The ABINIT Users Mailing List ( CLOSED )

[Katalin Gaal-Nagy <katalin.gaal-nagy@physik.uni-regensburg.de>]

Dear all,

I try to calculate some stm images for a Si surface. For this reason I 
changed the input file t46.in in abinit-5.4.4/tests/v4/Input
for my purposes and I have a question:

For the groundstate the criterion for selfconsistency was chosen as
   tolwfr1 1.0d-12
whereas for the stm the parameter was set to
   tolwfr2 1.0d-6

My question is: why the gs is calculated with an accuracy much larger than 
finally the stm? I agree that the gs should be accurate, but the 
difference in the tolwfr seems to be quite large.
For just doing stm starting from a relaxed gs, would it be possible to 
reduce tolwfr1? Has someone tested the influence of this parameter to 
the final stm images?

All the best,
Katalin



P.S: Little question in addition: for me it is easier to think in "toldff" 
instead of tolwfr, does the stm calculation also run with toldff?