Dear developers:
     I am learning to relax structure.
     I use diamond as an example, to make only z component of one of the C atom free to move. 
     I set the input file as

&CONTROL
prefix='diamond_benchmark',
calculation='relax',
restart_mode='from_scratch',
wf_collect=.true.,
verbosity='high',
outdir='/mnt/e/DFT_code/QE/qe_tempdir',
pseudo_dir='/mnt/e/DFT_code/QE/pseudo',
etot_conv_thr=1.d-7,
forc_conv_thr=1.d-7,
/
&SYSTEM
ibrav=  2, celldm(1) =6.1, nat=  2, ntyp= 1,
    ecutwfc =40.0, 
/
&ELECTRONS
conv_thr = 1.0d-10,
/
&IONS
/
ATOMIC_SPECIES
 C  12.0  C.pz-vbc.UPF
ATOMIC_POSITIONS {crystal}
C  -0.125 -0.125 -0.130  0 0 1
C   0.125  0.125  0.125  0 0 0 
K_POINTS automatic
4 4 4 1 1 1


However, what I got after few cycles is that


     Forces acting on atoms (cartesian axes, Ry/au):

     atom    1 type  1   force =    -0.01898324    0.01898324    0.00000023
     atom    2 type  1   force =     0.01898324   -0.01898324   -0.00000023
     The non-local contrib.  to forces
     atom    1 type  1   force =    -0.02683533    0.02683533    0.00002953
     atom    2 type  1   force =     0.02684781   -0.02684781   -0.00002013
     The ionic contribution  to forces
     atom    1 type  1   force =    -0.03597500    0.03597500   -0.00054719
     atom    2 type  1   force =     0.03597500   -0.03597500    0.00054719
     The local contribution  to forces
     atom    1 type  1   force =     0.04383115   -0.04383115    0.00051988
     atom    2 type  1   force =    -0.04383552    0.04383552   -0.00052531
     The core correction contribution to forces
     atom    1 type  1   force =     0.00000000    0.00000000    0.00000000
     atom    2 type  1   force =     0.00000000    0.00000000    0.00000000
     The Hubbard contrib.    to forces
     atom    1 type  1   force =     0.00000000    0.00000000    0.00000000
     atom    2 type  1   force =     0.00000000    0.00000000    0.00000000
     The SCF correction term to forces
     atom    1 type  1   force =    -0.00000000    0.00000000    0.00000001
     atom    2 type  1   force =    -0.00000000    0.00000000    0.00000000

     Total force =     0.000000     Total SCF correction =     0.000000

     bfgs converged in   6 scf cycles and   3 bfgs steps
     (criteria: energy <  1.0E-07 Ry, force <  1.0E-07 Ry/Bohr)

     End of BFGS Geometry Optimization

     Final energy   =     -22.5816712783 Ry
Begin final coordinates

ATOMIC_POSITIONS (crystal)
C       -0.125076599  -0.125076599  -0.129923401    0   0   1
C        0.125000000   0.125000000   0.125000000    0   0   0
End final coordinates


There is two problems
1. I set only z component to move, why x and y changes?
2. the optimized result is not right

Thank you in advance for helping

best regards


_______________________________________________
Pw_forum mailing list
[email protected]
http://pwscf.org/mailman/listinfo/pw_forum

Reply via email to