[Pw_forum] something wrong in relax, BFGS

Carlo Sbraccia sbraccia at sissa.it
Mon Apr 18 23:44:38 CEST 2005


Dear xywu,

in order to understand what is happening I need the complete output file.

carlo

On Mon, 18 Apr 2005, xywu wrote:

>
>
> Dear All,
>
> There're something wrong in relax.
>
> The follows occurs:
>
> ---------------
>    number of scf cycles    = 117
>     number of bfgs steps    =  62
>
>     energy old              =    -161.4284071964 ryd
>     energy new              =    -161.4284071569 ryd
>
>     CASE: energy_new > energy_old
>
>     new trust radius        =       0.0000032530 bohr
>
>
>     WARNING :  something is going wrong
> ---------------------
> Then the scf cycle continues ,but the bfgs step does not.
>
>
> Please give me some advice.Thank you in advance!
>
> The input file is as follows:
>
>
> &CONTROL
>                       title = 'A--0.01-' ,
>                 calculation = 'relax' ,
>                restart_mode = 'from_scratch' ,
>                      outdir = '/root/tmp/' ,
>                  pseudo_dir = '/root/software/pwscf/espresso/pseudo/' ,
>                      prefix = 'A--0.01-' ,
>                       nstep = 150 ,
>                     tstress = .true. ,
>                     tprnfor = .true. ,
>                     tefield = .true. ,
> /
> &SYSTEM
>                       ibrav = 6,
>                   celldm(1) = 10.2,
>                   celldm(3) = 5,
>                         nat = 19,
>                        ntyp = 2,
>                     ecutwfc = 25 ,
>                     ecutrho = 120 ,
>                 occupations = 'smearing' ,
>                     degauss = 0.005 ,
>                    smearing = 'gaussian' ,
>                        edir = 3 ,
>                     emaxpos = 0.71 ,
>                      eopreg = 0.01 ,
>                        eamp = 0.01 ,
> /
> &ELECTRONS
>            electron_maxstep = 100,
>                    conv_thr = 1.0d-8 ,
>                 mixing_mode = 'plain' ,
>                 mixing_beta = 0.1 ,
>                 mixing_ndim = 16,
>             diagonalization = 'cg' ,
> /
> &IONS
>                ion_dynamics = 'bfgs' ,
>                     upscale = 10 ,
>     potential_extrapolation = 'wfc2' ,
> /
> ATOMIC_SPECIES
>   Si   28.08600  Si.pz-vbc.UPF
>    N   14.00670  N.pz-vbc.UPF
> ATOMIC_POSITIONS angstrom
>   Si      0.000000000    0.000000000    5.397600000
>   Si      2.698800000    2.698800000    5.397600000
>   Si      1.349400000    4.048200000    4.048200000
>   Si      4.048200000    1.349400000    4.048200000
>   Si      1.349400000    4.048200000    9.445800000
>   Si      4.048200000    1.349400000    9.445800000
>   Si      0.000000000    2.698800000    2.698800000    0  0  0
>   Si      2.698800000    0.000000000    2.698800000    0  0  0
>   Si      0.000000000    2.698800000    8.096400000
>   Si      2.698800000    0.000000000    8.096400000
>   Si      1.349400000    1.349400000    1.349400000    0  0  0
>   Si      4.048200000    4.048200000    1.349400000    0  0  0
>   Si      1.349400000    1.349400000    6.747000000
>   Si      4.048200000    4.048200000    6.747000000
>   Si      0.000000000    0.000000000    0.000000000    0  0  0
>   Si      2.698800000    2.698800000    0.000000000    0  0  0
>   Si      0.000000000    0.000000000   10.795200000
>   Si      2.698800000    2.698800000   10.795200000
>    N      2.698800000    2.698800000   16.795200000    1  1  0
> K_POINTS automatic
>  4 4 1   1 1 1
>
>
>
> xywu
> xywu at imr.ac.cn
>
>
> _______________________________________________
> Pw_forum mailing list
> Pw_forum at pwscf.org
> http://www.democritos.it/mailman/listinfo/pw_forum
>



More information about the Pw_forum mailing list