[Ifeffit] FMS + SCF Card
joshua jason kas
hebhop at u.washington.edu
Thu Jun 19 16:59:54 CDT 2008
> I have some questions concerning the FMS and SCF Card in XANES-Calculations.
> I have always used the same value for lfms1 (SCF-Card) and lfms2
> (FMS-Card). Does it make sense to use different values in the same
> feff.inp (for example lfms1 = 0 and lfms2 = 1)? I don't think so.
> There is the lfms2 parameter in the FMS-Card and the description in the
> FEFF manual for lfms2 = 0 (appropriate for solids): "With the default
> value of 0, the FMS is calculated for a cluster of size rfms around each
> representative unique potential." My question is, what does it mean "each
> representative unique potential"? Every potential, which is defined in the
> POTENTIAL Card? I.e. with 4 potentials in the POTENTIAL Card, the FMS is
> done for each of these 4 potentials? Or does it mean, the FMS is done for
> really every potential of every atom?
The SCF with lfms1 = 0, works in the following manner.
1) Representative atoms are chosen for each potential type defined in the
POTENTIAL card. These are chosen to be the closest atoms of that type to
the absorbing atom.
2) FMS is performed for each representative atom (including the absorber)
with FMS radius measured from that representative atom.
So, to answer your questions,
1) lfms2 does not do anything unless you set the second value of the PRINT
card to 3 or greater, and then it only changes the way that the Mulliken
counts and the output in ratio.dat are calculated, not xmu.dat (the final
2) With 4 potentials in the POTENTIAL Card, the FMS is
done for each of these 4 potentials NOT for each atom in the cluster.
The density of states is then calculated inside the norman radius
centered about each of the representative atoms, and is assumed to be
the same for all other atoms of that type.
> Thanks for your help and time.
> sincerely yours,
> Eckhard Bosman
> e.bosman at stud.uni-goettingen.de
> +49 (0)551-39-14441
> Raum: E0.104
> Institut f?r R?ntgenphysik
> Friedrich-Hund-Platz 1
> 37077 G?ttingen
> Ifeffit mailing list
> Ifeffit at millenia.cars.aps.anl.gov
> End of Ifeffit Digest, Vol 64, Issue 10
More information about the Ifeffit