Thank you Dr. BlahaThis work well.All the best for you.
Sent from Yahoo Mail on Android
On Wed, Nov 15, 2023 at 0:32, Peter Blaha wrote:
Well, at the end it is exactly as I said:
Your manual RMT settings are very bad. One of the rules is, that
smallest and largest RMTs must not be too
Well, at the end it is exactly as I said:
Your manual RMT settings are very bad. One of the rules is, that
smallest and largest RMTs must not be too different.
With your spheres you get "effective" R(Si)Kmax of more than 12 and this
gives numerical linear dependency.
With your struct file:
Dr. Blahasometimes, semicore error appears in iteration3, sometime in 6, and
... (with changing rmts).
I tried with the struct attached here and the default init_lapw. After
finalized initialization, I changed RKm=7 (and Emax=3) in case.in1c attached
here, and run dstart. This time, semicore
Again, your message gets too big. You must delete the older content.
---
grep :DIS in case.scf:
:DIS : CHARGE DISTANCE ( 0.0122755 for atom3 spin 1)
0.0083335
:DIS : CHARGE DISTANCE ( 0.0117894 for atom3 spin 1)
Looking at your original case.in1, something went very unstable and that
led to the crash. I think some more information might help:
1. What functional are you using?
2. What k-mesh?
3. Spin polarized or not?
4. Are you minimizing positions with -min?
5. Did you change anything in case.in0 &
I suppose the modified l2main.F to SRC_lapw2 for WIEN2k 23.2 won't
affect this particular case, but if one wants to try it out to know for
sure, I have posted the patch file for it today at [1] or put the link
in the README at [1] that directs to the mailing list post with the
l2main.F.gz file
I used exactly the struct file you sent (with your RMTs, even if I would
not choose them this way).
init -prec 1n
run_lapw
save rkm6
edit case.in1 to change RKmax
run_lapw
no problems.
I do not understand, how you can get your error: (SELECT E-top/bottom
not found) when you just change
Yes, I use WIEN2k_23.2Let me know how are rmt of atoms in structure file which
you used.
Sent from Yahoo Mail on Android
On Tue, Nov 14, 2023 at 11:31, Peter Blaha wrote:
As I wrote before: I cannot reproduce this. For me it converges fine
even with RKmax=7. No errors. Thus, I don't
As I wrote before: I cannot reproduce this. For me it converges fine
even with RKmax=7. No errors. Thus, I don't know how to help you.
Are you using WIEN2k_23.2 ??
Am 13.11.2023 um 15:39 schrieb hajar.nejatipoor--- via Wien:
Dear Dr. Blaha
I act as you said, but the problem "LAPW2' -
Dear Dr. BlahaI act as you said, but the problem "LAPW2' - semicore
band-ranges too large, ghostbands" exists again!!LO for N-2s orbitals in
case.in1c were removed, but it worked just for rkm=6 (with or without LO for
N).I tried your way with changing rmt of atoms but the problem remained.
I
I tried your struct file, converged with RKM=6, saved, increased RKMax
to 7 and continued with run_lapw.
No problem. As expected with your RMTs rather small change from 6 to 7
and quick convergence.
You must have changed something else, like mixing a density with
different RMTs,
Dear Dr. BlahaThe way you proposed, just worked for RKm=6, and the error below
appeared for the case of RKm=7: 'SELECT' - no energy limits found for atom 1
L= 0
'SELECT' - E-bottom -520.0 E-top -520.0
It is worth to mention that since for Si-muffin tin
First of all, setrmt gives:
1 42.0 2.12 2.12
2 14.0 1.68 1.68
3 7.0 1.61 1.60
4 7.0 1.60 1.60
So the N radii are much larger and Si and Mo smaller.
It might be that the ghostband comes from N-2s, as the small RMT may not
allow for 2 radial functions. You could try to
Dear Wien2k developers and usersI want to converge RKmax for the monolayer
MoSi2N4.For Rkm=6, the scf is converged without problem, but when running the
scf with Rkm=7, the error "'LAPW2' - semicore band-ranges too large,
ghostbands" appears. I tried with changing muffin tin radii, but the
Dear Prof. Blaha,
With your advice the simulation is running
smoothly. I have changed the iqtlsave parameter to one again after getting
convergence with zero. However, after changing it to one the simulation
converges smoothly.
Thanks once again.
with kind regards,
On
For small C spheres, the setting of an LO for C-2s may lead to problems,
in particular when one is not yet self-consistent.
If you set iqtlsave to zero for some time, it is fine, but at the end of
the scf you should switch it back and check if the error is still
present. Also check for qtl-b
I may solve the problem by setting the iqtlsave parameter to zero in
case.in2. But upto 8th structure I have obtained converged SCF by setting
the iqtlsave to one. Is it proper to set it zero from the 9th structure
onwards?
On Fri, 18 Nov 2022 at 11:32, shamik chakrabarti
wrote:
> Dear Wien2k
Dear Wien2k users,
I have tried to optimized a structure with option
7 in x optimize. It ran up to 8th structure well. However during running
9th structure it stops with an error; .
