[Wien] density matrix is not positive

[Peter Blaha]

This should not happen.

I suspect some input error ???

Did you change case.indm/inorb after a first LDA+U run ?? You may still 
have "wrong" dmats (corresponding to an old input).

In that case remove any case.dmatXX_old files, run    x lapwdm -up/dn 
and start a new scf cycle (with the unmixed dmats).



On 1/2/20 9:54 AM, Llin Layton wrote:
> Dear wien2k users,
> I am working on Nd4Ni3O10 with the latest wien2k 19.1, during my LDA+U 
> calculations. However, I constantly encountered warning messages:
> WARNING: Density Matrix is not positive semi-definite
> WARNING: Density Matrix has -ve terms
> and it takes about 200 iterations to reach the default converge 
> conditions. I set the Ueff=8eV for 4f electrons of Nd and Ueff=6eV for 
> Ni's 3d electrons, rkmax=7.5 with 1200 kpoints, rmt is set by setrmt. 
> Besides, I find the 4f orbital is always localized at fermi energy, 
> which is also the case for Nd4Ni3O8, and I have performed normal LDA 
> calculations before +U.
> So, my questions are:
> 1. Is this a serious warning or I can just ignore it since it can still 
> converge?
> 2. Is that normal to find the 4f orbital localized at the fermi energy 
> given that I have set Ueff=8eV for it?
> Any suggestions will be appreciated.
> 
> Yours sincerely,
> 
> Fan
> 
> 
> 
> 
> 
> _______________________________________________
> Wien mailing list
> Wien at zeus.theochem.tuwien.ac.at
> http://zeus.theochem.tuwien.ac.at/mailman/listinfo/wien
> SEARCH the MAILING-LIST at:  http://www.mail-archive.com/wien@zeus.theochem.tuwien.ac.at/index.html
> 

-- 

                                       P.Blaha
--------------------------------------------------------------------------
Peter BLAHA, Inst.f. Materials Chemistry, TU Vienna, A-1060 Vienna
Phone: +43-1-58801-165300             FAX: +43-1-58801-165982
Email: blaha at theochem.tuwien.ac.at    WIEN2k: http://www.wien2k.at
WWW:   http://www.imc.tuwien.ac.at/TC_Blaha
--------------------------------------------------------------------------