Hi,

Thank you for your kindly reply.

>it is normal that as you increase your system to include many
> >equivalent (or almost) atoms


I do not understand why many equivalent atoms will affect the convergence.

>To get an idea, look at the results (Mulliken charges; DOS)
> >underway, after some number of iterations.


Could you give me some ditails about checking the results of Mulliken
charges or DOS after some number of iterations?
Usually, I just check the following information of the scf.

-----------------------------------------------------------------------------------------------------------------------------------
siesta: iscf   Eharris(eV)      E_KS(eV)   FreeEng(eV)   dDmax  Ef_up
Ef_dn(eV)
 ......
siesta:  799   -32180.0924   -32180.0919   -32180.0919  0.0011  -3.0670
-3.0670
siesta:  800   -32180.0924   -32180.0919   -32180.0919  0.0011  -3.0670
-3.0670
siesta:  801   -32180.0924   -32180.0919   -32180.0919  0.0011  -3.0670
-3.0670
siesta:  802   -32180.0924   -32180.0919   -32180.0919  0.0011  -3.0670
-3.0670
siesta:  803   -32180.0925   -32180.0920   -32180.0920  0.0011  -3.0670
-3.0670
siesta:  804   -32180.0925   -32180.0920   -32180.0920  0.0011  -3.0670
-3.0670
siesta:  805   -32180.0926   -32180.0921   -32180.0921  0.0011  -3.0670
-3.0670
siesta:  806   -32180.0925   -32180.0921   -32180.0921  0.0011  -3.0670
-3.0670
siesta:  807   -32180.0926   -32180.0921   -32180.0921  0.0011  -3.0670
-3.0670
siesta:  808   -32180.0926   -32180.0921   -32180.0921  0.0011  -3.0670
-3.0670
siesta:  809   -32180.0926   -32180.0921   -32180.0921  0.0011  -3.0670
-3.0670
siesta:  810   -32180.0927   -32180.0922   -32180.0922  0.0011  -3.0669
-3.0669
siesta:  811   -32180.0927   -32180.0923   -32180.0923  0.0011  -3.0669
-3.0669
siesta:  812   -32180.0928   -32180.0923   -32180.0923  0.0011  -3.0669
-3.0669
siesta:  813   -32180.0929   -32180.0924   -32180.0924  0.0011  -3.0669
-3.0669
siesta:  814   -32180.0927   -32180.0923   -32180.0923  0.0011  -3.0669
-3.0669
siesta:  815   -32180.0928   -32180.0924   -32180.0924  0.0011  -3.0669
-3.0669
siesta:  816   -32180.0928   -32180.0924   -32180.0924  0.0011  -3.0669
-3.0669
siesta:  817   -32180.0929   -32180.0925   -32180.0925  0.0011  -3.0669
-3.0669
siesta:  818   -32180.0928   -32180.0923   -32180.0923  0.0011  -3.0669
-3.0669
siesta:  819   -32180.0928   -32180.0924   -32180.0924  0.0011  -3.0669
-3.0669
siesta:  820   -32180.0929   -32180.0925   -32180.0925  0.0011  -3.0669
-3.0669
siesta:  821   -32180.0929   -32180.0925   -32180.0925  0.0011  -3.0669
-3.0669
siesta:  822   -32180.0929   -32180.0925   -32180.0925  0.0010  -3.0669
-3.0669
siesta:  823   -32180.0930   -32180.0926   -32180.0926  0.0010  -3.0669
-3.0669
siesta:  824   -32180.0931   -32180.0927   -32180.0927  0.0010  -3.0669
-3.0669
siesta:  825   -32180.0931   -32180.0927   -32180.0927  0.0010  -3.0669
-3.0669
siesta:  826   -32180.0929   -32180.0925   -32180.0925  0.0010  -3.0669
-3.0669
siesta:  827   -32180.0931   -32180.0927   -32180.0927  0.0010  -3.0668
-3.0668
......
----------------------------------------------------------------------------------------------------------------------

I'd like to increase the DM.MixingWeight to 0.3 to accelerate the
calculation. Is it correct? Since you said I need a small mixing parameter,
how to understand the mixing parameter? I read some notes about
DM.MxingWeight, but there is still something puzzled me.

alpha has to be small (0.1-0.3) for insulator and semiconductors, tipically
> much smaller for metals.
>

BTW, whether my setting of the antiferromagnetism is correct.

Best,

On Mon, Jul 26, 2010 at 2:03 AM, <[email protected]> wrote:

> > Dear all,
> >
> > I intend to calculate the exchange integral of CdFe2O4, so I need the
> > energy
> > of four magnetic structures, one ferromagnetism and three
> > antiferromagnetism. The ferromagnetic system have finished in 40 hours,
> > but
> > the antiferromagnetic system have taken about 170 hours. The calculation
> > converged so slowly and there was already 700 interations.
>
> Hi,
> it is normal that as you increase your system to include many
> equivalent (or almost) atoms, the convergence becomes difficult,
> because of electrons flopping between nearly degenerate states.
> So sometimes you need just patience and very small mixnig parameter.
> The crucial question is, is your calculation on the right way
> and simply converges slowly, or is it wildly fluctuating.
> To get an idea, look at the results (Mulliken charges; DOS)
> underway, after some number of iterations.
>
> Good luck
>
> Andrei Postnikov
>



-- 
Bin Shao
College of Information Technical Science, Nankai University
94 Weijin Rd. Nankai Dist. Tianjin 300071, China
Email: [email protected]

Responder a