[Pw_forum] something wrong D_2h

[Dal Corso Andrea]

Which version of QE are you using? 
If it is an old one, please update. On QE 4.2 your input seems to work
(at least pw.x with verbosity='high' finds the correct group and
representations).

Andrea


On Thu, 2010-05-27 at 12:36 +0200, Ion Errea Lope wrote:
> Dear all,
> 
> I am calculating with espresso-4.1 the dynamical matrix of a supercell
> containing 32 atoms of Ca. The scf calculation runs without problems,
> but when computing the dynamical matrix of that system with ph.x the
> program stops with the following error message:
> 
> %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
> %%%%%%
>      from divide_class : error #         1
>      something wrong D_2h
>  %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
> %%%%%%%
> 
> It seems there is a problem with the irreps emerging from the
> subroutines divide_class.f90 and divide_class_so.f90 that I do not
> understand well. Looking into the forum archive I found a similar
> problem discussed in 2007 that was solved modifying the subroutines
> mentioned above. Here is the link to that post: 
> 
> http://www.democritos.it/pipermail/pw_forum/2007-August/007054.html
> 
> I wonder whether a similar modification of the subroutines may solve
> this issue or the problem is different. I would be very grateful if
> someone could help me with this issue.
> 
> Below I include the input files of the pw.x and ph.x calculations:
> 
> -> pw.x 
> 
> &control
>   calculation='scf'
>   prefix='ca'
>   pseudo_dir='/scratch/ionerrea/pseudo'
> outdir='/lscratch/ionerrea/results-supercell_7_pol2-lscratch'                           
> /
> &system
>   ibrav = 8
>   celldm(1) =19.371149508
>   celldm(2) = 1.000000000
>   celldm(3) = 0.500000000
>   nat = 32
>   ntyp = 1
>   nbnd =  416
>   ecutwfc = 30.0
>   ecutrho = 300.0
>   occupations='smearing'
>   smearing='marzari-vanderbilt'
>   degauss=0.02
> /
> &electrons
>   conv_thr = 1.0e-8
>   mixing_beta = 0.7
> /
> ATOMIC_SPECIES
>   Ca  40.078  Ca.pbe-nsp-van.UPF
> ATOMIC_POSITIONS {crystal}
> Ca    0.000000000    0.000000000    0.000000000
> Ca    0.000000000    0.000000000    0.500000000
> Ca    0.003535534    0.253535534    0.000000000
> Ca   -0.003535534    0.246464466    0.500000000
> Ca    0.000000000    0.500000000    0.000000000
> Ca    0.000000000    0.500000000    0.500000000
> Ca   -0.003535534    0.746464466    0.000000000
> Ca    0.003535534    0.753535534    0.500000000
> Ca    0.253535534    0.003535534    0.000000000
> Ca    0.246464466   -0.003535534    0.500000000
> Ca    0.250000000    0.250000000    0.000000000
> Ca    0.250000000    0.250000000    0.500000000
> Ca    0.246464466    0.496464466    0.000000000
> Ca    0.253535534    0.503535534    0.500000000
> Ca    0.250000000    0.750000000    0.000000000
> Ca    0.250000000    0.750000000    0.500000000
> Ca    0.500000000    0.000000000    0.000000000
> Ca    0.500000000    0.000000000    0.500000000
> Ca    0.496464466    0.246464466    0.000000000
> Ca    0.503535534    0.253535534    0.500000000
> Ca    0.500000000    0.500000000    0.000000000
> Ca    0.500000000    0.500000000    0.500000000
> Ca    0.503535534    0.753535534    0.000000000
> Ca    0.496464466    0.746464466    0.500000000
> Ca    0.746464466   -0.003535534    0.000000000
> Ca    0.753535534    0.003535534    0.500000000
> Ca    0.750000000    0.250000000    0.000000000
> Ca    0.750000000    0.250000000    0.500000000
> Ca    0.753535534    0.503535534    0.000000000
> Ca    0.746464466    0.496464466    0.500000000
> Ca    0.750000000    0.750000000    0.000000000
> Ca    0.750000000    0.750000000    0.500000000
> K_POINTS {automatic}
>    4   4   8       1 1 1
>   
> -> ph.x
> 
> &inputph
>  tr2_ph=1.0d-16
>  prefix='ca'
>  amass(1)=40.078
>  outdir='/lscratch/ionerrea/results-supercell_7_pol2-lscratch
> '
>  fildyn='dyn_B_plus_0.02_1'
>  alpha_mix(1)=0.4
>  lnscf=.true.
> /
>      0.00000     0.00000     0.00000
> 
> Thank you very much for your help.
> 
> Kind regards,
> 
> Ion
> 
> 
-- 
Andrea Dal Corso                    Tel. 0039-040-3787428
SISSA, Via Beirut 2/4               Fax. 0039-040-3787528
34151 Trieste (Italy)               e-mail: dalcorso at sissa.it