[Pw_forum] grid phonon calculation at non-gamma point cannot reach convergence
Isaiah Moses
imoses87 at gmail.com
Mon Oct 2 14:15:51 CEST 2017
Hi Yuan,
Your structure seems big, many nbnd.
For such structure, smaller value of alpha_mix(1) in phonon input is
helpful for convergence.
Non inclusion of alpha_mix(1) in your input file implies that the default
value of 0.7 would be used.
Include alpha_mix(1) and use smaller value such as 0.2.
If that doesn't solve the problem, you may need to increase the convergence
threshold tr2_ph.
It is also helpful to split your job along the q-points. This will enable
you monitor the convergence of different q-point and adjust parameters such
as alpha_mix(1) and tr2_ph accordingly.
Best,
Isaiah
On Mon, Oct 2, 2017 at 12:51 PM, yuan liang <yl4ps at virginia.edu> wrote:
> Dear QE developers and users,
>
> Recently I'm trying to calculate phonon dos via ph.x, however, although I
> can get fine convergence result at gamma point, ph.x cannot reach
> convergence within 100 steps and stop automaticly at the 1st non gamma
> point on my mesh.
>
> The material is YVO3, orthorhombic, with 4 Y atom, 4 V atom and 12 O atom
> per unit cell.
>
> The QE version is 5.4.0
>
> I tried both GGA PBEsol USPP and LDA USPP(from THEOS SSSP)
>
> Are there any methods to solve this problem? Any suggestion will be welcome.
>
> Here are my input files:
> *scf calculation*
> &control
> calculation= 'scf',
> restart_mode= 'from_scratch',
> prefix='YVO3_exc',
> pseudo_dir = '...',
> outdir= '...',
> wf_collect=.true.,
> tprnfor=.true., tstress=.true.
> /
> &system
> ibrav = 8,
> celldm(1) = 9.9893,
> celldm(2) = 1.0581,
> celldm(3) = 1.4296,
> nat = 20,
> ntyp = 3,
> ecutwfc = 60,
> ecutrho = 480,
> ecfixed = 60,
> qcutz = 20,
> q2sigma = 5,
> occupations= 'smearing', smearing='gauss', degauss=0.01,
> /
> &electrons
> mixing_beta = 0.4,
> mixing_mode = 'plain',
> conv_thr = 1.0d-10,
> /
> ATOMIC_SPECIES
> Y 88.906 y_lda_v1.4.uspp.F.UPF
> V 50.942 v_lda_v1.4.uspp.F.UPF
> O 15.999 o_lda_v1.2.uspp.F.UPF
> ATOMIC_POSITIONS crystal
> Y 0.97830 0.07150 0.25000
> ...
> O 0.18720 0.20040 0.55663
> K_POINTS automatic
> 6 6 4 0 0 0
> *phonon calculation on grid*
> &inputph
> prefix='YVO3_exc',
> outdir='...',
> recover=.true.,
> *ldisp=.true*.,
> nq1=6,
> nq2=6,
> nq3=4,
> tr2_ph=1.0d-14,
> fildyn='YVO3_dyn.G',
> amass(1)=80.906,
> amass(2)=50.942,
> amass(3)=15.999,
> /
> 0.000000 0.000000 0.000000
>
> Best regards,
>
> Yuan Liang
> Physics Department,
> University of Virginia
>
> _______________________________________________
> Pw_forum mailing list
> Pw_forum at pwscf.org
> http://pwscf.org/mailman/listinfo/pw_forum
>
--
Isaiah Abu Moses
Graduate Student,
Physics Department,
University of Ibadan,
Nigeria
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.quantum-espresso.org/pipermail/users/attachments/20171002/2c1b3f08/attachment.html>
More information about the users
mailing list