[Pw_forum] the dynmaical matrix are like "***********" when the phonons calculation at gama

xu yuehua njuxuyuehua at gmail.com
Tue Mar 3 13:28:05 CET 2009


hi :
thank  Eyvaz Isaev for your hints. but unfortunately, the problem seems
more complicated when i use different thr2_ph,from1.d-16 to 1.d-18
the out file wrote:


  PHONON       : 51m46.02s CPU time,    51m50.38s wall time


     Alpha used in Ewald sum =   2.8000

     Electric Fields Calculation

      iter #   1 total cpu time :  3827.9 secs   av.it.:   5.2
      thresh= 0.100E-01 alpha_mix =  0.700 |ddv_scf|^2 =  0.130E-01
     kpoint   1 ibnd  33 solve_e: root not converged  NaN
     kpoint   1 ibnd  33 solve_e: root not converged  NaN
     kpoint   1 ibnd  33 solve_e: root not converged  NaN
     kpoint   2 ibnd  33 solve_e: root not converged  NaN
     kpoint   2 ibnd  33 solve_e: root not converged  NaN
     kpoint   2 ibnd  33 solve_e: root not converged  NaN
     kpoint   3 ibnd  33 solve_e: root not converged  NaN
     kpoint   3 ibnd  33 solve_e: root not converged  NaN
     kpoint   3 ibnd  33 solve_e: root not converged  NaN
     kpoint   4 ibnd  33 solve_e: root not converged  NaN
     kpoint   4 ibnd  33 solve_e: root not converged  NaN
     kpoint   4 ibnd  33 solve_e: root not converged  NaN
     kpoint   5 ibnd  33 solve_e: root not converged  NaN
     kpoint   5 ibnd  33 solve_e: root not converged  NaN
     kpoint   5 ibnd  33 solve_e: root not converged  NaN
     kpoint   6 ibnd  33 solve_e: root not converged  NaN
     kpoint   6 ibnd  33 solve_e: root not converged  NaN
     kpoint   6 ibnd  33 solve_e: root not converged  NaN
     kpoint   7 ibnd  33 solve_e: root not converged  NaN
     kpoint   7 ibnd  33 solve_e: root not converged  NaN
     kpoint   7 ibnd  33 solve_e: root not converged  NaN


so i think the thr2_ph=1.d-15 is ok,not so tight to cause not converged.

but my problem  still exists: the ******** in Dynamical  Matrix  is unphysical.
for it causes that the phonons at G is :

   Mode symmetry, C_2  (2)    point group:

     omega(  1 -  1) = ************  [cm-1]   --> B
     omega(  2 -  2) = ************  [cm-1]   --> A
     omega(  3 -  3) = ************  [cm-1]   --> A
     omega(  4 -  4) = ************  [cm-1]   -->   ?
     omega(  4 -  4) = ************  [cm-1]   -->   0A
     omega(  4 -  4) = ************  [cm-1]   -->   ?
     omega(  4 -  4) = ************  [cm-1]   --> B
     omega(  5 -  5) = ************  [cm-1]   -->   ?
     omega(  5 -  5) = ************  [cm-1]   -->   0A
     omega(  5 -  5) = ************  [cm-1]   -->   ?
     omega(  5 -  5) = ************  [cm-1]   --> B
     omega(  6 -  6) = ************  [cm-1]   -->   ?
     omega(  6 -  6) = ************  [cm-1]   -->   0A
     omega(  6 -  6) = ************  [cm-1]   -->   ?
     omega(  6 -  6) = ************  [cm-1]   --> B

it is noted that the first several phonon's frequnecies should not so
large,they are shoud around at zero cm-1.

when is the problem, i think over it,but no answer.
i need your help.
my input file you can find below.
thank you very much if you could give me some hints!



