mBJ is a potential and not a functional, so it is very clumsy; Ef has to be in
the gap, and you cannot make any changes;
If you run without mBJ
runsp -so -orb
then in the mBJ run you have to run the same
runsp -so -orb
and for example you cannot change the Hubbard U value
When I say clumsy I mean that as far as you obey these conditions it gives good
results
________________________________
De: [email protected]
<[email protected]> en nombre de Vijeta Singh
<[email protected]>
Enviado: domingo, 14 de junio de 2015 07:43 a. m.
Para: [email protected]
Asunto: [Wien] Regarding mBJ- GGA calculation
Dear WIEN2k users
I am using WIEN2k 13.1 version. I have some doubts regarding mBJ- GGA
calculation.
(i) If we include SOC for a well converged cubic system, mBJ-GGA fails towards
convergence. A crash is seen in the second iteration in the LAPW0 part.
However, for simple GGA it worked fine with the prescription "runsp -s lapw1"
.I am aware of the problem being posted on the mailing list, but couldnt find
it of much help.
(ii) what would be the difference in the following 2 calculations
(a) first GGA is done, then mBJ followed by SOC
(b) first GGA done, then SOC and then followed by mBJ
do we expect similar results in total energy, magnetic moments, band structure
etc etc in these two calculations?
Thanking you
_______________________________________________
Wien mailing list
[email protected]
http://zeus.theochem.tuwien.ac.at/mailman/listinfo/wien
SEARCH the MAILING-LIST at:
http://www.mail-archive.com/[email protected]/index.html
