Caveat: you have to modify case.in2, not case.in2eece. Add a few spaces
(2-3) before the 452, and then do runeece -p as this just does the part
where case.in2eece is created.
On Tue, Jan 26, 2016 at 2:51 PM, Hu, Wenhao wrote:
> Hi, Prof. Laurence:
>
> By following your suggestion, It seems that
Hi, Prof. Laurence:
By following your suggestion, It seems that everything is normal except for the
case.in2eece file. My case.in2eece is as follows:
TOT EECE (TOT,FOR,QTL,EFG,FERMI)
-12.0 52.00 0.50 0.05 1 EMIN, NE, ESEPERMIN, ESEPER0, iqtlsave
TETRA0.000 (GAUSS,ROOT,TEMP,TETRA,ALL
Hard to say. Thoughts:
Check case.scf2up/dn and look at the Cr d-occupancy
grep :EORB *scforb*
If everything is zero then no orbital terms were being applied.
Check case.vorbup/dn -- non zero terms?
Check case.scforbup/dn
Last line should be
PRATT mixing scheme with 1.000
head case.in2
Hi, Prof. Laurence:
No, I don’t include the SO effect. The orbital potential is applied in lapw1.
Basically, I’m using the -eece flag of runs_lapw. The program flow is as
follows:
> (runeece_lapw) options: -p
Sun Jan 24 14:24:18 CST 2016> (x) lapwdm -up -p -c
Sun Jan 24 14:24:40 CST 2016> (x
That looks converged. Are you using -so, or is the orbital potential being
applied in lapw1? (Look for lapw1 -up -orb lines in :log).
On Tue, Jan 26, 2016 at 11:43 AM, Hu, Wenhao wrote:
> Hi, Prof. Laurence:
>
> This result is converged to large extent with respect to force, charge and
> energy.
Hi, Prof. Laurence:
This result is converged to large extent with respect to force, charge and
energy. The forces on individual atoms are all below 0.5 and the total force in
z is 3.5, which is the direction without symmetry. The :MV flag gives the
following information:
:MVORD NDM 100 L1
Did you fully converge the calculation or just do one iteration? You have
to fully converge. Also, what does "grep :MV case.scf | tail" give?
On Tue, Jan 26, 2016 at 11:08 AM, Hu, Wenhao wrote:
> Hi, all:
>
> I have a basic question on the DOS calculation of onsite-hybrid functional
> (ONF) calc
7 matches
Mail list logo