ABINIT Discussion Forums

As the parallel version 7.2.1 ends up with segmentation fault errors, I am trying to run the serial version to do the linear response calculations for an insulator. I am using the template trf2_1.in. The program was Ok for a 4x4x4 kmesh, diemac =2 and converged within 3000 steps. But when I increase ngkpt to 8x8x8, program runs smoothly, but dataset 2 does not converge even after nstep=5000. I used diemac = 1. Can someone suggest a solution please..?
Uma

What do you mean? That the input file trf2_1.in itself fails? That is pretty unlikely. Have you altered it for your system? If so you will have to post it, otherwise no one can give you advice--

Dear Professor,
I have now attached the input file that I used for the calculation which ran to completion for all datasets, but dataset2 Energy has not converged ( tolwfr2 1.0d-22). Can you suggest a solution to achieve convergence so that the phonon frequencies based on this calculation could be reliable?
Uma

# computation of the phonon spectrum

ndtset 37
#Set 1 : ground state self-consistency

getwfk1 0 # Cancel default
kptopt1 1 # Automatic generation of k points, taking
# into account the symmetry
nqpt1 0 # Cancel default
tolvrs1 1.0d-18 # SCF stopping criterion (modify default)
rfphon1 0 # Cancel default
getddk1 0
optdriver1 0
rfelfd1 0
tolwfr1 0

#Q vectors for all datasets

#Complete set of symmetry-inequivalent qpt chosen to be commensurate
# with kpt mesh so that only one set of GS wave functions is needed.
#Generated automatically by running GS calculation with kptopt=1,
# nshift=0, shiftk=0 0 0 (to include gamma) and taking output kpt set
# file as qpt set. Set nstep=1 so only one iteration runs.

nqpt 1 # One qpt for each dataset (only 0 or 1 allowed)
# This is the default for all datasets and must
# be explicitly turned off for dataset 1.

qpt2 0.00000000E+00 0.00000000E+00 0.00000000E+00
qpt3 0.00000000E+00 0.00000000E+00 0.00000000E+00
qpt4 1.25000000E-01 0.00000000E+00 0.00000000E+00
qpt5 2.50000000E-01 0.00000000E+00 0.00000000E+00
qpt6 3.75000000E-01 0.00000000E+00 0.00000000E+00
qpt7 5.00000000E-01 0.00000000E+00 0.00000000E+00
qpt8 1.25000000E-01 1.25000000E-01 0.00000000E+00
qpt9 2.50000000E-01 1.25000000E-01 0.00000000E+00
qpt10 3.75000000E-01 1.25000000E-01 0.00000000E+00
qpt11 5.00000000E-01 1.25000000E-01 0.00000000E+00
qpt12 -3.75000000E-01 1.25000000E-01 0.00000000E+00
qpt13 2.50000000E-01 2.50000000E-01 0.00000000E+00
qpt14 3.75000000E-01 2.50000000E-01 0.00000000E+00
qpt15 1.25000000E-01 1.25000000E-01 1.25000000E-01
qpt16 2.50000000E-01 1.25000000E-01 1.25000000E-01
qpt17 3.75000000E-01 1.25000000E-01 1.25000000E-01
qpt18 -1.25000000E-01 1.25000000E-01 1.25000000E-01
qpt19 2.50000000E-01 2.50000000E-01 1.25000000E-01
qpt20 -2.50000000E-01 2.50000000E-01 1.25000000E-01
qpt21 -1.25000000E-01 2.50000000E-01 1.25000000E-01
qpt22 -3.75000000E-01 3.75000000E-01 1.25000000E-01
qpt23 -2.50000000E-01 3.75000000E-01 1.25000000E-01
qpt24 -1.25000000E-01 3.75000000E-01 1.25000000E-01
qpt25 5.00000000E-01 5.00000000E-01 1.25000000E-01
qpt26 -3.75000000E-01 5.00000000E-01 1.25000000E-01
qpt27 -2.50000000E-01 5.00000000E-01 1.25000000E-01
qpt28 -3.75000000E-01 -3.75000000E-01 1.25000000E-01
qpt29 2.50000000E-01 2.50000000E-01 2.50000000E-01
qpt30 -2.50000000E-01 2.50000000E-01 2.50000000E-01
qpt31 -3.75000000E-01 3.75000000E-01 2.50000000E-01
qpt32 -2.50000000E-01 3.75000000E-01 2.50000000E-01
qpt33 5.00000000E-01 5.00000000E-01 2.50000000E-01
qpt34 -3.75000000E-01 5.00000000E-01 2.50000000E-01
qpt35 -3.75000000E-01 3.75000000E-01 3.75000000E-01
qpt36 5.00000000E-01 5.00000000E-01 3.75000000E-01
qpt37 5.00000000E-01 5.00000000E-01 5.00000000E-01

