Ok, Thank you Sir.
On Fri, 21 Jan 2022 at 01:31, Peter Blaha
wrote:
> Keep going. DIS is slowly decreasing,
>
> Difficult cases do not converge in 15 cycles.
> Am 20.01.2022 um 19:12 schrieb shamik chakrabarti:
>
> Dear Prof. Blaha,
>
>
> :DIS are given as
> DIS : CHARGE DISTANCE
For magnetic systems the option -diaghf is not recommended; this would be a bad
approximation.
From: Wien on behalf of shamik
chakrabarti
Sent: Thursday, January 20, 2022 6:51 PM
To: A Mailing list for WIEN2k users
Subject: Re: [Wien] Converging full hy
Keep going. DIS is slowly decreasing,
Difficult cases do not converge in 15 cycles.
Am 20.01.2022 um 19:12 schrieb shamik chakrabarti:
Dear Prof. Blaha,
:DIS are given as
DIS : CHARGE DISTANCE ( 0.1572817 for atom 6 spin 2)
0.0350296
:DIS : CHARGE DISTANCE ( 0.1
Dear Prof. Blaha,
:DIS are given as
DIS : CHARGE DISTANCE ( 0.1572817 for atom6 spin 2)
0.0350296
:DIS : CHARGE DISTANCE ( 0.1670837 for atom6 spin 2)
0.0363429
:DIS : CHARGE DISTANCE ( 0.1545864 for atom6 spin 2)
0.0349665
:DIS : CHARGE DISTANCE (
:DIS are given as
DIS : CHARGE DISTANCE ( 0.1572817 for atom6 spin 2)
0.0350296
:DIS : CHARGE DISTANCE ( 0.1670837 for atom6 spin 2)
0.0363429
:DIS : CHARGE DISTANCE ( 0.1545864 for atom6 spin 2)
0.0349665
:DIS : CHARGE DISTANCE ( 0.1373922 for atom
Dear Prof. Tran,
My system is an antiferromagnetic semiconductor & I am
interested in calculating voltage (obtained from ENE) & DOS. Please advice.
with regards,
On Thu, 20 Jan 2022 at 23:10, Tran, Fabien wrote:
> If your system has a band gap and is nonmagnetic, and you are in
Dear Prof. Blaha,
I have reached convergence using PRATT smoothly with
a convergence parameter 0.20. Although, I have switched back to MSR1 for
some cases, for some cases I have got error in mixer again after getting
back to MSR1?
Sir, whether a well converged solution with
If your system has a band gap and is nonmagnetic, and you are interested in the
electronic structure (band structure, DOS)
then you may consider to use the option -diaghf (see user's guide for details),
which consists of doing just one iteration on
top of a PBE calculation.
_
Dear Prof. Blaha,
All the other quantities viz.,:ENE :MMT :MMI, :FER are
converged and STABLE. I am giving the ENE for 10 iterations as below;
--- ENE ---
:ENE : ** TOTAL ENERGY IN Ry = -35538.99630350
:ENE : ** TOTAL ENERGY IN Ry = -35538.99
Nobody can help you for this specific case without running it themself,
and all "general guidlines" were given.
There are always some cases, where it could be difficult to reach better
convergence.
Eventually a better k-mesh or larger TEMP parameter can help.
Are other quantities converged a
MSR1a also optimizes positions. A possible mixer error is that spheres
start to overlap during this optimization. Read the error messages in
detail.
PRATT does NOT change atomic positions, but attempts only to converge
the electrons. Be aware: PRATT can with a small mixing give the
impression
Here are my two cents:
Amorphous structures is an advanced topic, but not impossible. Our group has
calculated the electronic structure of amorphous Se in the past with WIEN2k
(https://olegrubel.mcmaster.ca/publications/2012/Darbandi_JPCM_24_2012.pdf),
but we had to construct the structural mode
Dear Prof. Tran,
With alpha=0.05 we are getting maximum convergence
up to 0.01, however, if I switch alpha=0.25 we are again getting back to
the cc of 0.1. My query is, should we consider that a maximum charge
convergence of 0.1 to 0.01 is actually achievable with full hybr
Dear Wien2k users & experts,
In the case of structural optimization
(volume or c/a etc) of some structures, I have encountered an error viz.,
"error in mixer". However, if I switched the scheme MSR1a in case.inm to
PRATT, it converges smoothly. Now my queries are;
(1
14 matches
Mail list logo
