Kari Ruotsalainen (guest) 16 Jan 2013 15:07
Hi all,
When calculating the ground state densities for various perovskite oxides I've noticed that rgkmax 7 or 8 seem to be some sort of a boundary beyond which something starts going wrong in the SCF calculation. I get unphysically large total energies and warnings about the total charge density being incorrect. There's also warnings about smallest valence eigenvalue being smaller than the minimum linearization energy.
I've tried digging up muffin tin radii and different expansion (e.g. lmaxapw, lmaxvr, gmaxvr ) parameters from literature for the same compounds I've been working on. I've also tried switching between different mixers and made a few attempts with mixed LAPW/apw+lo basis functions. So far I've had no luck with those attempts.
I'm new to the lapw/apw method, so maybe I'm missing something important or just doing something wrong? Or perhaps there are some practical issues or rules of thumb for oxides that I sould be aware of? I'd greatly appreciate any help or advice.
I think my input files are ok. Units are in au, geometries have been used with other dft codes etc. But just to make sure I've added an GS calculation input file for SrTiO3 to the end of this post.
Best regards,
Kari Ruotsalainen
<input>
<title>STO-GS</title>
<structure speciespath="../../species">
<crystal stretch="7.3794 7.3794 7.3794">
<basevect>1 0.0 0.0</basevect>
<basevect>0.0 1.0 0.0</basevect>
<basevect>0.0 0.0 1.0</basevect>
</crystal>
<species speciesfile="Ti.xml" chemicalSymbol="Ti">
<atom coord="0.5 0.5 0.5" />
</species>
<species speciesfile="Sr.xml" chemicalSymbol="Sr">
<atom coord="0.0 0.0 0.0" />
</species>
<species speciesfile="O.xml" chemicalSymbol="O">
<atom coord="0.5 0.5 0.0" />
<atom coord="0.5 0.0 0.5" />
<atom coord="0.0 0.5 0.5" />
</species>
</structure>
<groundstate ngridk="10 10 10" rgkmax="7" lmaxapw="12"
do="fromscratch" gmaxvr="16" lmaxvr="4 vkloff="0.5 0.5 0.5"></groundstate>
</input>