2009/3/2, Eyvaz Isaev <eyvaz_isaev at yahoo.com>:
> Hi,
>
> As the dynamical matrix is written in the fixed format (like 3(2F12,8, 2x)),
> and you have large components (say, larger than 999.99999999, or
> -99.99999999 ), then it can not be printed properly. In my opinion, such
> kind components will be quite unusual.
> I have no experience in calculation of H2O properties, but I would recommend
> the use of thr2_ph=1.d-18, at least, use different thr2_ph and see what
> happens.
>
> If you are really interested in these values then change the format in line
> 32, write_dyn_on_file.f90 (in /PH, of course), say, to f16.8, then recompile
> ph.x, and enjoy seeing these values.
>
> Bests,
> Eyvaz.
>
> -------------------------------------------------------------------
> Prof. Eyvaz Isaev,
> Theoretical Physics Department, Moscow State Institute of Steel & Alloys,
> Russia,
> Department of Physics, Chemistry, and Biology (IFM), Linkoping University,
> Sweden
> Condensed Matter Theory Group, Uppsala University, Sweden
> Eyvaz.Isaev at fysik.uu.se, isaev at ifm.liu.se, eyvaz_isaev at yahoo.com
>
>
> --- On Sun, 3/1/09, xu yuehua <njuxuyuehua at gmail.com> wrote:
>
>> From: xu yuehua <njuxuyuehua at gmail.com>
>> Subject: [Pw_forum] the dynmaical matrix are like "***********" when the
>> phonons calculation at gama
>> To: "PWSCF Forum" <pw_forum at pwscf.org>
>> Date: Sunday, March 1, 2009, 9:45 AM
>> hello all :
>>
>>
>> i am facing unkonwn problem when i calculated the phonons
>> at gama:
>>
>> the  dynamical matrix  file  wrote:
>> ................
>>  Dynamical  Matrix in cartesian axes
>>
>>      q = (    0.000000000   0.000000000   0.000000000 )
>>
>>   1  1
>> ************  0.00000000  ************  0.00000000
>> ************  0.00000000
>> ************  0.00000000  ************  0.00000000
>> ************  0.00000000
>> ************  0.00000000  ************  0.00000000
>> ************  0.00000000
>>   1  2
>> ************  0.00000000  ************  0.00000000
>> ************  0.00000000
>> ************  0.00000000  ************  0.00000000
>> ************  0.00000000
>> ************  0.00000000  ************  0.00000000
>> ************  0.00000000
>>   1  3
>> ************  0.00000000  ************  0.00000000
>> ************  0.00000000
>> ************  0.00000000  ************  0.00000000
>> ************  0.00000000
>> ************  0.00000000  ************  0.00000000
>> ************  0.00000000
>>   1  4
>> ************  0.00000000  ************  0.00000000
>> ************  0.00000000
>> ************  0.00000000  ************  0.00000000
>> ************  0.00000000
>> ************  0.00000000  ************  0.00000000
>> ************  0.00000000
>>   1  5
>> ************  0.00000000  ************  0.00000000
>> ************  0.00000000
>> ************  0.00000000  ************  0.00000000
>> ************  0.00000000
>> ************  0.00000000  ************  0.00000000
>> ************  0.00000000
>>   1  6
>> ************  0.00000000  ************  0.00000000
>> ************  0.00000000
>> ************  0.00000000  ************  0.00000000
>> ************  0.00000000
>> ************  0.00000000  ************  0.00000000
>> ************  0.00000000
>>   1  7
>> ************  0.00000000  ************  0.00000000
>> ************  0.00000000
>> ************  0.00000000  ************  0.00000000
>> ************  0.00000000
>> ************  0.00000000  ************  0.00000000
>> ************  0.00000000
>>   1  8
>> ************  0.00000000  ************  0.00000000
>> ************  0.00000000
>> ************  0.00000000  ************  0.00000000
>> ************  0.00000000
>> ************  0.00000000  ************  0.00000000
>> ************  0.00000000
>>   1  9
>> ************  0.00000000  ************  0.00000000
>> ************  0.00000000
>> .......
>>
>> my phone input file :
>> phonons at Gamma
>>  &inputph
>>   tr2_ph=1.0d-15,
>>   prefix='scf_two_b_type',
>>   epsil=.true.,
>>   amass(1)=1.0,
>>   amass(2)=15.999,
>>   outdir='/disk2/xxx/xxx/espresso-3.2.3/tmp',
>>   fildyn='fourringwater.dynG',
>>  /
>> 0.0 0.0 0.0
>>
>>
>> and my scf input file is :&CONTROL
>> calculation  = "scf",
>>   prefix       = "scf_two_b_type"
>>   pseudo_dir   =
>> "/disk2/xxx/xxx/espresso-3.2.3/pseudo",
>>   outdir       =
>> "/disk2/jdong/xyh/espresso-3.2.3/tmp",
>> /
>> &SYSTEM
>>   ibrav     = 4,
>>   a=20,b=20,c=5.60,cosab=-0.5,cosac=0,cosbc=0,
>>   nat       = 24,
>>   ntyp      = 2,
>>   ecutwfc   =35 ,
>>   ecutrho   =420
>> /
>> &ELECTRONS
>>   conv_thr    = 1.D-9,
>>   mixing_beta = 0.2D0,
>> /
>>  ATOMIC_SPECIES
>> H  1.0  H.pbe-rrkjus.UPF
>> O 15.999 O.pbe-rrkjus.UPF
>> ATOMIC_POSITIONS { crystal }
