The ABINIT Users Mailing List ( CLOSED )

[DORADO Boris Thésard <boris.dorado@cea.fr>]

Thank you for your answers,

I used the 25x25x25 grid just to have a perfect agreement between the energy 
differences and the force differences. And as the calculation is very fast 
even with a 25x25x25 grid, I did not bother to reduce the grid.

I intended to go back to a smaller grid but I just wanted to check the 
reasons why the dynamical matrix would be incomplete.

Thanks again for your answers. It indeed works well with the grid provided in 
the tutorial. I guess the grid must be well chosen for the phonon 
calculations...

Regards

Boris Dorado
-----Message d'origine-----
De : Aldo Humberto Romero [mailto:aromero@qro.cinvestav.mx] 
Envoyé : mardi 27 octobre 2009 15:43
À : forum@abinit.org
Objet : Re: [abinit-forum] The dynamical matrix was incomplete : phonon 
frequencies may be wrong ...

Did you check also that rfdir corresponds to the symmetry
you are calculating?.. in cubic systems
you just need 1 0 0 but for non-cubic .. 1 1 1 would be correct.. if you
do not
include them correctly, your matrices will be incomplete.

Also, as Josef has pointed out, the 25x25x25 grid is quite large...you get
excellent phonon spectra with much less, you need to use that
fine grid for properties quite sensitive to the mesh, as the electron-phonon.
Remember that the anadbb
utility is able to use Fourier interpolation!!!

regards

-aldo.

-> Dear Abinit users,
->
->
->
-> I'm currently trying to perform some calculations of phonon frequencies in
-> AlAs, as is done in the first tutorial of the Abinit website.
->
->
->
-> I changed, however, some parameters in order to have a better accuracy of
-> phonon frequencies and to assess the variation of the frequencies with
-> respect to the cutoff and the kpt sampling. I therefore set the cutoff to
-> 400 eV (about 15 Ha, 3 Ha in the tutorial) and the kpt grid to 25x25x25,
-> but the calculation of the phonon perturbation displays this warning along
-> with the phonon frequencies:
->
->
->
-> chkph3 : WARNING -
->
->   The dynamical matrix was incomplete : phonon frequencies may be wrong
-> ...
->
->
->
-> Could someone please tell me why the dynamical matrix would be incomplete?
-> Does it have anything to do with my changed calculation parameters?
-> Because I kept the same input file as in the tutorial except I changed the
-> cutoff and the kpt sampling. I also checked the output file of the
-> tutorial and this warning of course does not appear.
->
->
->
-> In addition, my phonon frequencies are indeed wrong (308 cm-1 and 337
-> cm-1, instead of something like 359 cm-1 and 390 cm-1).
->
->
->
-> Thank you for your answers.
->
->
->
-> Regards
->
->
->
-> Boris Dorado
->
->
->
->


-- 

Prof. Aldo Humberto Romero
CINVESTAV-Unidad Queretaro
Libramiento Norponiente 2000
CP 76230, Queretaro, QRO, Mexico
tel: 442 211 9909
fax: 442 211 9938

email: aromero@qro.cinvestav.mx
        aldorome@gmail.com
www: qro.cinvestav.mx/~aromero