[Pw_forum] Convergence issues

[Yantao Wu]

Dear QE,

I ran a vc-relax with the following specifications (16 atoms, 6 kinds (4
Al, 2Ti, 2Cr, 2 Fe, 3 Co, 3 Ni) :

&CONTROL
  calculation = "vc-relax",
  dt          = 30.D0,
/
&SYSTEM
  ibrav       = 1,
  A           = 5.9726,
  nat         = 16,
  ntyp        = 6,
  ecutwfc     = 12.D0,
  occupations = "smearing",
  smearing    = "methfessel-paxton",
  degauss     = 0.05D0,
/
&ELECTRONS
  conv_thr    = 1.D-7,
  mixing_beta = 0.7D0,
/
&IONS
  ion_dynamics      = "damp",
  pot_extrapolation = "second_order",
  wfc_extrapolation = "second_order",
/
&CELL
  cell_dynamics = 'bfgs'
/
ATOMIC_SPECIES
Al  27  Al.pbe-n-rrkjus_psl.0.1.UPF
Ti  48  Ti.pbe-sp-van_ak.UPF
Cr  52  Cr.pbe-sp-van.UPF
Fe  56  Fe.pbe-nd-rrkjus.UPF
Co  59  Co.pbe-n-van_gipaw.UPF
Ni  59  Ni.pbe-nd-rrkjus.UPF
....
K_POINTS automatic
8 8 8 0 0 0


These are some of the total energies in the first scf step:

....
-1167.37380454
-1167.37730282
-1167.37313462
-1167.32387998
-1167.36729923
-1167.36931830
-1167.36943509
-1167.37686298
-1167.37121766
-1167.36631842
-1167.36484431
-1167.31794482
-1167.35485093
-1167.36884054
-1167.36678740
-1167.36964606
-1167.35381941
-1167.29375341
-1167.31307315
-1167.30808057

It took 55 steps but still showed sign of convergence. I ended the
calculation prematurely at this point, because it was taking too long. I'm
just wondering if this is normal for such a system. Or something is going
wrong, like &SYSTEM parameters.

Thank you very much,
Yantao  Wu
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.quantum-espresso.org/pipermail/users/attachments/20130528/2289f505/attachment.html>