[Wien] character plot, crystallography related and min_lapw

[Lawal Mohammed]

Dear Developers and Users,
I need help on these three problems.
1) I tried to plot band character by editing case.insp as explained in the user guide, but band character for the selected orbital is not showing. I run a TiC example as test-case and it works fine, what could have been the problem?

2) I run Mg(OH)2 exercise (as contained in the workshop documents) for the optimization of positions. I noticed that the primitive cell and the conventional cell of the Mg(OH)2 is the same when viewed using xcrysden, is this always the case H lattice or there are some exceptions?
3) Please is this command min_lapw -i 150 -p -NI correct? I tried it for non spin polarized case, but the scf terminated at 40 circles with forces not converged. I already have previous scf converged for volume optimization.

Thanks for your time, looking forward for help.
With kind regards.

Lawal

[Laurence Marks]

Concerning 2), I believe hexagonal is always the primitive cell, with the
caveat of rhombohedral cells.

Concerning 3), it is hard (impossible) to answer without more information.
It could be any of: a) too large RMTs, b) not enough iterations to converge
the forces, c) inappropriate convergence parameters, d) some other
incorrect parameters.

Note: normally MSR1a using run**_lapw -min is better than min_lapw.

_____
Professor Laurence Marks
"Research is to see what everybody else has seen, and to think what nobody
else has thought", Albert Szent-Gyorgi
www.numis.northwestern.edu

Post by Lawal Mohammed
Dear Developers and Users,
I need help on these three problems.
1) I tried to plot band character by editing case.insp as explained in the
user guide, but band character for the selected orbital is not showing. I
run a TiC example as test-case and it works fine, what could have been the
problem?
2) I run Mg(OH)2 exercise (as contained in the workshop documents) for the
optimization of positions. I noticed that the primitive cell and the
conventional cell of the Mg(OH)2 is the same when viewed using xcrysden,
is this always the case H lattice or there are some exceptions?
3) Please is this command *min_lapw -i 150 -p* *-NI *correct? I tried it
for non spin polarized case, but the scf terminated at 40 circles with
forces not converged. I already have previous scf converged for volume
optimization.
Thanks for your time, looking forward for help.
With kind regards.
*Lawal *

[Lawal Mohammed]

Dear Prof. Laurence Marks,
Thank you so much for your reply.
I will check on the points that you mentioned and also try the MSR1a option. In the meantime I will be equally grateful if someone respond to my first question.
With kind regards.

Lawal



On Wednesday, October 10, 2018, 8:03:17 PM GMT+8, Laurence Marks <L-***@northwestern.edu> wrote:

Concerning 2), I believe hexagonal is always the primitive cell, with the caveat of rhombohedral cells.
Concerning 3), it is hard (impossible) to answer without more information. It could be any of: a) too large RMTs, b) not enough iterations to converge the forces, c) inappropriate convergence parameters, d) some other incorrect parameters.
Note: normally MSR1a using run**_lapw -min is better than min_lapw.

_____
Professor Laurence Marks
"Research is to see what everybody else has seen, and to think what nobody else has thought", Albert Szent-Gyorgi
www.numis.northwestern.edu

On Wed, Oct 10, 2018, 6:46 AM Lawal Mohammed <***@yahoo.com> wrote:

Dear Developers and Users,
I need help on these three problems.
1) I tried to plot band character by editing case.insp as explained in the user guide, but band character for the selected orbital is not showing. I run a TiC example as test-case and it works fine, what could have been the problem?

2) I run Mg(OH)2 exercise (as contained in the workshop documents) for the optimization of positions. I noticed that the primitive cell and the conventional cell of theMg(OH)2 is the same when viewed using xcrysden, is this always the case H lattice or there are some exceptions?
3) Please is this command min_lapw -i 150 -p-NI correct? I tried it for non spin polarized case, but the scf terminated at 40 circles with forces not converged. I already have previous scf converged for volume optimization.

Thanks for your time, looking forward for help.
With kind regards.

Lawal

[Gavin Abo]

Post by Lawal Mohammed
1) I tried to plot band character by editing case.insp as explained in
the user guide, but band character for the selected orbital is not
showing. I run a TiC example as test-case and it works fine, what
could have been the problem?