>> H        0.040592899  -0.047115761   0.485459003
>> H        0.114704004  -0.001940711   0.646569902
>> H        0.077728017   0.073838408   0.485209132
>> H        0.068286628   0.133313508   0.646050213
>> H       -0.040592899   0.047115761   0.485459003
>> H       -0.114704004   0.001940711   0.646569902
>> H       -0.077728017  -0.073838408   0.485209132
>> H       -0.068286628  -0.133313508   0.646050213
>> H        0.087725644   0.047173264   0.985851842
>> H        0.116757198   0.001980553   1.146554001
>> H       -0.003964894   0.073814692   0.985604262
>> H        0.065042531   0.133450775   1.145919070
>> H       -0.087725644  -0.047173264   0.985851842
>> H       -0.116757198  -0.001980553   1.146554001
>> H        0.003964894  -0.073814692   0.985604262
>> H       -0.065042531  -0.133450775   1.145919070
>> O        0.095295444  -0.003121003   0.483499312
>> O        0.058352838   0.111476844   0.483127474
>> O       -0.095295444   0.003121003   0.483499312
>> O       -0.058352838  -0.111476844   0.483127474
>> O        0.098444160   0.003183717   0.983638903
>> O        0.053053024   0.111436979   0.983231173
>> O       -0.098444160  -0.003183717   0.983638903
>> O       -0.053053024  -0.111436979   0.983231173
>>
>>
>>
>>
>>
>>
>>
>> K_POINTS {automatic}
>> 1 1 12 0 0 0
>>
>>
>> and my relax out file  wrote:
>>
>>
>>      convergence has been achieved
>>
>>      Forces acting on atoms (Ry/au):
>>
>>      atom   1 type  1   force =    -0.00000712
>> -0.00004363    0.00001631
>>      atom   2 type  1   force =     0.00004994
>> 0.00001718   -0.00006583
>>      atom   3 type  1   force =     0.00001729
>> -0.00001573    0.00002699
>>      atom   4 type  1   force =    -0.00000656
>> 0.00003062    0.00005285
>>      atom   5 type  1   force =     0.00000712
>> 0.00004363    0.00001631
>>      atom   6 type  1   force =    -0.00004994
>> -0.00001718   -0.00006583
>>      atom   7 type  1   force =    -0.00001729
>> 0.00001573    0.00002699
>>      atom   8 type  1   force =     0.00000656
>> -0.00003062    0.00005285
>>      atom   9 type  1   force =    -0.00000156
>> -0.00006167   -0.00005767
>>      atom  10 type  1   force =     0.00001092
>> 0.00004788   -0.00001249
>>      atom  11 type  1   force =     0.00005214
>> -0.00008085   -0.00005460
>>      atom  12 type  1   force =    -0.00002474
>> -0.00007499    0.00003358
>>      atom  13 type  1   force =     0.00000156
>> 0.00006167   -0.00005767
>>      atom  14 type  1   force =    -0.00001092
>> -0.00004788   -0.00001249
>>      atom  15 type  1   force =    -0.00005214
>> 0.00008085   -0.00005460
>>      atom  16 type  1   force =     0.00002474
>> 0.00007499    0.00003358
>>      atom  17 type  2   force =     0.00001005
>> 0.00004653    0.00008656
>>      atom  18 type  2   force =     0.00001760
>> -0.00007784    0.00003639
>>      atom  19 type  2   force =    -0.00001005
>> -0.00004653    0.00008656
>>      atom  20 type  2   force =    -0.00001760
>> 0.00007784    0.00003639
>>      atom  21 type  2   force =     0.00002956
>> -0.00007076   -0.00004942
>>      atom  22 type  2   force =    -0.00000699
>> 0.00005606   -0.00001264
>>      atom  23 type  2   force =    -0.00002956
>> 0.00007076   -0.00004942
>>      atom  24 type  2   force =     0.00000699
>> -0.00005606   -0.00001264
>>
>>      Total force =     0.000382     Total SCF correction =
>>    0.000010
>>
>>
>>      entering subroutine stress ...
>>
>>           total   stress  (Ry/bohr**3)
>> (kbar)     P=   -0.40
>>   -0.00000194   0.00000001   0.00000000         -0.29
>> 0.00      0.00
>>    0.00000001  -0.00000179   0.00000000          0.00
>> -0.26      0.00
>>    0.00000000   0.00000000  -0.00000434          0.00
>> 0.00     -0.64
>>
>>
>>      bfgs converged in  47 scf cycles and  45 bfgs steps
>>
>>      End of BFGS Geometry Optimization
>>
>>      Final energy =    -274.4689566789 Ry
>>
>>
>> so i think the problem is not caused by the structure
>> relaxation.
>> but where is the problem? i can find it . i need your help.
>> thank you if you could give me some hints about the
>> problem!
>>
>>
>> cheers
>>
>>
>>
>>
>>
>>
>>
>>
>>
>>
>> ~
>>
>>
>>
>>
>>
>> the
>> --
>> Xu Yuehua
>> physics Department of Nanjing university
>> China
>> _______________________________________________
>> Pw_forum mailing list
>> Pw_forum at pwscf.org
>> http://www.democritos.it/mailman/listinfo/pw_forum
>
>
>
> _______________________________________________
> Pw_forum mailing list
> Pw_forum at pwscf.org
> http://www.democritos.it/mailman/listinfo/pw_forum
>


-- 
Xu Yuehua
physics Department of Nanjing university
China



More information about the users mailing list