[Wien] Problem optimizing internal parameters NN - error

Bo Qiu 200210qb at gmail.com
Sun Oct 14 07:14:43 CEST 2007


Dear Wien2k users and developers,

I'm a new user and I've been struggling with this problem and find no way
out for several days and I'm really in despair. I sincerely request for your
help as my last hope, thanks a lot!

The problem is that:
When I'm trying to do the optimization with the example TiO2 and try to
optimize its atomic positions using min_lapw. Firstly I tried to follow the
userguide in ch 10.3 exactly using w2web, and run scf first. As pointed in
the userguide, the last force was a jump. Then I proceeded to mini
positions, however, first of all, before every min_lapw round there is a
warning saying:

WARNING: Step size reduced due to overlapping spheres -- check RMT

after several (around 10) rounds, PORT crashed:

>>  (mini) PORT stopped -> exit
 :WARNING you might have inconsistent Forces/Energies

The last few lines on FGL002 inside *mini are (FGL001 is alway zero):

:FGL002:   2.ATOM               -45.621114095   -45.621114095     0.000000000 t
:FGL002:   2.ATOM               -45.586465863   -45.586465863     0.000000000 t
:FGL002:   2.ATOM               -45.575859262   -45.575859262     0.000000000 t

Then I removed case.tmpM and modified case.inM as

NEW1 2.0 0.35      # PORT/NEWT;  tolf, Initial Trust Radius
0.0 0.0 0.0 1.0   #Atom  1 Generated by pairhess
1.0 1.0 0.0 1.0   #Atom  2 Generated by pairhess

Then rerun min_lapw using w2web

however, this time, in the first round, after 4 cycles I get it
crashed with NN -Error:

in cycle 4    ETEST: 0   CTEST: 0
 LAPW0 END
 LAPW1 END
 LAPW2 END
 CORE  END
 MIXER END
ec cc and fc_conv 1 1 1

>   stop
NN - Error

The version used: Wien2k_07, compiler ifort, mkl 9.1, in series mode

Could you please help me? Thank you very much!!! If any more information
needed, please let me know, thanks!

-- 
Thanks and Best Wishes.

Sincerely yours,
Bo Qiu

http://web.ics.purdue.edu/~qiub/ <http://web.ics.purdue.edu/%7Eqiub/>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://zeus.theochem.tuwien.ac.at/pipermail/wien/attachments/20071014/b83dc4d4/attachment.html


More information about the Wien mailing list