[CP2K-user] [CP2K:12534] Unable to converge for perovskite calculation

Jiapeng Liu sunta... at gmail.com
Mon Nov 25 04:44:10 UTC 2019
Previous message (by thread): [CP2K-user] [CP2K:12534] Unable to converge for perovskite calculation
Next message (by thread): [CP2K-user] [CP2K:12536] Unable to converge for perovskite calculation
Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]
Dear Matthias,

Thank you very much. I used a larger cell and it converged successfully. 
May I ask you another question that is it possible to use CP2K to conduct a 
small cell calculation with no KPOINTS as in my previous case. Since I run 
the test in my desktop, too many atoms are not manageable. Thank you.

Regards,
Jiapeng

On Monday, November 25, 2019 at 1:00:08 AM UTC+8, Matthias Krack wrote:
>
> Hi
>
>  
>
> That result is not a surprise, since you did not follow my first and most 
> important suggestion which was to employ multiple unit cells.
>
>  
>
> Matthias
>
>  
>
> *Von:* c... at googlegroups.com <javascript:> <c... at googlegroups.com 
> <javascript:>> * Im Auftrag von *Jiapeng Liu
> *Gesendet:* Sonntag, 24. November 2019 15:42
> *An:* cp2k <c... at googlegroups.com <javascript:>>
> *Betreff:* Re: [CP2K:12534] Unable to converge for perovskite calculation
>
>  
>
> Dear Matthias,
>
>  
>
> Thank you for your reply. I tried with higher smearing temperature (600K) 
> together with the default GPW method, the scf calculation accelerates with 
> a short time. However, the scf still does not converge and the energy does 
> not decrease monotonically. The energy just oscillates. I am not sure if 
> this is normal? Also, does higher smearing temperature (600K) affect the 
> final energy? Thank you.
>
>
> On Friday, November 22, 2019 at 4:47:13 PM UTC+8, Matthias Krack wrote:
>
> Hi
>
>  
>
> Without k points you will need a larger simulation cell, i.e. multiple 
> unit cells, e.g. 4x4x2. A larger electronic temperature for the smearing 
> might also help initially. With a larger cutoff, it might also be possible 
> to use the default GPW method instead of GAPW in the beginning. 
>
>  
>
> Matthias
>
>  
>
> *From:* c... at googlegroups.com <c... at googlegroups.com> *On Behalf Of *Jiapeng 
> Liu
> *Sent:* Freitag, 22. November 2019 01:47
> *To:* cp2k <c... at googlegroups.com>
> *Subject:* [CP2K:12529] Unable to converge for perovskite calculation
>
>  
>
> Dear CP2K users,
>
>  
>
> I repost the problem with the expectation that anyone can help me to solve 
> this. I would like to perform the DFT+U calculations for a small perovskite 
> cubic cell, which contains 10 atoms in total. As a first step suggested by 
> this group, I conduct the DFT without U firstly. However, the scf fails to 
> converge even after 3 outer scf each with 200 steps. I am quite confused 
> about this. Can anyone help me with this, your suggestions or advice are 
> significantly appreciated. Thank you. Following is the input file
>
>  
>
> !*************************GLOBAL SETTINGS**********************************
>
> &GLOBAL
>
>   PRINT_LEVEL LOW
>
>   PROJECT_NAME LCTN_exsolution
>
>   RUN_TYPE CELL_OPT
>
>   SAVE_MEM T
>
> &END GLOBAL
>
>  
>
> !*************************MOTION SETTINGS**********************************
>
> &MOTION
>
>   &CELL_OPT
>
>     OPTIMIZER LBFGS
>
>     TYPE DIRECT_CELL_OPT
>
>   &END CELL_OPT
>
> &END MOTION
>
>  
>
> !************************FORCE_EVAL 
> SETTINGS********************************
>
> &FORCE_EVAL
>
>   STRESS_TENSOR ANALYTICAL
>
>   METHOD QS
>
> !********************FORCE_EVAL -> DFT 
> SETTINGS*****************************
>
>   &DFT
>
>     BASIS_SET_FILE_NAME BASIS_MOLOPT_UCL
>
>     BASIS_SET_FILE_NAME BASIS_MOLOPT
>
>     POTENTIAL_FILE_NAME GTH_POTENTIALS
>
>     UKS T
>
>     CHARGE 0
>
>     MULTIPLICITY 2
>
> !    PLUS_U_METHOD LOWDIN
>
> !****************FORCE_EVAL -> DFT -> SCF 
> SETTINGS**************************
>
>     &SCF
>
>       ADDED_MOS 100
>
>       MAX_SCF 200
>
>       EPS_SCF 1.0E-6
>
>       SCF_GUESS RESTART
>
>       &DIAGONALIZATION T
>
>         ALGORITHM STANDARD
