[Wien] Slow convergence of charge

[Shafqat Hussain Shah]

Dear Wien2k users

I have successfully optimized the volume and various ratios of lattice
parameters of an orthorhombic system (Bi2Fe4O9) without any problem (0.33%
error in a, 0.17% error in b and 2.86% error in c lattice parameter from
the corresponding Exp values). Now when I try to relax the atomic positions
of the system using MSR1a, it is, unexpectedly, very slow (around 300
iterations) in charge convergence. The structure for this calculation has
optimized lattice parameters  and ratios with experimental atomic positions
that I want to relax. I have used following command

runsp_lapw -p -i 999 -ec 0.0001 -cc 0.001 -fc 0.50

Just to check I started a new simple scf calculation which also ran into
trouble due to the slow convergence of charge density. Interestingly all
the calculations with experimental parameters and Optimizer generated
structures for option 1, 3 and 4 converge nicely (around 80 iterations with
the above mentioned command). I really don't know what is wrong with the
calculation. Any help in this regard would be highly appreciated. I have
tried to attach scf files but they do not go through due to large size. I
have attached first iteration and log file for a quick examination so that,
if possible, someone can spot a problem.  If you need any further
information I would be happy to provide it.

I am using WIEN2k_13.1 on a 12 core machine with opensuse 11.2.

Thanks in advance,

Dr. Shafqat Hussain Shah
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://zeus.theochem.tuwien.ac.at/pipermail/wien/attachments/20141216/78bb2017/attachment.html>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: BIFOfm_MSR1a.zip
Type: application/zip
Size: 18098 bytes
Desc: not available
URL: <http://zeus.theochem.tuwien.ac.at/pipermail/wien/attachments/20141216/78bb2017/attachment.zip>