Page 1 of 1

occopt problem for 2x2x2 supercells

Posted: Wed Jun 22, 2011 11:18 am
by iamikaruk
Hi all
I'm trying to calculate the phonon band of Ba(Ti1-xZrxO3) (x=0.125) with a 2x2x2 supercell structure using abinit 5.4.4p. During the non self-consistent calculation, an error occurs:
"occeig : ERROR -
In a non-metallic case (occopt<3), for a RF calculation,
if the eigenvalues are degenerate, the occupation numbers must also be degenerate.
However, the following pair of states gave :
k -state, band number 160, occ= 2.000000E+00, eigenvalue= 2.993290E+00,
kq-state, band number 161, occ= 0.000000E+00, eigenvalue= 2.993290E+00.
Action : change occopt, consistently, in GS and RF calculations."
I've tried to change the default nband value (nband=165) to the fully occupied value nband=160 but the convergence of tolwfr is very slow. Is there any solution to this problem? Thank you very much!
Regards

Re: occopt problem for 2x2x2 supercells

Posted: Tue Aug 16, 2011 3:21 am
by Robin
(1) Switch to metallic occupation with small tsmear

(2) Use nbdbuf to speed the convergence of tolwfr

(3) Update to the latest Abinit, which improves a lot on speed according to my experiences

Sincerely,
Guangfu Luo