The ABINIT Users Mailing List ( CLOSED )

[matthieu verstraete <matthieu.jean.verstraete@gmail.com>]

Hello anaddb users

This happens quite often: the thermodynamic functions, and usually
just the phonon DOS, is compared for successive refinements of the
interpolating phonon q-grid. As soon as the difference in DOS (maximum
diff at all frequencies) is below the tolerance it stops refining.

1) you may need a finer grid, but I doubt it

2) you should look at the DOS at different grid densities to see what
is changing.

3) the most common cause is negative modes, which are accumulated in
the first bin of the DOS. This means that as the grid is refined this
element of the DOS changes strongly (and does not really converge -
worse the remaining weight in the rest of the modes also fluctuates,
if some negative mode q-points are neglected).

                             If you have negative modes, the
convergence _will not work_

This is normal, as the modes are unphysical. Plotting the band
structure is still fine, but in the worst cases the negative modes are
off the high-symmetry lines and do not appear in normal BS plots.
There will be a spike in the phonon DOS accumulated in thm9, however.
We should put a test in the thermodynamical part to avoid these
deadlocks...

Matthieu



2009/8/8 徐 <lanqingxu@fjnu.edu.cn>:
> Dear forum users,
>   I've successfully calculated thermodynamical properties with the following
> input file. Later I changed the input parameter for RF calculation and
> overwrote the merged DDB. But when I change the revised parameters to the
> original ones and get a newly merged DDB, analyze it with  the following
> input file, it  always end in error.
> Error message:
>  thm9 : ERROR -
>   No thermodynamical function is printed out :
>   the tolerance level that was asked
>   has not been match with the grids specified.
>   Action : in the input file, increase the resolution
>   of grids ng2qpt, or decrease the accuracy requirement thmtol.
>
>  leave_new : decision taken to exit ...
> I tried ng2qpt   20 20 20  & ng2qpt   30 30 30,thmtol  0.30   & thmtol
> 0.003 , it is always the same.
> Does that because of parameters in anaddb input file, or because of privious
> RF calculation?The input file for anaddb is as following, and the input file
> for RF calculation is in the attachment.
> Any suggestions will be appreciated.
> Thanks,
> Best regards,
> shelleny
>
> !Input file for the anaddb code.
> !Flags
>  dieflag   1     ! Frequency-dependent Dielectric tensor flag
>  ifcflag   1     ! Interatomic force constant flag
>  thmflag   1     ! Thermodynamical properties flag
>
> !Wavevector grid number 1 (coarse grid, from DDB)
>   brav    1      ! Bravais Lattice : 1-S.C., 2-F.C., 3-B.C., 4-Hex.)
>   ngqpt   4  4  4   ! Monkhorst-Pack indices
>   nqshft  1         ! number of q-points in repeated basic q-cell
>   q1shft  3*0.0
> !nfreq  10
> !frmin 0
> !frmax 6.31
> !Effective charges
>      asr   1     ! Acoustic Sum Rule. 1 => imposed asymetrically
>   chneut   1     ! Charge neutrality requirement for effective charges.
>
> !Interatomic force constant info
>   dipdip  1      ! Dipole-dipole interaction treatment
>   ifcana  1      ! Analysis of the IFCs
>   ifcout 16      ! Number of IFC's written in the output, per atom
>   natifc  2      ! Number of atoms in the cell for which ifc's are analysed
>    atifc  1 2    ! List of atoms
>
> !Phonon band structure output for band2eps - See note near end for
> ! dealing with gamma LO-TO splitting issue.
>    eivec  4
>    prtdos 2
>
> !Wavevector list number 1 (Reduced coordinates and normalization
> factor)
>   nph1l    41      ! number of phonons in list 1
>
>   qph1l   0.5000  0.5000  0.5000   1.0    !(L point)
>           0.5500  0.5500  0.5500   1.0
>           0.6000  0.6000  0.6000   1.0
>           0.6500  0.6500  0.6500   1.0
>           0.7000  0.7000  0.7000   1.0
>           0.7500  0.7500  0.7500   1.0
>           0.8000  0.8000  0.8000   1.0
>           0.8500  0.8500  0.8500   1.0
>           0.9000  0.9000  0.9000   1.0
>           0.9500  0.9500  0.9500   1.0
>           1.0000  1.0000  1.0000   1.0    !(gamma point)
>           0.9500  0.9500  1.0000   1.0
>           0.9000  0.9000  1.0000   1.0
>           0.8500  0.8500  1.0000   1.0
>           0.8000  0.8000  1.0000   1.0
>           0.7500  0.7500  1.0000   1.0
>           0.7000  0.7000  1.0000   1.0
>           0.6500  0.6500  1.0000   1.0
>           0.6000  0.6000  1.0000   1.0
>           0.5500  0.5500  1.0000   1.0
>           0.5000  0.5000  1.0000   1.0    !(X point)
>           0.4875  0.4875  0.9750   1.0
>           0.4750  0.4750  0.9500   1.0
>           0.4625  0.4625  0.9250   1.0
>           0.4500  0.4500  0.9000   1.0
>           0.4375  0.4375  0.8750   1.0
>           0.4250  0.4250  0.8500   1.0
>           0.4125  0.4125  0.8250   1.0
>           0.4000  0.4000  0.8000   1.0
>           0.3875  0.3875  0.7750   1.0
>           0.3750  0.3750  0.7500   1.0    !(K point)
>           0.3375  0.3375  0.6750   1.0
>           0.3000  0.3000  0.6000   1.0
>           0.2625  0.2625  0.5250   1.0
>           0.2250  0.2250  0.4500   1.0
>           0.1875  0.1875  0.3750   1.0
>           0.1500  0.1500  0.3000   1.0
>           0.1125  0.1125  0.2250   1.0
>           0.0750  0.0750  0.1500   1.0
>           0.0375  0.0375  0.0750   1.0
>           0.0000  0.0000  0.0000   1.0    !(gamma point)
>
> !Wavevector list number 2 (Cartesian directions for non-analytic gamma
> phonons)
>  nph2l    1       ! number of directions in list 2
>
>   qph2l   1.0  0.0  0.0    0.0
>
> !Wavevector grid number 2 (series of fine grids, extrapolated from interat
> forces)
>   ng2qpt   20 20 20  ! sample the BZ up to ngqpt2(I've also tried 30 30 30 )
>   ngrids   5         ! number of grids of increasing size
>   q2shft   3*0.0
>
> !Thermal information
>   nchan   1250   ! # of channels for the DOS with channel width 1 cm-1
>   nwchan  1      ! # of different channel widths from this integer down to 1
> cm-1
>   thmtol  0.30  ! Tolerance on thermodynamical function fluctuations(I've
> also tried 0.003)
>
>   ntemper 100     ! Number of temperatures
>   temperinc 1  ! Increment of temperature in K for temperature dependency
>   tempermin 4.  ! Minimal temperature in Kelvin
>
> # This line added when defaults were changed (v5.3) to keep the previous,
> old behaviour
>   symdynmat 0
>



-- 
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Dr. Matthieu Verstraete

European Theoretical Spectroscopy Facility (ETSF)
Dpto. Fisica de Materiales,
U. del Pais Vasco,
Centro Joxe Mari Korta, Av. de Tolosa, 72,   Phone: +34-943018393
E-20018 Donostia-San Sebastian, Spain        Fax  : +34-943018390

Mail : matthieu.jean.verstraete@gmail.com
http://www-users.york.ac.uk/~mjv500