" Semicore band ranges too large".
I have attached the structure.
Looking forward to
Dear WIEN2k users and experts,
I am working on perovskites materials containing 3d,4f atoms with
monoclinic
structure. When I have calculate using GGA functional, there is no any
error at any k-points and then applying spin orbit coupling (SOC) using 50
k-points, it is successfully converged
I think the parameters for the 141_I41/amd spacegroup in WIEN2k need to
be in the origin 2 setting, but it looks like you have entered them in
the origin 1 setting. You can use Bilbao Crystallographic Server's
SETSTRU [ http://www.cryst.ehu.es/cryst/setstru.html ] or program like
VESTA [
Dear All,
I am working on TiO2 Anatase and when i run SCF i get semicore band-ranges too
large error. Wien2k version is 13.1. I have attached my in1, struct, scf files.
Please help me on this.
Best regards,Temuujin Bayaraa
Researcher
Institute of Physics and Technology,
Mongolian Academy of
On 02.03.2015 08:20, bayarr temuujin wrote:
I am working on TiO2 Anatase and when i run SCF i get semicore
band-ranges too large error. Wien2k version is 13.1. I have attached my
in1, struct, scf files.
I cannot understand the problem.
But I have looked at the structure in xcrysden and IMHO it
Dear Dr.Blaha,
I'm working on alloy Ni2CoGa, and at FSM calculation get the semicore band
ranges too large.I see at scf2dn:
:GAP : -9. Ry = -. eV ( metallic )
Bandranges (emin - emax) and occupancy:
:WARN :BAN1: 1 -8.256368 -3.954598 1.
Is it possible
This happens if:
a) your structure is wrong (unphysical distances, RMTs, ) or
b) you got strong oszillations (:DIS) during the scf-procedure and some
ghostbands occured. This can also happen because the (automatic)
determination of E-parameters (case.scf1) of the semicore states did not
Dear Dr.Blaha,
In 3nd volume of volume optimize calculation, wien2k get semicore error
again.In my calculations RKmax=9 , Kpoint=1000 and ex-co potential is PBE-08.
For use full diagonalization what do I do?I'm attaching my struct and in1 and
scf1 files.
Best wishes,
Leyla
This happens if:
a)
I used your struct file,
init -b -rkmax 8
run
save rkm_8
edit case.in1c and increase rkmax to 9
run
no problems.
Maybe you start out first with a smaller rkmax and only in a second
round increase it to 9
On 10/14/2014 02:13 PM, Mh Mi wrote:
Dear Dr.Blaha,
In 3nd volume of volume optimize
Dear Dr.Blaha,
I am extremely grateful for your help, just your calculations are done to
optimize the volume?
On Tuesday, October 14, 2014 5:33 PM, Peter Blaha
pbl...@theochem.tuwien.ac.at wrote:
I used your struct file,
init -b -rkmax 8
run
save rkm_8
edit case.in1c and increase
No. I was just using the struct file you included.
On 10/14/2014 04:17 PM, Mh Mi wrote:
Dear Dr.Blaha,
I am extremely grateful for your help, just your calculations are done
to optimize the volume?
On Tuesday, October 14, 2014 5:33 PM, Peter Blaha
pbl...@theochem.tuwien.ac.at wrote:
I
Dear all
I'm working on magnetic heusler alloys include Co,Ni,Fe,Ga.When I calculate
volume optimization ,get the semicore band ranges too large in 2nd volume.
I use muffin tin spheres Co,Fe,Ni=2a.u. , Ga=2.2a.u. ,and energy separation
between core and valence=-6Ry.Please help me for this
Use sphere sizes as determined by our setrmt utility.
On 10/08/2014 10:05 AM, Mh Mi wrote:
Dear all
I'm working on magnetic heusler alloys include Co,Ni,Fe,Ga.
When I calculate volume optimization ,get the semicore band ranges too
large in 2nd volume.
I use muffin tin spheres Co,Fe,Ni=2a.u.
Dear Dr.Balaha
Now,I used from reduce RMTs by 5% ,but that problem not fixed.Ni,Co=2.17a.u.
and Ga=2.12a.u.___
Wien mailing list
Wien@zeus.theochem.tuwien.ac.at
http://zeus.theochem.tuwien.ac.at/mailman/listinfo/wien
SEARCH the MAILING-LIST at:
It happens again at the second volume ??? And you restarted from the
beginning ??
Are your volume-steps sp large that the extrapolation does not work
anymore ?
In the worst case, you have to activate thex dstart lines and
comment the clmextrapol lines in optimize.job.
On
Dear all,
after switching to Wien2k_13, I frequently get the semicore band ranges
too large error, often during relaxation. Though I can understand why
that should not happen, can you explain what could have gone wrong that
it appears and how the source of this error can be traced?
Best regards,
This check was introduced recently to detect ghost bands.
If this problem appears at the very beginning of the SCF iteration
for a new geometry then there is probably no problem of ghost bands.
This is just the starting density which is not good.
To avoid the stop of the calculation, in case.in2
34 matches
Mail list logo
