The ABINIT Users Mailing List ( CLOSED )

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

It could definitely be the psp: the HGH are gaussian potentials, made to be used analytically. Using them with planewaves (as in abinit) can and sometimes does result in extremely high ecut, although you may have the record.

I would try to use the PAW potentials, eventually make your own following the tutorial for PAW dataset generation (and post it to us if you get a good one!), but for the moment all you can do is ground state stuff. Otherwise try other norm conserving psp.

Matthieu

On Mon, Sep 15, 2008 at 6:11 PM, Enrico <enrico.avenati@yahoo.com> wrote:

Hi, i'm trying to do some studies on Dy, and CuDy compounds, unfortunetly i
can't get good convergence in the ecut cycle. I'm unsing HGH psuedo, LDA ixc=1
Here there are the result for the Dy (hP2)


ecut    energies                 DeltaH
60      -280,54963299   -15,7738091
80      -283,84757625   -3,2979433
100     -284,24311097   -0,3955347
120     -284,29092258   -0,0478116
140     -284,29218679   -0,0012642
160     -284,29407167   -0,0018849
180     -284,29541506   -0,0013434
200     -284,29504158   3,7348E-4
220     -284,28639706   0,0086445
240     -284,28950882   -0,0031118
260     -284,29244033   -0,0029315



I do have similar result s for almost any compound i tried! I can't reach
convergence. I'm wondering if it could be due to the pseudopotential.
Those result are from abinit 5.5.3 but i add similar problem with 5.3.4 and
5.4.4.

I hope you can help.
Thank you

--
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
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