[Pw_forum] vc-relax analysis doubt

[Ricardo Afonso]

Dear QE experts,

I have simulated NFO structure as it is shown in the input file bellow,
however after a few bfgs cycle (~2 or 3) my scf iteration reaches limit. I
don't understand what might be wrong. I have changed subroutine in
cell_base.f90 for 'z' direction like this:
-----------------------------------
            CASE ('z')
              iforceh      = 0
              iforceh(1,3) = 1
              iforceh(2,3) = 1
              iforceh(3,3) = 1
------------------------------------

This change is to leave the system to relax only in 'z' direction.

So could anyone tell me what may go wrong in this case? My pseudopotentials
are from pslibrary as well and they have worked for other cases. I want to
keep fixed positions of Fe in octahedrally oriented sites cause I must
compare with the case of other structure arragements, that's why I kept
their position fixed. I'm using "Intel Harwell" processors with 24 cores in
google cloud computing system in case this might be adressed as important
information.

input file:
--------------------------------------------------------------------------------------------
 &CONTROL
                 calculation = 'vc-relax' ,
                restart_mode = 'from_scratch' ,
                      outdir = '/home/drumsris/Bulk-74/' ,
                  pseudo_dir = '/home/drumsris/Bulk-227/' ,
                      prefix = 'nfo' ,
                     tstress = .true. ,
                     tprnfor = .true. ,
                       lorbm = .false.,
          wf_collect = .false.,
 /
 &SYSTEM
                       ibrav = 11,
               nosym = .true.
                   celldm(1) = 11.2679,
                   celldm(2) = 1.0,
                   celldm(3) = 1.0,
          space_group = 74,
            !nbnd = 100 ,
                         nat = 5,
                        ntyp = 4,
                     ecutwfc = 71 ,
                     ecutrho = 564 ,
                use_all_frac = .true. ,
                   input_dft = 'pbe' ,
                 occupations = 'smearing' ,
                     degauss = 0.0002 ,
                    smearing = 'methfessel-paxton' ,
                       nspin = 2 ,
            !noncolin = .true.,
            !lspinorb = .true.,
   starting_magnetization(1) = 1,
   starting_magnetization(2) = -1,
   starting_magnetization(3) = 1,
   starting_magnetization(4) = 0,
           !tot_magnetization = -1,
 /
 &ELECTRONS
            electron_maxstep = 100,
             diagonalization = 'david' ,
              diago_thr_init = 1.0e-5 ,
         mixing_beta = 0.6,
 /
 &IONS
        ion_dynamics = 'bfgs' ,
       pot_extrapolation = 'atomic' ,
                 upscale = 10D0 ,
               bfgs_ndim = 1 ,
        trust_radius_max = 0.8D0 ,
        trust_radius_min = 1.D-3 ,
        trust_radius_ini = 0.5D0 ,
                     w_1 = 0.01D0 ,
                     w_2 = 0.5D0
 /
 &CELL
           cell_dynamics = 'bfgs' ,
               press = 0.D0 ,
             cell_factor = 4.0 ,
          press_conv_thr = 0.5D0 ,
         cell_dofree = 'z' ,
/
ATOMIC_SPECIES
   Ni    58.69000  Ni.pbe-n-kjpaw_psl.1.0.0.UPF
   Fe1   55.85000  Fe.pbe-n-kjpaw_psl.1.0.0.UPF
   Fe2   55.85000  Fe.pbe-n-kjpaw_psl.1.0.0.UPF
    O    16.00000  O.pbe-n-kjpaw_psl.1.0.0.UPF
ATOMIC_POSITIONS crystal_sg
Fe2    0    0    0.5
Fe1    0    0.2500    0.125    0    0    0
Ni    0.25    0.25    0.75
O    0    0.00845    0.73971
O    0.23527    0.25000    0.50252
K_POINTS {automatic}
 5 5 5 0 0 0
--------------------------------------------------------------------------------------------
Best Regards,

Ricardo.

-- 
Ricardo Afonso
Student of Magnetism and Superconductivity Group
Federal University of Sao Carlos
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.quantum-espresso.org/pipermail/users/attachments/20170830/a21ac156/attachment.html>