[Pw_forum] ground state structure calculation -reg.,

Giuseppe Mattioli giuseppe.mattioli at ism.cnr.it
Wed Nov 6 14:23:15 CET 2013


Dear Suresh
The two methods are formally different, but they should lead to very close results. The "vc-relax" calculation is a minimization of the stress tensor 
(roughly: an equilibration of the internal vs external Pressure) carried on together with a minimization of forces within the cell. In the other case, 
you perform a bunch of "relax" calculations and fit the Energy vs Volume data by using some equation of state (often the Birch-Murnaghan you mention, 
but simpler paraboloid fits give often good results, specially in multi-dimensional problems like the present one) to find lattice parameter values 
which minimize the Energy (and the Pressure...). When you have a single lattice parameter (e.g., in a simple cubic cell), you may try to check that 
the two methods lead actually to negligibly different results. In the case of the rutile tetragonal cell the "vc-relax" option saves a lot of time, so 
long as the calculation is well converged (errors on stress are one order of magnitude higher than errors on forces, but I see that you have set 
already a VERY tight, perhaps too tight convergence criterion on wavefunctions).

Some further words. In the case of hybrid exx calculations, the vc-relax minimization is still not implemented. There is no explicit sign of exx in 
your input file, but I see the exxdiv_treatment flag, which is meaningful only in the case of hybrid calculations... Moreover, the wfc cutoff is horribly 
high. Are you sure you need 200/800 Ry? Finally, if you are really performing exx calculations, you should also check the convergence against the nqx
mesh in real space (see the PW manual and the EXX example).

HTH

Giuseppe


On Wednesday 06 November 2013 12:29:24 Suresh A wrote:
> I am new to QE. I have done variable cell volume optimization for rutile
> tio2.
> I confused with finding ground state structure.
> 
> In literature survey the authors did volume- constraint total energy
> minimization and have set cell parameters a, c and internal parameter(  u
> for rutile)
> finally fitted their value Birch - Murnaghan equation(B-M) . From the final
> value they have compared their values with exprimental values of cell
> parameter
> a,c and elastic parameters.
> 
> my question is:
> 
>  i) which method is correct whether doing doing vc-relax alone  or collect
> all volume and total energy from vc-relax output and fitting this with B-M
> equation ?
> 
>  ii) How to choose internal parameter u for TiO2 rutile.?
> i have included the input file.
> 
>  &CONTROL
>                        title = 'rutile' ,
>                  calculation = 'vc-relax' ,
>                 restart_mode = 'from_scratch' ,
>                       outdir = '/home/basheer/Desktop/relax/' ,
>                   pseudo_dir = '/home/basheer/espresso-5.0.1/pseudo/' ,
>                       prefix = 'rutile' ,
>                etot_conv_thr = 1.d-6 ,
>                forc_conv_thr = 1.d-4 ,
>                      tstress = .true. ,
>                      tprnfor = .true. ,
>  /
>  &SYSTEM
>                        ibrav = 6,
>                            A = 4.59373 ,
>                            B = 4.59373 ,
>                            C = 2.95812 ,
>                        cosAB = 0 ,
>                        cosAC = 0 ,
>                        cosBC = 0 ,
>                          nat = 6,
>                         ntyp = 2,
>                      ecutwfc = 200 ,
>                      ecutrho = 800 ,
>                  occupations = 'smearing' ,
>                      degauss = 0.01 ,
>                     smearing = 'gaussian' ,
>             exxdiv_treatment = 'gygi-baldereschi' ,
>  /
>  &ELECTRONS
>                     conv_thr = 1.d-12 ,
>  /
>  &IONS
>  /
>  &CELL
>                cell_dynamics = 'bfgs' ,
>               press_conv_thr = 2 ,
>  /
> ATOMIC_SPECIES
>    Ti   47.88800  Ti.pz-mt_fhi.UPF
>     O   15.99900  O.pz-mt_fhi.UPF
> ATOMIC_POSITIONS angstrom
>    Ti      0.000000000    0.000000000    0.000000000
>    Ti      2.296860000    2.296860000    1.479060000
>     O      1.402470000    1.402470000    0.000000000
>     O      3.191260000    3.191260000    0.000000000
>     O      0.894400000    3.699330000    1.479060000
>     O      3.699330000    0.894400000    1.479060000
> K_POINTS automatic
>   4 4 4   1 1 1
> 
> 
> 
> A.Suresh,
> Project Fellow,
> Madurai Kamaraj University,
> Madurai.


********************************************************
- Article premier - Les hommes naissent et demeurent
libres et ègaux en droits. Les distinctions sociales
ne peuvent être fondèes que sur l'utilitè commune
- Article 2 - Le but de toute association politique
est la conservation des droits naturels et 
imprescriptibles de l'homme. Ces droits sont la libertè,
la propriètè, la sùretè et la rèsistance à l'oppression.
********************************************************

   Giuseppe Mattioli                            
   CNR - ISTITUTO DI STRUTTURA DELLA MATERIA   
   v. Salaria Km 29,300 - C.P. 10                
   I 00015 - Monterotondo Stazione (RM)          
   Tel + 39 06 90672836 - Fax +39 06 90672316    
   E-mail: <giuseppe.mattioli at ism.cnr.it>
   ResearcherID: F-6308-2012




More information about the users mailing list