N.N.B., instead of a smaller ecut one can use .lcore. Sometimes this also
improves convergence.
_
Professor Laurence Marks
"Research is to see what everybody else has seen, and to think what nobody
else has thought", Albert Szent-Gyorgi
www.numis.northwestern.edu
On Nov 13, 2017 2:12 PM,
"besides the statement: be careful with the energy parameter (very fine
E-search in case.in1)"
N.B., one does not need a fine search in the more recent Wien2k, only in
the older versions.
I seem to remember that GMax 20 mainly cancels out for LDA/GGA -- I would
always use it (or more) for a meta
1- Computing the atomic energies of atoms like N and P in an FCC cell is
ok, however for O atom the high symmetry of the FCC cell results in 1/3
occupancies (for the 4th p electron of O) in the spin down case. Only
using a lower symmetry cell (orthorhombic) for O atom eliminates this issue.
Good to see more details added to the FAQ on cohesive/formation energies. I
have two suggestions for the new FAQ on cohesive energies
1- Computing the atomic energies of atoms like N and P in an FCC cell is
ok, however for O atom the high symmetry of the FCC cell results in 1/3
occupancies (for
Thanks Sir for the nice explanation and introducing one more factor "IFFT".
Sometimes we get XCOPT error and we set it 0 0 0 2. So I was doing wrong. If I
was doing three compounds A, B, and C then let for A, and B everything was okay
but for C the linear dependency failed then I set IFFT to
Let me add one important thing to what Peter sent. My experience is that
the atomic positions are only weakly sensitive to the exact parameters
used, although this may not be true if you are doing phonons. Hence it has
always worked well for me to calculate (including MSR1a relaxations) using
Thanks for the correction.
With respect to your other questions:
In order to save cpu time, I would do (at least for the more complicated
compounds, not necessarily for BaO, which is fast anyway) all
calculations with "standard RMTs" (volume optimization, optimization of
internal parameters,
Thank you very much Sir for a very nice NOTE.
I suggest one correction and two more queries:
Correction: The RKmax 5.6 in the sentence "Then you should do atom A with
RKmax=7/2.5*2.0=5.6" should be replaced by 7/2*2.5=8.75!!
Query: 1. I could not understand what is the purpose of another
Please check the faq pages at www.wien2k.at/reg_user/faq.
I've made a couple of modifications and there is a page for cohesive and
formation energies.
Any suggestions are welcome.
Peter Blaha
On 11/13/2017 11:50 AM, Peter Blaha wrote:
You must modify your O2 struct file.
Make the 2
You must modify your O2 struct file.
Make the 2 oxygens "equivalent", i.e. you should have only ONE
non-equivalent atomic site (NAT=1), but MULT=2 for the 2 oxygens.
This will produce inversion symmetry and all calculations will take less
time and memory !!!
And yes:
Dear Prof. Alay/Peter/Marks/Stefaan or any expert user
As you know molecule calculations take much time (I received your reply to my
previous query) and as suggested by you all these molecules need some good HPC
facility that I do not have.
Therefore, to avoid the multiple heavy repetitive
11 matches
Mail list logo