The ABINIT Users Mailing List ( CLOSED )

["Aldo Humberto Romero" <aromero@qro.cinvestav.mx>]

As Yu has explained.. you have to check convergence with respect to ecut
before
the kpoints... 17Ha is really low for norm pseudopotentials...if you have
problems
with the number of K points and things like that, you can use flags
mkmem and related variables to decrease the need of memory used
during the calculations...

good luck with your calculations


-> Yes you are right, i didn't do the convergency test for phonon
-> frequencies,
-> i have just did it for Etotal.
->
-> Thank you very much i will take your advice. But 17Ha of ecut with 8 cubed
-> ngkpt is almost touch the limit of my computer. Do you think maybe i set
-> ngkpt too high while ecut too low.
->
-> Thanks again
->
-> Fan
->
-> 2009/12/20 Xie Yu <lazycle@gmail.com>
->
->> Dear Fan,
->>
->> When you use the ab initio package for a particular properties, you have
->> to
->> test your pseudopotential and the parameters very carefully.
->>
->> There are some parameters related to the system besides the kpoints and
->> ecut, i.e. smearing and occupation. You should test them before you do
->> the
->> real calculation.
->>
->> In my point of view, you can calculate the frequencies at the gamma
->> point
->> with different parameters until the frequencies are converged . Then you
->> can
->> use the converged parameters for the whole phonon dispersion
->> calculation.
->>
->> PS: 17 Ha/ 34 Ry is quite small for a norm conserving pseudopotential.
->>
->> Best,
->>
->> Yu
->>
->>
->> On Sun, Dec 20, 2009 at 1:25 AM, jinhong zhang
->> <zhangjinhong2009@gmail.com
->> > wrote:
->>
->>> Dear Abiniters
->>>
->>> I recently calculated phonon dispersion of B12P2 following the RF
->>> tutorial. I first optimized the geometry of the material by GGA and the
->>> tensor stress in the material reached e-7, which i believe is accurate
->>> enough. And the ngkpt I chose first time is 4 4 4 and ecut is 15. The
->>> result
->>> of phonon dispersion turned out to be bad, some low frequencies
->>> appeared
->>> negative and even reached -700cm-1 (but the acoustic frequencies are 0
->>> at
->>> Gamma point). And then I changed the ngkpt to 8 8 8 and ecut to 17, and
->>> optimized and calculated again, the graph turned out to be better but
->>> still
->>> not good, only 3 acoustic frequencies represented negative beyond Gamma
->>> point this time, and the magnitude reached about -500cm-1max.
->>>
->>> I was wondering the reason. Is the problem of ABINT or somewhere i set
->>> incorrect? (I preferred the later:P) or the stability of this material
->>> is
->>> not good? But my teacher had calculated the phonon dispersion of
->>> B12O2-a
->>> similar one-by Material Studio, which the result was good.
->>>
->>> Hope any Abiniter can give me some advices and help me solve the
->>> problem,
->>> i would be very appreciated.
->>>
->>> Best regards
->>>
->>> Fan
->>>
->>
->>
->


-- 

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