[Pw_forum] (no subject)

[Ari P Seitsonen]

Dear Osmair V Oliveira,

   Some quick comments/first reaction:

  - Are you sure that there is only one atom in the cell? If it is the hcp 
structure, there should be two, if I am not wrong

  - Did you check the convergence of the results with respect to the 
cut-off energy? This pp seems to contain the semi-core s and p states in 
the valence, so I would guess something like 150+ Ry might be needed

  - (I am not using the "cold" smearing, but maybe it is ok)

  - 10^-15 as the convergence threshould, nice if you indeed can converge 
the electronic structure so tightly, but seems to be somewhat exaggerated 
to me

  - The coordinates for the lonely atom seem to be in relative units but 
you have specified 'Angstrom'; more digits might be needed for the 
symmetry routines anyway ("Ti 1/3 2/3 1/4" would also be allowed, well 
then at least the elder versions of 'XCrysDen' were not able to read the 
line though)

  - No comments about the input for the phonon code (first please check the 
ones in the ground state calculation)

  - I would recommend to use an 'outdir' different than the current 
directory, to keep the directories cleaner

  - Please next time use a more appropriate title for your email (and a 
more complete affiliation if possible)

     Greetings from Paris,

        apsi

-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-
   Ari Paavo Seitsonen / Ari.P.Seitsonen at iki.fi / http://www.iki.fi/~apsi/
   Ecole Normale Supérieure (ENS), Département de Chimie, Paris
   Mobile (F) : +33 789 37 24 25    (CH) : +41 79 71 90 935


On Wed, 17 Sep 2014, Osmair Vital Osmair wrote:

> Dear users,I am trying to calculate critical temperature for Titanium. I used the following
> files:
> Ti.scf.in
>  &control
>     calculation='scf'
>     restart_mode='from_scratch',
>     prefix='Ti',
>     pseudo_dir = './',
>     outdir='./'
>  /
>  &system
>     ibrav =  4,
>     celldm(1) = 5.576203,
>     celldm(3) = 1.587874,
>     nat = 1,
>     ntyp = 1,
>     occupations = 'smearing',
>     degauss = 0.03,
>     smearing = 'cold',
>     ecutwfc = 60.0,
>     la2F = .true.,
>  /
>  &electrons
>      mixing_beta = 0.7,
>      conv_thr =  1.0d-15,
>      diagonalization = 'cg',
> /
> ATOMIC_SPECIES
>  Ti  47.88 Ti.pbe-sp-hgh.UPF
> ATOMIC_POSITIONS ANGSTROM
>  Ti 0.33333 0.66667 0.25000
> K_POINTS {automatic}
> 8 8 8 0 0 0
> 
> Question 1) The calculation is finished without error, but in the output file
> has the message  "No symmetry found
>           (note: 23 additional sym.ops. were found but ignored
>            their fractional translations are incommensurate with FFT grid)".
> This is very interesting because I tested various inputs with different elements with hexagonal structure (Zr, Tc, Re...) the same message appear. So, is there any
> problem or bug with the ibrav = 4 (hexagonal)?
> In addiction, in the ph.x calculation using the input below, I found negative frequencies, these frequencies can be attributed to this message?
> 
> Question 2)
> After the scf calculation, I performed ph.x using the file:
>  &inputph
>   tr2_ph=1.0d-16,
>   prefix='Ti',
>   fildvscf='aldv',
>   amass(1)=47.88,
>   outdir='./',
>   fildyn='Ti.dyn',
>   electron_phonon='interpolated',
>   trans=.true.,
>   ldisp=.true.
>   nq1=4, nq2=4, nq3=4
>  /
> In this calculation, the file Ti.dy0 has 36 points and it was generate
> 36 elph.* files. However, the kpoints.x generate 50 k-points:
> bravais lattice  >> 4
>      filout [mesh_k]  >> 
>      enter celldm(3)  >> 1.587874 
>      mesh: n1 n2 n3   >> 8 8 8
>      mesh: k1 k2 k3 (0 no shift, 1 shifted)  >> 0 0 0
>      write all k? [f] >> 
>      # of k-points   ==    50  of   512
> 
> If I use mesh: n1 n2 n3   >> 4 4 4, I have 12 k-points
> So, I expected that the number o k-points generated by the kpoints.x is the
> same of the obtained by the ph.x. Is this corrected and how can I improve this?
> 
> Thanks in advance.
> Osmair V. Oliveira
> Brazil
> 
> 
>