Running optimization and EBS in same run (different dataset)

Total energy, DFT+U, BigDFT,...

Moderator: amadon

Post Reply
Posts: 12
Joined: Thu Jun 16, 2016 8:47 am
Location: Basel, CH

Running optimization and EBS in same run (different dataset)

Post by marco.digennaro » Fri Aug 12, 2016 5:32 pm

Dear all,

I am now running a calculation (quite basic primitive Bcc iron), and want to calculate, say, EBS.
Normally, I would run the optimization (OPT) first and once finished use the optimized lattice parameter in a second run with 2 dataset (GS+EBS).
Is there a way to do it in the same run?
If you append a second dataset (EBS) directly to the OPT one, abinit will crash,
since the density is printed on a *DS1_TIM${N}_DEN instead on a *_DS1_DEN file.

I therefore thought to make one job with 3 dataset instead, the first one with geometry optimization and the others with the rest of the calculations.
The optimization goes fine, but then abinit uses the cubic unit cell instead of the primitive in the 2nd dataset (GS), while the corrected one in the last one.
Therefore the pressure exploses in the GS and I am not sure the rest of the calculation is correct:

Code: Select all

grep acell log.16948468 
            acell      6.2549934999E+00  6.2549934999E+00  6.2549934999E+00 Bohr
 Scale of Primitive Cell (acell) [bohr]
Scale of Primitive Cell (acell) [bohr]
            acell1     5.1868263033E+00  5.1868263033E+00  5.1868263033E+00 Bohr
            acell2     6.2549934999E+00  6.2549934999E+00  6.2549934999E+00 Bohr
            acell3     6.2549934999E+00  6.2549934999E+00  6.2549934999E+00 Bohr

grep Press log.16948468
-Cartesian components of stress tensor (GPa)         [Pressure= -1.1134E-03 GPa]
-Cartesian components of stress tensor (GPa)         [Pressure= -2.9172E+01 GPa]

Thank you in advance.
Marco Di Gennaro
Marvel, University of Basel (CH)

Post Reply