[Wien] change of structure during run_lapw
Dear WIEN2k users, I know it is correct to run run_lapw for one structure, then to do little changes in lattice parameters in structure file case.struct and then remove broyden files and simple continue run_lapw (or runafm_lapw). However, is this allowed in case that one has to change not only lattice parameters but as well direct coordinates of atoms? I have a monoclinic structure with one angle 90.4 degrees and after little deformation I come to the next structure that should have the angle 89.5 degrees. But because WIEN2k prefers obtuse angle for monoclinic lattices, I rather choose angle 90.5 degrees instead of 89.5 degrees and change all the X direct coordinates of all atoms to 1-X. After this change of structure file, can I still only remove broyden files and continue runafm_lapw or do I have to strictly run x dstart or clmextrapol_lapw before continuing runafm_lapw? Thank you very much for answer Tomas Kana
[Wien] x lapw2 -p -qtl fails with shared memory configuration
Dear Wien community,
I'm just updating WIEN2k from version 9.1 to 10.1. The tar-file was
downloaded last thursday, march 24th. I compiled the programs under
SuSE-Linux 11.3 with ifort 11.1. The standard options supplied by
siteconfig_lapw worked well.
However, when I wanted to calculate partial charges in parallel (x lapw2
-qtl -p) lapw2 fails:
c3serv4{wedig}2388: x lapw2 -p -qtl
running LAPW2 in parallel mode
calculating QTL's from parallel vectors
moving vector to test101b.vector_1.old
moving vector to test101b.vector_2.old
moving vector to test101b.vector_3.old
L2main - Error
0.072u 0.052s 0:00.11 109.0%0+0k 0+0io 0pf+0w
Error in LAPW2
'l2main' - error reading parallel vectors
I found the following reason for this behaviour. Compared to version
9.1, the following line was added in lapw2para_lapw for the -qtl option:
vec2old_lapw -p -local -$updn
This should copy vector files from remote scratch directories to the
case directory. However, a look into vec2old_lapw shows, that the -local
option is only considered, if $useremote ($USE_REMOTE in the file
parallel_options) is set to 1. I have configured WIEN2k for shared
memory without MPI. In this case $useremote is set to 0 and the vector
files are moved to *.vector_n.old in any case, leading to the error above.
As a simple workaround I modified vec2old_lapw as shown below. This work
for me as the scratch and the case directories are always identical
(SCRATCH=./). When $SCRATCH points to a different location, further
modifications may be necessary.
while ( $i <= $#machines )
if( $copy == '') then
if ($useremote == 1) then
if($?local) then
.
.
else
.
.
endif
else
if( ! $?local) then <=== this line added
mv $file.vector${updn}_$i $file.vector${updn}_$i.old
if($status == 0 ) echo moving vector to $file.vector${updn}_$i.old
endif <=== this line added
endif
Best regards,
Ulrich Wedig
--
-
Dr. Ulrich Wedig Tel. 0711/6891535
Max-Planck-Institut fuer Festkoerperforschung FAX 0711/6891502
Heisenbergstr. 1
70569 Stuttgart U.Wedig at fkf.mpg.de
-
[Wien] change of structure during run_lapw
If you have a small change in the atomic positions clmextrapol works well. For a large one you may have problems, and need to redo dstart. For your case I would use the 89.5 angle. While Wien2k may not prefer it, it should work fine. Changing to 1-x is a big change. On Mar 29, 2011 5:50 AM, "Tomas Kana" wrote: > Dear WIEN2k users, > > I know it is correct to run run_lapw for one structure, then to > do little changes in lattice parameters in structure file case.struct and then > remove broyden files and simple continue run_lapw (or runafm_lapw). > However, is this allowed in case that one has to change not only lattice parameters but as well > direct coordinates of atoms? > I have a monoclinic structure with one angle 90.4 degrees and after little deformation I come > to the next structure that should have the angle 89.5 degrees. But because WIEN2k prefers obtuse > angle for monoclinic lattices, I rather choose angle 90.5 degrees instead of 89.5 degrees and > change all the X direct coordinates of all atoms to 1-X. After this change of structure file, > can I still only remove broyden files and continue runafm_lapw or do I have to strictly run x dstart or clmextrapol_lapw before continuing runafm_lapw? > Thank you very much for answer > Tomas Kana > ___ > Wien mailing list > Wien at zeus.theochem.tuwien.ac.at > http://zeus.theochem.tuwien.ac.at/mailman/listinfo/wien -- next part -- An HTML attachment was scrubbed... URL: <http://zeus.theochem.tuwien.ac.at/pipermail/wien/attachments/20110329/957e25c8/attachment.htm>
[Wien] x lapw2 -p -qtl fails with shared memory configuration
Dear Wien community, indeed, the problem I reported in my last mail is solved, when using the vec2old_lapw version that was distributed in the mailing list by Peter Blaha on march, 7th. Best regards, Ulrich Wedig -- - Dr. Ulrich Wedig Tel. 0711/6891535 Max-Planck-Institut fuer Festkoerperforschung FAX 0711/6891502 Heisenbergstr. 1 70569 Stuttgart U.Wedig at fkf.mpg.de -
[Wien] MAE
Dear all, I'm beginner user of Wien2k. For my scientific job I need to calculate Magnetocrystalline Anisotropy energy (MAE) of magnetite Fe3O4. I have read "User Guide" and "Orbital package in WIEN" (P.Novak, 2002), but I don't understand, how to do this calculations. I'll be?very?appreciate?if you could?help me. What should I do step by step? Thanks in advance. -- Best regards, Irina
