[Pw_forum] dynamic coordinate study with CP

[Axel Kohlmeyer]

On Wed, Nov 7, 2012 at 7:10 PM, farid taherkhani
<faridtaherkhani at gmail.com>wrote:

>  *Hi
>
> Axel Kohlmeyer * on seven November wrote
>
>
>
> >how would the MD done with CP be different from other MD simulations?
>
> >when you talk about "pure" MD and what method would you use to
> >compute the forces (classical, semi-empirical, ab initio)?
>
> >are you aware of acceleration methods that preserve
> >dynamics published by art voter like parallel-replica MD.
>
> axel.
>
>
>
>
> In practice via pure MD  we can not obtain real structure due to time scale problem but  Quantum simulation method give us a real structure then we can follow our dynamics on the basis of Quantum proposed structure.
>
> sorry, but this is nonsense.
the statistical mechanics of the motions of the atoms
are in *no* way affected by how you compute the forces.

in fact the statistical sampling issues are even worse,
in CP, since your time step is much smaller and it takes
even longer to compute one time step.

a structure computed from a classical model is as
"real" as it is when you use DFT.

>  Can we use Quantum Espresso for obtaining one optimum structure then we turn on Molecular dynamics iteratively?
>
> cp.x *does* molecular dynamics. actually, computing an optimized structure
with DFT
and then continue with a classical model is a bad idea, since the
optimization takes
you to a 0K structure and hence you have a non-equilibrium configuration
when you
continue the the classical MD. better to just run only the classical MD and
get better
sampling.


>  We want to follow CP method for obtaining the exact dynamics our physical system in low temperature.Quantum Sutton-Chen potential for metallic interaction, covalent-bonding Tersoff potential between carbon-carbon in geraphen and Lenard-Jones potential for Carbon – metal. We know the method of parallel-replica MD which presented by professor Voter. However, we have limitation for replication technique in practice due to limitation of computer cpu node and its strategy in practice.
>
> using CP will make your computational restrictions even worse. what you
describe
sounds more like a QM/MM coupling method, which is an entirely different
animal
(or rather a different can of worms) and and even bigger headache.

best you stick to classical MD and save yourself a lot of pain and trouble.

axel.



> All the best,
>
>
> Farid
>
>
>
>
>
> --
> Dr. Farid Taherkhani
>
> Assistant Professor
>
> Department of Chemistry
>
> Razi University
>
> Kermanshah, Iran
>
> Tel: +98-831-4274569
>
> Email: f.taherkhani at razi.ac.ir
>
>
> _______________________________________________
> Pw_forum mailing list
> Pw_forum at pwscf.org
> http://pwscf.org/mailman/listinfo/pw_forum
>



-- 
Dr. Axel Kohlmeyer  akohlmey at gmail.com  http://goo.gl/1wk0
International Centre for Theoretical Physics, Trieste. Italy.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.quantum-espresso.org/pipermail/users/attachments/20121107/a987d7b7/attachment.html>