[Wien] Problem with geometry minimzation
Saeed Bahramy
bahramy at riken.jp
Tue Nov 2 14:16:44 CET 2010
Thanks for your comments.
As for your questions:
1. Yes, I've read the UG as well as the optimization notes!
2. In my calculations, the muffin tins are all well separated and
there is enough space between them for the sake of atomic relaxation.
3. I've tried different RKMAX values: 7~10
4. BZ sampling is done by a 10x10x5 k-mesh
5. The core charge leakage is less than 0.00001 e
So the problem must be somewhere else!
Seeed Bahramy
On Nov 2, 2010, at 10:02 PM, Laurence Marks wrote:
> This almost always means that the user (you) has set the IDFT problem
> up incorrectly -- GIGO.
>
> Have you read the optimization notes at
> http://www.wien2k.at/reg_user/textbooks/ ? Do you have almost touching
> spheres, bad RKMAX, k-points etc?
>
> On Tue, Nov 2, 2010 at 1:53 AM, Saeed Bahramy <bahramy at riken.jp>
> wrote:
>> Dear all,
>>
>> I'm trying to do geometry minimization for atoms inside a hexagonal
>> unitcell. For all atoms, I have set RMT to 2.5 a.u. and the
>> minimization has
>> been executed using the following command:
>> min_lapw -I -j "run_lapw -I -fc 1.0 -ec 0.0001 -i 40"
>>
>> After few steps, the job stops with the following warning message:
>> :WARNING you might have inconsistent Forces/Energies
>>
>> The forces in the case.scf_mini look like:
>>
>>
>> ******************************************************
>> * MULTISECANT MIXING
>> OPTIONS
>> *
>> * NFL Limit 0.100000
>> *
>> * Diag Parm 0.000500
>> *
>> * Standard Increase/
>> Decrease
>> Bounds *
>> * Plane Wave SQRT Dynamic
>> Rescaling *
>> * Max Number of Memory
>> Steps 8
>> *
>>
>> ******************************************************
>>
>> :PLANE: INTERSTITIAL TOTAL
>> 16.38294
>> DISTAN 0.239D-05 %
>> :CHARG: CLM CHARGE
>> TOTAL 8.54710
>> DISTAN 0.270D-06 %
>> :REDuction and DMIX
>> in Broyd:
>> 1.1000 0.2000
>> :INFO : Dynamic rescale
>> 0.718D+00
>> 0.100D+01 0.500D+01
>> :INFO : Number of
>> Memory
>> Steps 3 Skipping 0
>> :INFO : Bounds 0.304D-01
>> 0.200D+00
>> 0.108D+00 0.304D-01
>> :DIRM : MEMORY 3/8 RESCALE 0.718 RED
>> 0.983
>> PRED 0.993 NEXT 0.923
>> :INFO : DMIXM and
>> Projections
>> 0.304D-01 0.923D+00
>> :DIRP : |BROYD|= 0.488D-06 |PRATT|=
>> 0.725D-07
>> ANGLE= 41.0 DEGREES
>> :DIRB : |BROYD|= 0.495D-06 |PRATT|=
>> 0.824D-07
>> ANGLE= 44.6 DEGREES
>>
>> MSEC1
>> MIXING SCHEME WITH 0.030
>>
>> .....
>>
>>
>> :ENE : ********** TOTAL ENERGY
>> IN Ry =
>> -115360.01623084
>>
>> TOTAL FORCE IN mRy/a.u. = |F| Fx
>> Fy Fz
>> with/without FOR
>> :FOR001: 1.ATOM 32.478 0.000 0.000
>> 32.478 total forces
>> :FOR002: 2.ATOM 0.094 0.000 0.000
>> -0.094 total forces
>> :FOR003: 3.ATOM 0.687 0.000 0.000
>> 0.687 total forces
>>
>> TOTAL FORCE WITH RESPECT TO
>> GLOBAL
>> CARTESIAN COORDINATES:
>> :FCA001: 1.ATOM 0.000 0.000
>> 32.478 total forces
>> :FCA002: 2.ATOM 0.000 0.000
>> -0.094 total forces
>> :FCA003: 3.ATOM 0.000 0.000
>> 0.687 total forces
>>
>> TOTAL FORCE WITH RESPECT TO
>> THE
>> GLOBAL COORDINATE SYSTEM:
>> :FGL001: 1.ATOM 0.000000000 0.000000000
>> 32.478000000 total forces
>> :FGL002: 2.ATOM 0.000000000 0.000000000
>> -0.094000000 total forces
>> :FGL003: 3.ATOM 0.000000000 0.000000000
>> 0.687000000 total forces
>> :WARNING you might have inconsistent Forces/Energies
>>
>> As you can see, the forces on atom #2 and #3 are quite small , The
>> only
>> problem is atom #1 on which the total force is rather large.
>>
>> My initial case.inM is as follows
>>
>> PORT 2.00 0.35 # PORT/NEWT; tolf, Initial Trust Radius
>> 0.0 0.0 1.0 1.0 #Atom 1 Generated by pairhess
>> 0.0 0.0 1.0 1.0 #Atom 2 Generated by pairhess
>> 0.0 0.0 1.0 1.0 #Atom 3 Generated by pairhess
>>
>> I have done additional calculations using NEW1 and NEWT methods,
>> but so far
>> no success! No matter what method I use, the total force remains as
>> large as
>> ~30 mRy/a.u..
>>
>> The muffin tins are well separated from each other and there is
>> sufficient
>> distance between them for atomic relaxation.
>> I'm wondering if somebody know how to fix this problem.
>>
>> Thanks,
>> Saeed Bahramy
>>
>>
>>
>>
>> _______________________________________________
>> Wien mailing list
>> Wien at zeus.theochem.tuwien.ac.at
>> http://zeus.theochem.tuwien.ac.at/mailman/listinfo/wien
>>
>
>
>
> --
> Laurence Marks
> Department of Materials Science and Engineering
> MSE Rm 2036 Cook Hall
> 2220 N Campus Drive
> Northwestern University
> Evanston, IL 60208, USA
> Tel: (847) 491-3996 Fax: (847) 491-7820
> email: L-marks at northwestern dot edu
> Web: www.numis.northwestern.edu
> Chair, Commission on Electron Crystallography of IUCR
> www.numis.northwestern.edu/
> Electron crystallography is the branch of science that uses electron
> scattering and imaging to study the structure of matter.
> _______________________________________________
> Wien mailing list
> Wien at zeus.theochem.tuwien.ac.at
> http://zeus.theochem.tuwien.ac.at/mailman/listinfo/wien
More information about the Wien
mailing list