[QE-users] question about last steps of vc-relax

[Konstantin Glazyrin]

Dear All,
I was playing with one structure 'vc-relax', tested that it converged to
the pressure I was aiming for.
If I understand correctly, after structural relaxation QE (e.g. step 1)
runs an additional 'scf' run (step 2) and recalculates stuff and in both
steps recalculates pressure.

It is interesting that at a single low pressure point, the pressure shown
in 1 - is very close to 2, while in higher pressure points the last step
shows lower pressure than 1 by some GPa.

An example - target relaxation pressure 600 kbar:
1 - total   stress  (Ry/bohr**3)                   (kbar)     P=
 600.65 - last run of structure relaxation
2 - total   stress  (Ry/bohr**3)                   (kbar)     P=
 484.28 - final scf run
May I ask - what does it indicate, which value I should trust more and
how should I proceed?

Is it because of this?
SCF correction compared to forces is large: reduce conv_thr to get better
values
For me it is conv_thr =   6.4000000000d-06

Thank you!
best wishes,
            Konstantin

-- 
Beamline Scientist
DESY, FS-PE, bldg. 47c, L115
Notkestrasse 85
22607 Hamburg, Germany
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.quantum-espresso.org/pipermail/users/attachments/20230904/ded16778/attachment.html>