The ABINIT Users Mailing List ( CLOSED )

["Nichols A. Romero" <naromero@gmail.com>]

Hi,




In the ABINIT online manual,


http://www.abinit.org/Infos_v4.5/varbas.html#tolwfr




it states in reference to using tolwfr that
"this criterion should be replaced
by those based on toldfe (preferred for 

ionmov==0),
toldff (preferred for 

ionmov/=0) or
tolvrs (preferred for theoretical reasons!)."

I would think that in general, the best criterion for stopping the SCF
is tolvrs since when self-consistency is reached it is true that
V_scf^in = V_scf^out

and

n^in = n^out

Especially in response calculation, I've noticed that using something
other than tolvrs, e.g. toldfe could be potentially problematic. Isn't
possible that under certain circumstances toldfe could be reached much
faster than its equivalent value for tolvrs?  If its possible to speak of an equivalent value for tolvrs.

-- 
Nichols A. Romero
4-133
ESB, MC
704                508
E. Michigan Avenue
1110 West Green Street          Apartment 12
Urbana, IL 61801-3080            Urbana, IL 61801-5176
(217)
333-6108                      
(217) 417-5210