 5

Ben Hourahine benjamin.hourahine at strath.ac.uk
Fri Apr 22 09:48:09 CEST 2011


On 22/04/11 04:50, bigcontinent wrote:
> Dear Ben
> Thank you very much ! I've solved the problem. But, I want to konw if
> different values of CouplingStrength will greatly change the results
> of final MD simulation.
> Best regareds
> Xin
Hello Xin,

the short answer is yes, for some calculated properties of the system.

A longer answer, is that the coupling constant for Berendsen thermostats
controls how strongly the temperature
will be forced towards the target temperature, and this has two effects.
Firstly the rates of temperature change
increases with strength of the coupling, but secondly it effects the
magnitude of fluctuations in the temperature
of the system. The physically correct situation (canonical ensemble) is
that the temperature of a finite collection
of atoms should fluctuate around the average temperature, with the
magnitude of the fluctuations being a
function of the number of atoms. Berendsen thermostats don't really
produce these fluctuations in a physical
way, but can have fluctuations for small coupling constants (so weak
control of the temperature).

If the property you are calculating depends on these fluctuations, then
the coupling strength will influence this.
Without knowing why you are running the simulation, and what data you
want to obtain its difficult
to be more specific. Many people use a thermostat to reach the desired
temperature and then restart the simulation
with constant energy (no thermostatting) when actually calculating
properties.

Regards

Ben

-- 
     Dr. B. Hourahine, SUPA, Department of Physics,
   University of Strathclyde, John Anderson Building,
          107 Rottenrow, Glasgow G4 0NG, UK.
    +44 141 548 2325, benjamin.hourahine at strath.ac.uk
   The University of Strathclyde is a charitable body,
registered in Scotland, with registration number SC015263

-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://www.dftb-plus.info/pipermail/dftb-plus-user/attachments/20110422/0d842211/attachment.htm 


More information about the DFTB-Plus-User mailing list