[Pw_forum] phonon with nspin=2 - not implemented ?

[Paolo Giannozzi]

On Dec 9, 2007, at 17:13 , Konstantin Kudin wrote:

> I ran the following PW/PH calculation for an insulating system:
>  [...]
> nspin=2
> multiplicity=1
> [...]
>  Then, the PH code crashed with the message:
> [...]
>      The phonon code with constrained magnetization is not yet  
> available
> [...]
>  Compared to other options already in PH, the one with nspin=2 for
> insulators seems rather less demanding. Is there any big reason why it
> is not implemented?

nothing big, but, as usual, there is a reason for such behavior.
"multiplicity=1" forces the system to have a definite magnetization
state. This is done by setting two Fermi energies, one for spin-up
and one for spin-down. This information is written in the data file
and if the phonon code notices it, it stops for the reason mentioned
in the error message.

> Or, can one use some kind of a workaround here?


If the magnetization you chose is the ground-state one, the two Fermi  
energies
will be the same (or they will be equivalent in the case of  
insulators). In that
case, commenting out the check in phq_readin.f90 should work. If you  
constrain
the magnetization to a value that is not the ground state, I have no  
idea what
numbers will come out

Paolo
---
Paolo Giannozzi, Dept of Physics, University of Udine
via delle Scienze 208, 33100 Udine, Italy
Phone +39-0432-558216, fax +39-0432-558222