[Pw_forum] Mass of the atomic for relax

Paolo Giannozzi giannozz at nest.sns.it
Mon Mar 17 18:41:43 CET 2008


Viet Bac wrote:

> I am confusing when using mass of the atomic species for relaxation. In 
> the Doc/INPUT_PW, it was said that:
>  
> Mass_X      REAL      : mass of the atomic species [amu: mass of C = 12]
>                                   not used if calculation='scf','nscf', 
> 'bands', 'phonon'
>  
> So for  VCS calculation, which is better when using this parameter?

I don't see anything confusing in the documentation: no dynamics,
no need for masses. For true MD, choose the real atomic masses.
For structural optimization with damped dynamics, you can set all
masses to the same number (quite arbitrary, since the only relevant
parameter is dt^2/mass): it is a form of "preconditioning".
Structural optimization with bfgs or similar algorithms will
ignore the masses. Or at least, this is what I remember.

Paolo
-- 
Paolo Giannozzi, Democritos and University of Udine, Italy



More information about the users mailing list