>
>       &END DIAGONALIZATION
>
>       &OUTER_SCF
>
>         EPS_SCF 1.0E-5
>
>         MAX_SCF 5
>
>       &END OUTER_SCF
>
>       &SMEAR T
>
>         METHOD FERMI_DIRAC
>
>         ELECTRONIC_TEMPERATURE [K] 300
>
>       &END SMEAR
>
>       &MIXING T
>
>         METHOD BROYDEN_MIXING
>
>         ALPHA 0.4
>
>         BETA [bohr^-1] 2
>
>         NMIXING 5
>
>         NBUFFER 8
>
>       &END MIXING
>
>     &END SCF
>
> !****************FORCE_EVAL -> DFT -> QS 
> SETTINGS***************************
>
>     &QS
>
>       EPS_DEFAULT 1.0E-12
>
>       EXTRAPOLATION USE_GUESS
>
>       METHOD GAPW
>
>     &END QS
>
> !***************FORCE_EVAL -> DFT -> GRID 
> SETTINGS**************************
>
>     &MGRID
>
>       NGRIDS 5
>
>       CUTOFF [Ry] 600
>
>       REL_CUTOFF [Ry] 60
>
>     &END MGRID
>
> !*****************FORCE_EVAL -> DFT -> XC 
> SETTINGS**************************
>
>     &XC
>
>       FUNCTIONAL_ROUTINE NEW
>
>       &XC_FUNCTIONAL PBE
>
>       &END XC_FUNCTIONAL
>
>     &END XC
>
> !***************FORCE_EVAL -> DFT -> POISSON 
> SETTINGS************************
>
>     &POISSON
>
>       POISSON_SOLVER PERIODIC
>
>       PERIODIC XYZ
>
>     &END POISSON
>
>   &END DFT
>
> !********************FORCE_EVAL -> SUBSYS 
> SETTINGS***************************
>
>   &SUBSYS
>
>     &CELL
>
>       ABC [angstrom] 3.857 3.857 7.714
>
>       ALPHA_BETA_GAMMA [deg] 90 90 90
>
>       PERIODIC XYZ
>
>       MULTIPLE_UNIT_CELL 1 1 1
>
>     &END CELL
>
>     &KIND La
>
>       ELEMENT La
>
>       BASIS_SET DZVP-MOLOPT-SR-GTH
>
>       POTENTIAL GTH-PBE-q11
>
>     &END KIND
>
>     &KIND Ca
>
>       ELEMENT Ca
>
>       BASIS_SET DZVP-MOLOPT-SR-GTH
>
>       POTENTIAL GTH-PBE-q10
>
>     &END KIND
>
>     &KIND Ti
>
>       ELEMENT Ti
>
>       BASIS_SET DZVP-MOLOPT-SR-GTH
>
>       POTENTIAL GTH-PBE-q12
>
> !      MAGNETIZATION 3.0
>
>     &END KIND
>
>     &KIND Ni
>
>       ELEMENT Ni
>
>       BASIS_SET DZVP-MOLOPT-SR-GTH
>
>       POTENTIAL GTH-PBE-q18
>
> !      MAGNETIZATION 3.0
>
> !      &DFT_PLUS_U T
>
> !        L 2
>
> !        U_MINUS_J [eV] 6.4
>
> !      &END DFT_PLUS_U
>
>     &END KIND
>
>     &KIND O
>
>       ELEMENT O
>
>       BASIS_SET DZVP-MOLOPT-SR-GTH
>
>       POTENTIAL GTH-PBE-q6
>
>     &END KIND
>
>     &TOPOLOGY
>
>       COORD_FILE_NAME lctn_112.xyz
>
>       COORD_FILE_FORMAT XYZ
>
>       MULTIPLE_UNIT_CELL 1 1 1
>
>     &END TOPOLOGY
>
>   &END SUBSYS
>
>  
>
> !********************FORCE_EVAL -> PRINT 
> SETTINGS****************************
>
>   &PRINT
>
>     &FORCES ON
>
>     &END FORCES
>
>   &END PRINT
>
> &END FORCE_EVAL
>
>  
>
> -- 
> You received this message because you are subscribed to the Google Groups 
> "cp2k" group.
> To unsubscribe from this group and stop receiving emails from it, send an 
> email to c... at googlegroups.com.
> To view this discussion on the web visit 
> https://groups.google.com/d/msgid/cp2k/3fec0ca8-ab49-4fa2-9e9f-dde0988af57c%40googlegroups.com 
> <https://groups.google.com/d/msgid/cp2k/3fec0ca8-ab49-4fa2-9e9f-dde0988af57c%40googlegroups.com?utm_medium=email&utm_source=footer>
> .
>
> -- 
> You received this message because you are subscribed to the Google Groups 
> "cp2k" group.
> To unsubscribe from this group and stop receiving emails from it, send an 
> email to cp... at googlegroups.com <javascript:>.
> To view this discussion on the web visit 
> https://groups.google.com/d/msgid/cp2k/4ab70dea-43f7-40e2-aa10-0657b4869474%40googlegroups.com 
> <https://groups.google.com/d/msgid/cp2k/4ab70dea-43f7-40e2-aa10-0657b4869474%40googlegroups.com?utm_medium=email&utm_source=footer>
> .
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.cp2k.org/archives/cp2k-user/attachments/20191124/aab2249d/attachment.htm>
Previous message (by thread): [CP2K-user] [CP2K:12534] Unable to converge for perovskite calculation
Next message (by thread): [CP2K-user] [CP2K:12536] Unable to converge for perovskite calculation
Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]
More information about the CP2K-user mailing list