The ABINIT Users Mailing List ( CLOSED )

[Matthieu Verstraete <mjv500@york.ac.uk>]

Hello everyone,

does anyone else have this? I've generated a Pd paw dataset for abinit, 
and Ag as well. Both give reasonable physical properties etc, but while 
the scf cycle for Ag converges very nicely (10-15 iterations) that for Pd 
oscillates for a long time around 10e-8 10e-9 and eventually runs out of 
scf cycles. Why? And why such a difference.

I imagine it must be linked to the quality of my paw psp: the pd one is 
transposed from the ag inputs for uspp2abinit. Are there any obscure 
tweaks I can do? Both paw psp are nice and smooth, transferable. Pfff!

What parameters vary between elements, beyond the atomic number and 
occupations? Do the cutoff radii have this kind of effect?

Thanks,

Matthieu

--
================================================================
Dr. Matthieu Verstraete                 mailto:mjv500@york.ac.uk
Dept. of Physics, University of York,     tel: +44 1904 43 22 08
Heslington, YO10 5DD York, United Kingdom fax: +44 1904 43 22 14