Dear colleagues,
I am having troubles using anaddb to obtain relaxed ion elastic constants. Running anaddb ends with a note:
** On entry to ZHPEV parameter number 7 had an illegal value
Could anyone help me and explain what this error means and give me suggestions how to continue my calcs?
I read that it is connected with zhpev.f90 routine which diagonalizes matrices, but nothing more.
My rf elastic clalc ends without warnings and I merged ddbs from datasets 1 (gs) and 3 (rf) before running anaddb.
I am interested in metallic Ce, and using abinit 5.8.4 on a Linux cluster.
Thank you all in advance!
Yours,
Igor Lukacevic
PS. Input file is attached
Error while running anaddb on elastics
Moderators: mverstra, joaocarloscabreu
-
- Posts: 271
- Joined: Sat Jan 16, 2010 12:05 pm
- Location: Dept. of Physics, University J. J. Strossmayer, Osijek, Croatia
- Contact:
Error while running anaddb on elastics
- Attachments
-
- t11.in
- Response function input
- (2.32 KiB) Downloaded 385 times
-
- telast_3.in
- anaddb input
- (390 Bytes) Downloaded 333 times
Re: Error while running anaddb on elastics
Hello Igor,
argument 7 is the dimension of the eigenvalue matrix. I see that
src/77_ddb/piezo9.F90
has a call with arg7 = 3*natom-3 = 0 in your case. I believe the piezo constants have no meaning if you only have 1 atom (no internal degrees of freedom to relax). There should be a check in the code to avoid this, but basically it probably means there is no "relaxed ion" you can define here.
ciao!
Matthieu
argument 7 is the dimension of the eigenvalue matrix. I see that
src/77_ddb/piezo9.F90
has a call with arg7 = 3*natom-3 = 0 in your case. I believe the piezo constants have no meaning if you only have 1 atom (no internal degrees of freedom to relax). There should be a check in the code to avoid this, but basically it probably means there is no "relaxed ion" you can define here.
ciao!
Matthieu
Matthieu Verstraete
University of Liege, Belgium
University of Liege, Belgium
-
- Posts: 271
- Joined: Sat Jan 16, 2010 12:05 pm
- Location: Dept. of Physics, University J. J. Strossmayer, Osijek, Croatia
- Contact:
[SOLVED] Re: Error while running anaddb on elastics
Dear Matthieu,
thank You for the explanation. That solves the issue. Actually, a part of this answer was given to me last week by Razvan, but I hadn't time to put it on the forum.
Cheers!
Igor L.
thank You for the explanation. That solves the issue. Actually, a part of this answer was given to me last week by Razvan, but I hadn't time to put it on the forum.
Cheers!
Igor L.
-
- Posts: 2
- Joined: Sun Jun 13, 2010 6:26 am
Re: Error while running anaddb on elastics
You can try the unit cell. With more than two atoms per cell, anddb works well!
regards
alejandro
regards
alejandro