Re: [Phonopy-users] Precision of thermodynamics

[Masato Y. <yo...@am...>]

Dear Evgeny,

I guess I understand what kind of answer you wish to hear: I used to do 
so. But I'm afraid there is no general answer to your question. Somebody 
else might answer to your question based on his/her experiences, but it 
is different from material to material.

First of all, ab-initio calculation has finite numerical and physical 
accuracy whichever code or method you use as long as you conduct the 
calculation within practical time. In addition, physical constraints you 
impose to your model, such as, assumptions, cell size, or criteria of 
convergencies, for example, add up physical errors. These factors put 
your obtained value away from ideal value achievable by computation.

It should be noted that the ideal value is not necessarily equial to 
experimental value which might contain error.

As a rule of thumb, computational value will tell you how accurate 
experimental value would be. At the same time, experimental value will 
tell you how accurate computational value would be.

Thus, my best advise to you is, to put both computational and 
experimental values on your sides, and to try to improve each. As for 
computational side, there are a number of ways to improve numerical and 
physical accuracy, about which you may be asked in this mailing list, I 
suppose.

Good luck!

Masato

(2010/07/16 0:48), Evgeny Blokhin wrote:
> Good day! Please could You comment on the precision of the heat
> capacities and Helmholtz free energies calculated with Phonopy? How
> many digits after dot do you usually consider?
> Thanks,
> Evgeny
>
> ------------------------------------------------------------------------------
> This SF.net email is sponsored by Sprint
> What will you do first with EVO, the first 4G phone?
> Visit sprint.com/first -- http://p.sf.net/sfu/sprint-com-first
> _______________________________________________
> Phonopy-users mailing list
> Pho...@li...
> https://lists.sourceforge.net/lists/listinfo/phonopy-users