[Pw_forum] RE: Pw_forum digest, Vol 1 #673 - 6 msgs

[Paolo Giannozzi]

On Tuesday 28 June 2005 17:40, sun jason wrote:

> how to write the input file for dynmat.x?

you don't have to write anything: program "ph.x" writes the
dynamical matrix into the file specified in variable "fildyn". See
example09 and the header of file pwtools/dynmat.f90:

!
!  read a dynamical matrix at q=0 , diagonalise it, calculate IR
!  and Raman cross sections (if Z* and Raman tensor available) 
!
!  Input data (namelist "input")
!
!  fildyn  character input file containing the dynamical matrix
!                    (default: fildyn='matdyn')
!  q(3)      real    calculate LO modes (add nonanalytic terms) along
!                    the direction q (default: q=(0,0,0) )
!  amass(nt) real    mass for atom type nt, a.u.
!                    (default: amass is read from file fildyn)
!  asr     logical   impose Acoustic Sum Rule (default:asr=.true.)
!  filout  character output file containing frequencies and modes
!                    (default: filout='dynmat.out')
!  filmol  character as above, in a format suitable for 'molden'
!                    (default: filmol='molden.out') 

> what's wrong with my input file?

what error message are you getting? in any event, you will get zero
forces and thus no optimization, for symmetry reasons, in a zincblende
lattice. The following is for sure wrong:

>                        cosAB = 90 ,
>                        cosAC = 90 ,
>                        cosBC = 90 ,

cosAB = "cosine of the angle between axis A and B" and so forth

Paolo
-- 
Paolo Giannozzi             e-mail:  giannozz at nest.sns.it
Scuola Normale Superiore    Phone:   +39/050-509876, Fax:-563513 
Piazza dei Cavalieri 7      I-56126 Pisa, Italy