The band character plot using the default "line switch" of 0.2 works
well for TiC but it does not always work well for other structures [
https://www.mail-archive.com/***@zeus.theochem.tuwien.ac.at/msg16414.html
]. Have you tried increasing the "line switch" value?

[Gavin Abo]

Sorry, a mistake in the previous post, but info at link is correct.  In
other words:

Check in case.insp that "line switch" is set to 2 and try increasing the
"size of heavier plotting" (much larger than the default value of 0.2
from SRC_templates and/or jsize in section "3.11.5 Bandstructure with
band character plotting / full lines" on page 27 of the WIEN2k 18.2
usersguide).

Corrected wording below.

Post by Lawal Mohammed
1) I tried to plot band character by editing case.insp as explained
in the user guide, but band character for the selected orbital is not
showing. I run a TiC example as test-case and it works fine, what
could have been the problem?

The band character plot using the default "size of heavier plotting"
of 0.2 works well for TiC but it does not always work well for other
structures [
]. Have you tried increasing the "size of heavier plotting" value?

[Lawal Mohammed]

Dear Gavin,
Thank you very much for your suggestion. Yes, I have tried different size of heavier plotting but in vain. What I observed was, the program does not read/recognise my jtype input for my case, because on w2web only atom number and size are printed. However, I can plot the band structure without editing the band character options.
Dear Prof Laurence Marks,I am sorry for going back, in my previous mail I used -i 150, but the scf circles terminate at 40 iterations. Why should it stops at 40 while it's suppose to reach 150, because of -i 150?
Many thanks for your time.
With kind regards.
Lawal



On Thursday, October 11, 2018, 9:29:38 AM GMT+8, Gavin Abo <***@crimson.ua.edu> wrote:


1) I tried to plot band character by editing case.insp as explained in the user guide, but band character for the selected orbital is not showing. I run a TiC example as test-case and it works fine, what could have been the problem?

The band character plot using the default "line switch" of 0.2 works well for TiC but it does not always work well for other structures [https://www.mail-archive.com/***@zeus.theochem.tuwien.ac.at/msg16414.html ]. Have you tried increasing the "line switch" value?

_______________________________________________
Wien mailing list
***@zeus.theochem.tuwien.ac.at
http://zeus.theochem.tuwien.ac.at/mailman/listinfo/wien
SEARCH the MAILING-LIST at:  http://www.mail-archive.com/***@zeus.theochem.tuwien.ac.at/index.html

[Lawal Mohammed]

Dear Prof. Peter,
Thanks a lot. I got it now.
With kind regards
Lawal

Post by Lawal Mohammed
Thank you very much for your suggestion. Yes, I have tried different
size of heavier plotting but in vain. What I observed was, the program
does not read/recognise my jtype input for my case, because on w2web
only atom number and size are printed. However, I can plot the band
structure without editing the band character options.

2 possibilities:
a) your case.insp file is wrong
b) you did not run the    x lapw2 -qtl step properly.

Post by Lawal Mohammed
I am sorry for going back, in my previous mail I used *-i 150*, but the
scf circles terminate at 40 iterations. Why should it stops at 40 while
it's suppose to reach 150, because of -i 150?

For min_lapw the -i 150 switch means  150 geometry steps.
In each geometry step a  "run_lapw -i 40" is issued by default (unless
you use -j ....) and apparently in your case the 40 scf cycles do not
give converged forces.

As was mentioned before: usually run_lapw -min  (using MSR1a) is much
more efficient then then the old min_lapw, but of course, if you have
enough time to wait, you can still use the old way.

Post by Lawal Mohammed
Many thanks for your time.
With kind regards.
*/Lawal
/*
On Thursday, October 11, 2018, 9:29:38 AM GMT+8, Gavin Abo

Post by Lawal Mohammed
1) I tried to plot band character by editing case.insp as explained in
the user guide, but band character for the selected orbital is not
showing. I run a TiC example as test-case and it works fine, what
could have been the problem?

The band character plot using the default "line switch" of 0.2 works
well for TiC but it does not always work well for other structures [
]. Have you tried increasing the "line switch" value?
_______________________________________________
Wien mailing list
http://zeus.theochem.tuwien.ac.at/mailman/listinfo/wien
_______________________________________________
Wien mailing list
http://zeus.theochem.tuwien.ac.at/mailman/listinfo/wien