[Pw_forum] convergence problem
H.Y. Geoffrey Tse
geoffrey.tse at pku.edu.cn
Fri Jun 27 13:14:13 CEST 2014
Providing the structure built is correct, then your energy cutoff set is too low.
I would suggest, go for something like ecutwfc = 80; ecutrho = 4*ecutwfc
Geoffrey
From: pw_forum-bounces at pwscf.org [mailto:pw_forum-bounces at pwscf.org] On Behalf Of Pallavi Bothra
Sent: 2014年6月27日 下午 06:09
To: pw_forum at pwscf.org
Subject: [Pw_forum] convergence problem
Dear all,
I am trying to do a scf calculation considering spin-orbit interaction. Here is the input file:
&CONTROL
calculation = 'scf',
restart_mode='from_scratch',
prefix='BiTe',
pseudo_dir = './',
outdir='./tempc',
nstep = 500
/
&system
ibrav= 0,
nat= 40,
ntyp= 2,
celldm(1) = 1.890359168,
ecutwfc = 30.0,
ecutrho= 300.0,
occupations='smearing',
degauss = 0.01 ,
smearing = 'cold' ,
starting_magnetization(1) = 0.5
starting_magnetization(2) = 0.5
noncolin = .true.
lspinorb = .true.
/
&ELECTRONS
electron_maxstep = 500 ,
conv_thr = 1.0d-10
mixing_beta = 0.1 ,
mixing_mode = 'local-TF',
mixing_ndim = 10 ,
/
&IONS
upscale = 100 ,
/
CELL_PARAMETERS
8.7900000 0.00000000 0.00000000
-4.3950000 7.6123633 0.00000000
0.00000000 0.00000000 36.64000000
ATOMIC_SPECIES
Te 127.6 Te.rel-pbe-dn-rrkjus_psl.0.2.2.UPF
Bi 208.980 Bi.pbe-d-mt.UPF
ATOMIC_POSITIONS angstrom
Te 0.001076915 -0.001111167 24.534365287
Bi 0.000000000 0.000000000 12.190002400 0 0 0
Te 2.197500000 1.268727220 10.146666670 0 0 0
Te -0.001767560 2.537644595 20.687693846
Te 0.000000000 0.000000000 6.383268000 0 0 0
Te 2.197537577 1.265453382 16.833926576
Te -0.000000000 2.537454430 13.910065330 0 0 0
----------------------------
K_POINTS automatic
4 4 1 0 0 0
Actually first I have relaxed the structure w/o considering spin-orbit interaction. Then I used the optimised coordinates to do the scf calculation but the system did not converge even after 300 cycles.
Thank you in advance
Regards
Pallavi Bothra
JNCASR, Bangalore
India
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.quantum-espresso.org/pipermail/users/attachments/20140627/bcc81f14/attachment.html>
More information about the users
mailing list