Hello,
it appears that the Hessian matrix is not printed in the OUTCAR file when IBRION=6 (i.e. symmetry) is used. The line SECOND DERIVATIVES (NOT SYMMETRIZED) is missing in the OUTCAR. I'd assume that VASP would fill the matrix based on the symmetry of the system. This is what I get from the VASP wiki: For IBRION=6, only symmetry inequivalent displacements are considered, and the remainder of the Hessian matrix is filled using symmetry consideration
Could someone please give some advice how to print the Hessian in the OUTCAR in this case?
Thanks
Tobias
Hessian matrix IBRION=6
Moderators: Global Moderator, Moderator
-
Tobias Kraemer
- Newbie
- Posts: 10
- Joined: Wed Mar 29, 2017 12:45 pm
- License Nr.: 5-2661
-
alex
- Hero Member
- Posts: 585
- Joined: Tue Nov 16, 2004 2:21 pm
- License Nr.: 5-67
- Location: Germany
Re: Hessian matrix IBRION=6
Hi Tobias,
it looks like it got lost over time. Go to the file finite_diff.F, search for 'clumsy and unreadable output', make the 7 lines active and recompile.
It might solve your issue.
Cheers,
alex
it looks like it got lost over time. Go to the file finite_diff.F, search for 'clumsy and unreadable output', make the 7 lines active and recompile.
It might solve your issue.
Cheers,
alex
-
Tobias Kraemer
- Newbie
- Posts: 10
- Joined: Wed Mar 29, 2017 12:45 pm
- License Nr.: 5-2661
Re: Hessian matrix IBRION=6
Alex,
thanks for your suggestion. This is indeed rather strange, but I will have a look into this and perhaps this will remedy the problem.
I don't understand why it normally works just fine for IBRION=5.
Tobias
thanks for your suggestion. This is indeed rather strange, but I will have a look into this and perhaps this will remedy the problem.
I don't understand why it normally works just fine for IBRION=5.
Tobias
-
alex
- Hero Member
- Posts: 585
- Joined: Tue Nov 16, 2004 2:21 pm
- License Nr.: 5-67
- Location: Germany
Re: Hessian matrix IBRION=6
Hi Tobias,
in the early days of VASP 4.6 the hessian was printed twice, one as is and then symmetrized. However at this time, symmetry was not used during the calculation. Maybe some of the developers had the opinion - hm, we don't need the symmetrized version of a symmetric hessian anyway and just removed the printing, leaving us with no hessian at all. :-)
We are all humans...
Cheers,
alex
in the early days of VASP 4.6 the hessian was printed twice, one as is and then symmetrized. However at this time, symmetry was not used during the calculation. Maybe some of the developers had the opinion - hm, we don't need the symmetrized version of a symmetric hessian anyway and just removed the printing, leaving us with no hessian at all. :-)
We are all humans...
Cheers,
alex