[CP2K:9130] cutoff value optimization question

[hut... at chem.uzh.ch]

Hi

this is a recurring question in this mailing list. Have a look at previous discussions for additional insight.

The problem is related to the handling of real space grids and the assignment of basis functions to these grids.
If you want to see a more or less smooth convergence using the single cutoff value you have to disable multiple
grids (NGRIDS 1). 
An expert usage of multiple grids with special settings of
- CUTOFF
- REL_CUTOFF
- NGRIDS
- MULTIGRID_CUTOFF
can also be used, but requires insight into the algorithms used.

regards

Juerg
--------------------------------------------------------------
Juerg Hutter                         Phone : ++41 44 635 4491
Institut für Chemie C                FAX   : ++41 44 635 6838
Universität Zürich                   E-mail: hut... at chem.uzh.ch
Winterthurerstrasse 190
CH-8057 Zürich, Switzerland
---------------------------------------------------------------

-----cp... at googlegroups.com wrote: -----To: cp2k <cp... at googlegroups.com>
From: Evgeniy Gr 
Sent by: cp... at googlegroups.com
Date: 06/20/2017 01:04PM
Subject: [CP2K:9130] cutoff value optimization question

Hi! I am a beginner in CP2K and I encountered with such problem: when I carried out the test CUTOFF value optimization in single point mode the total energy has significant changes even at large CUTOFF values (350-500 Ry). The REL_CUTOFF value is constant (60 Ry). For test example I chose the small maleimide molecule. I tried to enlarge the cell volume (ABC = 30, 30, 30), but for larger cell I obtained the same picture for TOTAL_E changes. What am I doing wrong? I will be very grateful for your advices. My ultimate goal is to study the mechanisms of organic reactions in a solvent medium by DFT molecular dynamics.CUTOFF (Ry) TOTAL_E
50 -64.842624462853280
150 -66.596463098602229
200 -66.596086697673911
250 -66.593017494126642
300 -66.591148878993607
350 -66.590341443197801
400 -66.590268413708017
450 -66.590267280865888
500 -66.590221265788742


Input File:
&GLOBAL
  PROJECT maleimide-CUT50
  RUN_TYPE ENERGY
  PRINT_LEVEL MEDIUM
&END GLOBAL
&FORCE_EVAL
  METHOD Quickstep
  &SUBSYS
    &KIND C
      ELEMENT   C
      BASIS_SET DZVP-GTH-q4
      POTENTIAL GTH-BP-q4
    &END KIND
    &KIND N
      ELEMENT   N
      BASIS_SET DZVP-GTH-q5
      POTENTIAL GTH-BP-q5
    &END KIND
    &KIND O
      ELEMENT   O
      BASIS_SET DZVP-GTH-q6
      POTENTIAL GTH-BP-q6
    &END KIND
    &KIND H
      ELEMENT   H
      BASIS_SET DZVP-GTH-q1
      POTENTIAL GTH-BP-q1
    &END KIND
    &CELL
      A     15.00000000    0.000000000    0.000000000
      B     0.000000000    15.00000000    0.000000000
      C     0.000000000    0.000000000    15.00000000
    &END CELL
    &COORD
      C   -1.34450  -0.60000   0.00000 
      C   -1.34450   0.75700   0.00000 
      C   -0.02500  -0.91990   0.00000 
      C   -0.02500   1.07680   0.00000 
      N    0.91900   0.07840   0.00000 
      O    0.31260  -2.08220   0.00000 
      O    0.31270   2.23910   0.00000 
      H   -2.23620  -1.24920   0.00000 
      H   -2.23620   1.40620   0.00000 
      H    1.92530   0.07840   0.00000 
    &END COORD
  &END SUBSYS
  &DFT
    BASIS_SET_FILE_NAME  GTH_BASIS_SETS
    POTENTIAL_FILE_NAME  GTH_POTENTIALS
    &QS
      EPS_DEFAULT 1.0E-9
    &END QS
    &MGRID
      NGRIDS 4
      CUTOFF 50
      REL_CUTOFF 60
    &END MGRID
    &XC
      &XC_FUNCTIONAL BP
      &END XC_FUNCTIONAL
    &END XC
    &SCF
      SCF_GUESS ATOMIC
      EPS_SCF 1.0E-7
      MAX_SCF 1
      &DIAGONALIZATION  ON
        ALGORITHM STANDARD
      &END DIAGONALIZATION
      &MIXING  T
        METHOD BROYDEN_MIXING
        ALPHA 0.4
        BETA 0.5
        NBROYDEN 8
      &END MIXING
    &END SCF
  &END DFT
  &PRINT
    &TOTAL_NUMBERS ON
    &END TOTAL_NUMBERS
  &END PRINT
&END FORCE_EVAL

Evgeniy 



-- 

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... at googlegroups.com.

To post to this group, send email to cp... at googlegroups.com.

Visit this group at https://groups.google.com/group/cp2k.

For more options, visit https://groups.google.com/d/optout.