"CPASSERT failed"---kpoint calculations/band structure with hybrid functional PBE0?

[wuli...@gmail.com]

Hi All,

I want to print the band structure of WO3 with hybrid functional PBE0 and I get the "CPASSERT failed  hfx_energy_potential.F:494" 

Routine Calling Stack : 9 integrate_four_center

The input file downloaded from https://www.cp2k.org/exercises:2018_uzh_cmest:pdos?s[]=band

And the functional PBE0 description are:

&XC                       ! this is the section to define the electronic exchange  
      &XC_FUNCTIONAL          ! our functional is hybrid 
         &PBE                 ! it has 75% of PBE 
           SCALE_X 0.75
           SCALE_C 1.0
         &END
         &PBE_HOLE_T_C_LR
           SCALE_X 0.25       ! + 25% of truncated PBE0 functional - that includes exact hfx
           CUTOFF_RADIUS 3.5  ! that has interaction truncated at 3.5 A from the atomic core 
         &END
      &END XC_FUNCTIONAL
      &HF
        FRACTION 0.25         ! this is the hfx section. The amount of hfx must be consistent with above
        &SCREENING            ! Screening of the electronic repulsion up to the given threshold. This section is needed
          EPS_SCHWARZ 1.0E-6
          SCREEN_ON_INITIAL_P TRUE  ! having an external wave-function, a preliminary screening
        &END                        ! can be performed to speed up calculations
        &MEMORY
          MAX_MEMORY 800            ! uses some memory to store data and not recompute each step
        &END
        &INTERACTION_POTENTIAL      ! Sets up interaction potential between the two regions
          POTENTIAL_TYPE TRUNCATED  ! the potential is  truncated 
          CUTOFF_RADIUS 3.5         ! at 3.5 A (see above)
          T_C_G_DATA ./t_c_g.dat    ! external file with parameters needed to truncate the potential
        &END
      &END

Thanks!

Lina

[wuli...@gmail.com]

The cp2k version is 5.1, SVN source code revision svn: 18091

[Zhendong Guo]

Dear Lina
  Please try a smaller  EPS_PGF_ORB   for instance, 1.0E-50
Best regards
zhendong

On Sat, Nov 10, 2018 at 3:59 AM <wuli...@gmail.com> wrote:

The cp2k version is 5.1, SVN source code revision svn: 18091

--
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 cp2k+uns...@googlegroups.com.
To post to this group, send email to cp...@googlegroups.com.
Visit this group at https://groups.google.com/group/cp2k.
For more options, visit https://groups.google.com/d/optout.

--

Zhendong Guo 

Phone: +41786943316

Email:  zhendon...@gmail.com

            zhendo...@epfl.ch-------------------------------------------------------------------------------------------------

[Zdenek Futera]

Dear Lina,

I don't think that K-point sampling is available for hybrid functionals.

It's probably not implemented yet.

Best,

Zdenek

[Matt W]

K-point sampling is not available for hybrid functionals.

[Li-Na Wu]

Thank you!

Matt W <mattwa...@gmail.com> 于2018年11月11日周日 上午5:38写道：

--

[wuli...@gmail.com]

Dear Zhendong,

Thank you. It seems  that K-point sampling is not available for hybrid functionals.

Bests,

Lina

在 2018年11月10日星期六 UTC+8下午2:42:17，Zhendong Guo写道：

[wuli...@gmail.com]

Thanks a lot!

 Bests,

Lina

在 2018年11月10日星期六 UTC+8下午6:05:39，Zdenek Futera写道：

[wuli...@gmail.com]

Hi Prof. Matt,

Thanks for your reply. 

Another question: Is SOC calculation available in CP2K for now?

Bests,

Lina

[Matt W]

Unfortunately no Spin Orbit Coupling available at the moment.

[wuli...@gmail.com]

Thanks a lot!

在 2018年11月11日星期日 UTC+8下午9:23:42，Matt W写道：