K-points problem in cRPA

Dear colleagues,

I was trying to get out U and J using cRPA. Can someone please take a look at the problem I am facing (described below).


— !WARNING
src_file: m_sigma_driver.F90
src_line: 3498
message: |
Problem while computing the fundamental and direct gap (likely metal). Will replace gw_qprange=0 with gw_qprange=1

— !ERROR
src_file: m_sigma_driver.F90
src_line: 3636
mpi_rank: 0
message: |
k-point [-8.3333E-02, -8.3333E-02, -8.3333E-02] not in the list of k points treated in the preparatory SCF run.
Change kptgw, or shiftk of previous run.

==== K-mesh for the wavefunctions ====
Number of points in the irreducible wedge : 18
Reduced coordinates and weights :

 1)     8.33333333E-02  8.33333333E-02  8.33333333E-02       0.03704
 2)     2.50000000E-01  8.33333333E-02  8.33333333E-02       0.07407
 3)     4.16666667E-01  8.33333333E-02  8.33333333E-02       0.07407
 4)     2.50000000E-01  2.50000000E-01  8.33333333E-02       0.03704
 5)     4.16666667E-01  2.50000000E-01  8.33333333E-02       0.07407
 6)     4.16666667E-01  4.16666667E-01  8.33333333E-02       0.03704
 7)     8.33333333E-02  8.33333333E-02  2.50000000E-01       0.03704
 8)     2.50000000E-01  8.33333333E-02  2.50000000E-01       0.07407
 9)     4.16666667E-01  8.33333333E-02  2.50000000E-01       0.07407
10)     2.50000000E-01  2.50000000E-01  2.50000000E-01       0.03704
11)     4.16666667E-01  2.50000000E-01  2.50000000E-01       0.07407
12)     4.16666667E-01  4.16666667E-01  2.50000000E-01       0.03704
13)     8.33333333E-02  8.33333333E-02  4.16666667E-01       0.03704
14)     2.50000000E-01  8.33333333E-02  4.16666667E-01       0.07407
15)     4.16666667E-01  8.33333333E-02  4.16666667E-01       0.07407
16)     2.50000000E-01  2.50000000E-01  4.16666667E-01       0.03704
17)     4.16666667E-01  2.50000000E-01  4.16666667E-01       0.07407
18)     4.16666667E-01  4.16666667E-01  4.16666667E-01       0.03704

Together with 16 symmetry operations and time-reversal symmetry
yields 216 points in the full Brillouin Zone.

Does this number (216) need to be a multiple of nkpt==18 points?


Here is the part of the input file

ndtset 4
jdtset 1 2 3 4
paral_kgb 0
prtvol 10
pawprtvol 3
############ Parameters common to all DATASETS#############

nstep 500
nline 5 #Number of LINE minimisations
nnsclo 5 #Number of Non-Self Consistent LOops
tolvrs 1.0d-12

occopt 3 #OCCupation OPTion
tsmear 300 K #Temperature of SMEARing

ngkpt 6 6 6
nshiftk 1
shiftk 0.5 0.5 0.5
istwfk *1

nband 35
ecut 35
pawecutdg 42

pawoptosc 2
usepawu 1
lpawu -1 2 -1
jpawu 0.0 0.0 0.0 eV
upawu 0.0 0.0 0.0 eV
dmatpuopt 1
ucrpa 1
tolvrs1 1.0d-15

############ SECOND DATASET######
getden2 -1
iscf2 -2
nbandkss2 -1
kssform2 3
nbdbuf2 4
tolwfr2 1.0d-18
usedmft2 1
dmftbandi2 17
dmftbandf2 21

############ THIRD DATASET###############
optdriver3 3
gwcalctyp3 2
getwfk3 -1
ecuteps3 5.0

nfreqre3 1
freqremax3 10 eV
freqremin3 0 eV
nfreqim3 0

ucrpa_bands3 17 21
gwpara3 1

############ FOURTH DATASE############
gwmem4 01
optdriver4 4
getwfk4 2
getscr4 3
ecutsigx4 30.0

nfreqsp4 1
freqspmax4 10 eV
freqspmin4 0 eV
nkptgw4 0
mqgrid4 1001
mqgriddg4 1001

gwpara4 2

Dear all,

Just for your info, I missed the two variable in the imput file, symchi and symsigma. It works now.

BTW, it seems that DFT+U can’t be performed with norm-conserving pseudopotential (NCPP)? What kind of calculations are available for NCPP using Abinit?

Best!
R