ABINIT Discussion Forums

Dear developers,

Thanks for your attention firstly. When i want to calculate the ground state of a system including the exact exchange term, there is always an error as: segmentation fault occurred. I tried a lot of time to improve the input file, but there is no effect. Would you please help me to check if there are some errors in my input file? There is no error when i did the calculation without exact exchange parameters.

Thanks a lot

The input file is shown as:
# 2-Options for output:
# ---------------------
enunit 2
prtvol 0
# prtdos 2
prtden 0
prtwf 1
prtnabla 1

##### Exact Exchange

useexexch 1
lexexch 2 -1
exchmix 0.25
# 3-Parameters governing the convergence:
# ---------------------------------------
ecut 100.
pawecutdg 250.
pawmixdg 1
# pawovlp =-1
nband 50
nstep 100
nline 4
toldfe 1.d-9

# 4-Options for the choice of calculation:
# ----------------------------------------
ixc 11

nsppol 1
iscf 17
occopt 3
tsmear 0.0036749

# occopt 0
# occ 2 1
# occopt 7
# tsmear 0.04
ionmov 0
optcell 0

# 5-Definition of the unit cell:
# ------------------------------
acell 3*7.38796
# rprim 0.0 0.5 0.5
# 0.5 0.0 0.5
# 0.5 0.5 0.0
# natom 1 ntypat 1
# typat 1 znucl 13.
# xred 0.0 0.0 0.0

rprim 1.0 0.0 0.0
0.0 1.0 0.0
0.0 0.0 1.0
natom 4
ntypat 1 1 1 1
typat 1 1 1 1 1 1 1 1 1 1
znucl 13
xred 0.0 0.0 0.0
0.5 0.5 0.0
0.5 0.0 0.5
0.0 0.5 0.5

amu 26.98

# 6-Definition of special K-points:
# ---------------------------------
kptrlatt 4 0 0 0 4 0 0 0 4
nshiftk 1
#shiftk 0.5 0.5 0.5
chksymbreak 0
# ngkpt 3 3 3
# kptopt 1
# kptnrm 1
# nshiftk 4
shiftk 0.5 0.5 0.5
0.5 0.0 0.0
0.0 0.5 0.0
0.0 0.0 0.5

# 7-Symmetries of the lattice:
#
nsym 1
chkprim 0

Your ecut and pawecutdg are incredibly large, this will require a huge amount of memory, it may be that this is crashing your system. Why is your ecut so huge? What PAW set are you using?

Other issues: You evidently have one type of atom (sodium, I guess) so ntypat 1 is what you should have. There are apparently 4 atoms in the cell so typat should be 4*1 or 1 1 1 1. Your input looks like the structure is primitive cubic, is that right? If so I don't think your kpt grid is coherent with this (looks like an FCC type lattice).

Dear Prof. Zwanziger,

Thanks for your helpful reply. In fact, this is the test file which i changed from the example v5/t09.in for the Al system. It is primitive cubic for FCC structure. When i used the small cutoff for plane waves, the errors were still there. The large cut is because that i used the 11 valences pseudopotential.

I will check the kpt set.

Thanks a lot.
Best wishes,
Jiayu Dai
Department of physics,
National University of Defense Technology
China

Dear developers,

After my test, i found the PBE0 calculations are dependent on the choice of the pseudopotential. For example, for the example case v5/t18.in, if we use the supplied pseudopotential, i.e., 28ni.pbe.paw, the calculations can be done. However, if we change the pseudopotential to the other one, for example, Al.pbe.gga.paw, the errors will be shown.

So, my question is that: is there any requirement for the construction of pseudopentials? If so, what's the requirement?

Best wishes.

Jiayu Dai