[Wien] How to select the appropriate nband in hybrid functionals calculations

Fri May 14 17:31:30 CEST 2021

Dear Wien2k developers，

When I tried to do hybrid functional calculation (YS-PBE0) on h-BN monolayer which was set a 4*4*1 supercell including 30 B atoms and 34 N atoms， all calculated attempts went wrong in the first loop. 

The following content comes from case.dayfile：
>   lcore       (04:48:28) 0.546u 0.788s 0:01.66 79.5%  0+0k 0+0io 2pf+0w
>   hf      -redklist -p -c     (04:48:30) running HF in parallel mode
**  HF crashed!
0.159u 0.253s 0:06.88 5.8%      0+0k 0+40io 0pf+0w
error: command   /public/software/wien2k/WIEN2k_14/hfcpara -c hf.def   failed

>   stop error

I searched past mailing lists and found that there are few questions about hybrid functional errors, and several questions and answers about hybrid functional errors can't solve my problems. I put my calculation steps below， and any comments would be highly appreciated. Thanks in advance!

1. Initialization
[yurc at node71 bigc-tahybridtest1]$ init_hf_lapw
       Insulator, EF-inconsistency corrected
:GAP  :    0.1301 Ry =     1.769 eV   (provided you have a proper k-mesh)
         Bandranges (emin - emax) and occupancy:
:BAN00120: 120   -0.462509   -0.403568  2.00000000
:BAN00121: 121   -0.403194   -0.379922  2.00000000
:BAN00122: 122   -0.403039   -0.378604  2.00000000
:BAN00123: 123   -0.393649   -0.368830  2.00000000
:BAN00124: 124   -0.391842   -0.367239  2.00000000
:BAN00125: 125   -0.388449   -0.357020  2.00000000
:BAN00126: 126   -0.388279   -0.353847  2.00000000
:BAN00127: 127   -0.388042   -0.333360  2.00000000
:BAN00128: 128   -0.387867   -0.326951  2.00000000
:BAN00129: 129   -0.129916   -0.120609  2.00000000
:BAN00130: 130   -0.128004   -0.113769  2.00000000
:BAN00131: 131    0.016334    0.034382  0.00000000
:BAN00132: 132    0.016486    0.034448  0.00000000
:BAN00133: 133    0.038904    0.046957  0.00000000
:BAN00134: 134    0.039002    0.047860  0.00000000
:BAN00135: 135    0.041063    0.051186  0.00000000
        Energy to separate low and high energystates:   -1.10359

:NOE  : NUMBER OF ELECTRONS          = 260.000

You must set NBAND to at least  NB_occ + 1  and you have 260.00 electrons
edit bigc-tahybridtest1.inhf ...
PS: For very accurate calc. and large NBAND you may have to increase EMAX in bigc-tahybridtest1.in1 by hand

Prepare k-mesh for HF. Use identical mesh and shift for IBZ and FBZ
Do you want to use a REDUCED k-mesh for HF (saving cpu-time) (Y/n) ? 
Y
 This script runs   x kgen   twice and generates equivalent meshes in the
 IBZ and FBZ.

KGEN ENDS
KGEN ENDS
How many k-points in full BZ?
If you type 0 you can give 3 numbers for nx,ny,nz
10
           2  symmetry operations without inversion
 inversion added (non-spinpolarized non-so calculation)
  NUMBER OF K-POINTS IN WHOLE CELL: (0 allows to specify 3 divisions of G)
 length of reciprocal lattice vectors:   0.383   0.383   0.083   3.586   3.586   0.778
  Shift of k-mesh allowed. Do you want to shift: (0=no, 1=shift)
           5  k-points generated, ndiv=           3           3           1
KGEN ENDS
0.000u 0.012s 0:00.06 16.6%	0+0k 0+0io 0pf+0w
           1  symmetry operations without inversion
  NUMBER OF K-POINTS IN WHOLE CELL: (0 allows to specify 3 divisions of G)
 length of reciprocal lattice vectors:   0.383   0.383   0.083   3.586   3.586   0.778
  Shift of k-mesh allowed. Do you want to shift: (0=no, 1=shift)
           9  k-points generated, ndiv=           3           3           1
KGEN ENDS
0.000u 0.003s 0:00.04 0.0%	0+0k 0+0io 0pf+0w
Give nx,ny,nz for the reduced mesh
nx=?
1
ny=?
1
nz=?
1
           1  symmetry operations without inversion
  NUMBER OF K-POINTS IN WHOLE CELL: (0 allows to specify 3 divisions of G)
 length of reciprocal lattice vectors:   0.383   0.383   0.083   0.000   0.000   0.000
  Specify 3 mesh-divisions (n1,n2,n3):
  Shift of k-mesh allowed. Do you want to shift: (0=no, 1=shift)
           1  k-points generated, ndiv=           1           1           1
KGEN ENDS
0.000u 0.004s 0:00.00 0.0%	0+0k 0+0io 0pf+0w
 Now you can use      run_lapw -hf -redklist ...
bigc-tahybridtest1.in0_grr and bigc-tahybridtest1.inhf and hf-kmesh prepared
Now do the hybrid calculation:   run_lapw -hf -redklist ...  
[yurc at node71 bigc-tahybridtest1]$ 

The content of the file case.inhf is as follows:
case.inhf
0.25         alpha
T            screened (T) or unscreened (F)
0.165        lambda
136          nband
6            gmax
3            lmaxe
3            lmaxv
1d-3         tolu

2. SCF calculations
The command I used when submitting the calculation is "run_lapw -hf -redklist -ec 0.6 -p -i 999 > output.log".

3. error information
I tried many times to calculate and modify the parameters. Every time, the error appears in first cycle in Hf calculation. The following is the file with error message:
-rw-r--r-- 1 yurc users 61 May 13 04:48 hf.error
-rw-r--r-- 1 yurc users 12 May 13 04:48 hf_5.error
-rw-r--r-- 1 yurc users 12 May 13 04:48 hf_4.error
-rw-r--r-- 1 yurc users 12 May 13 04:48 hf_3.error
-rw-r--r-- 1 yurc users 12 May 13 04:48 hf_2.error
-rw-r--r-- 1 yurc users 12 May 13 04:48 hf_1.error

The content of hf.error is:
**  Error in Parallel HF
**  testerror: Error in Parallel HF

And the content of file from hf_1.error to hf_5.error is:
error in hf

Thank you again for reading my question. Any suggestions are welcome！

Yifan Ding

Institute of Physics, Chinese Academy of Science (CAS)
Address: No.8 Zhongguancun South 3rd Street, Haidian District, Beijing
100190 China
E-mail: yfding at iphy.ac.cn
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://zeus.theochem.tuwien.ac.at/pipermail/wien/attachments/20210514/d72dcd32/attachment-0001.htm>