[Pw_forum] questions on the acoustic sum rule crystal

Nicola Marzari marzari at MIT.EDU
Sun Mar 9 16:08:28 CET 2008

Eduardo Ariel Menendez Proupin wrote:
> Hi,
> Question 1. This question is about the Acoustic Sum Rule in phonon 
> calculations. Is there a reference to learn about the option 
> zasr='crystal'?
> Is 'crystal' always better than 'simple' ?

I think "crystal" was implemented by Nicolas Mounet - a MSc student here
in our group - and was based on the idea that the IFCs are a vector in
a multi-dimensional space, and you should take the projection of these 
onto the hypersurface where the ASR are satisfied. I'd think it's
better, in that it imposed simultaneously all constraints in the
least disruptive way, if oyu are happy with the cartesian metrics
for the distance; it might take a lot of memory or time in really large

It's described in his thesis - I'd be happy to send it. We never 
published it, but maybe it's time - a recent PRB by Mingo also discusses 
it, and introduces another alternative based on Lagrangian multipliers.
> Question 2. Consider the calculation of the phonon DOS, as in example 
> 06. The ASR can be imposed in the run of q2r and/or in the run of 
> matdyn, let's say
> There are four combinations of zasr-asr in q2r an matdyn
>         q2r          matdyn
>  1       no         no
>  2       no         crystal
>  3      crystal    no
>  4      crystal    crystal
> Should  there be  differences in the final DOS in the combinations 2, 3, 
> and 4?

I'd like to know myself -but last time I looked into the q2r and
maydyn was 16 years ago :-( . Maybe you could report any differences
(rather than in the DOS, I'd look at low energy acoustic phonons, both
at Gamma but especially close to gamma, but not really there).


Prof Nicola Marzari   Department of Materials Science and Engineering
13-5066   MIT   77 Massachusetts Avenue   Cambridge MA 02139-4307 USA
tel 617.4522758 fax 2586534 marzari at mit.edu http://quasiamore.mit.edu

More information about the users mailing list