[Pw_forum] question

[Andrea Dal Corso]

On Wed, 2006-10-25 at 19:44 +0800, wang yanchao wrote:
> Dear all:
>     I have a question  when  I read the source program of PWscf.The
> route of the source program is PWscf/PW/kpoint_grid.f90.In it I find
> "if(t_rev(ns).eq.1) xkr=-xkr".I think "t_rev" is used to decide time
> revers.But  I write "t_rev" when I caculate example01 namely  Cu`s
> scf.The result is "0"rather than "1".I feel very confused. if it is
> "0",then "(t_rev.eq.1)"is false .This indicate that Cu`structure has
> no time revers. But in fact,  Cu`structure is fcc,it should have time
> revers.So I do not know it clearly.Could you please tell me why ?Thank
> you!
> 
> 

The reason is that t_rev is set and used only with magnetic spin-orbit
calculations. In this case there might be symmetry operations which
require multiplication by the time reversal operator.
In all other cases t_rev is set to zero, but time reversal is 
implicitly used because the code takes only k or -k, not both.

Andrea   



>                                                                   
> 
> 
> sincerly 
> 
> Yanchao Wang
-- 
Andrea Dal Corso                    Tel. 0039-040-3787428
SISSA, Via Beirut 2/4               Fax. 0039-040-3787528
34014 Trieste (Italy)               e-mail: dalcorso at sissa.it