# qpt2 0.00000000E+00 0.00000000E+00 0.00000000E+00
# qpt3 0.00000000E+00 0.00000000E+00 0.00000000E+00
# qpt4 2.50000000E-01 0.00000000E+00 0.00000000E+00
# qpt5 5.00000000E-01 0.00000000E+00 0.00000000E+00
# qpt6 2.50000000E-01 2.50000000E-01 0.00000000E+00
# qpt7 2.50000000E-01 2.50000000E-01 2.50000000E-01
# qpt8 -2.50000000E-01 2.50000000E-01 2.50000000E-01
# qpt9 5.00000000E-01 5.00000000E-01 2.50000000E-01
# qpt10 5.00000000E-01 5.00000000E-01 5.00000000E-01


#Set 2 : Response function calculation of d/dk wave function

iscf2 -3 # Need this non-self-consistent option for d/dk
kptopt2 2 # Modify default to use time-reversal symmetry
rfphon2 0 # Cancel default
rfelfd2 2 # Calculate d/dk wave function only
tolwfr2 1.0d-22 # Use wave function residual criterion instead
getwfk2 1
getddk2 0
optdriver 1
tolvrs2 0
#Set 3 : Response function calculation of Q=0 phonons and electric field pert.

getddk3 2 # d/dk wave functions from last dataset
kptopt3 2 # Modify default to use time-reversal symmetry
rfelfd3 3 # Electric-field perturbation response only
getwfk3 1
tolwfr 0
#Sets 4-37 : Finite-wave-vector phonon calculations (defaults for all datasets)
getddk 0
getwfk 1 # Use GS wave functions from dataset1
kptopt4 3 # Need full k-point set for finite-Q response
kptopt5 3
kptopt6 3
kptopt7 3
kptopt8 3
kptopt9 3
kptopt10 3
kptopt 3
rfphon 1 # Do phonon response
rfelfd 0
rfatpol 1 4 # Treat displacements of all atoms
rfdir 1 1 1 # Do all directions (symmetry will be used)
tolvrs 1.0d-8 # This default is active for sets 3-10

#######################################################################
#Common input variables

#Definition of the unit cell
acell 3.8037 3.8037 3.8037 angstrom
rprim -0.5000000000000000 0.5000000000000000 0.5000000000000000
0.5000000000000000 -0.5000000000000000 0.5000000000000000
0.5000000000000000 0.5000000000000000 -0.5000000000000000

# rprim 0.0 0.5 0.5 # In lessons 1 and 2, these primitive vectors
# 0.5 0.0 0.5 # (to be scaled by acell) were 1 0 0 0 1 0 0 0 1
# 0.5 0.5 0.0 # that is, the default.

#Definition of the atom types
ntypat 1 # There is one type of atom
znucl 7.0 # The keyword "znucl" refers to the atomic number of the
# possible type(s) of atom. The pseudopotential(s)
# mentioned in the "files" file must correspond
# to the type(s) of atom.

#Definition of the atoms
natom 4 # There are 4 atoms
typat 1 1 1 1 # 4 atoms of the same type N.
xred 0.17000000000000017 0.1700000000000017 0.1700000000000017
0.5000000000000000 0.0000000000000000 0.3299999999999983
0.0000000000000000 0.3299999999999983 0.5000000000000000
0.3299999999999983 0.5000000000000000 0.0000000000000000

# xred 0.0 0.0 0.0
# 0.25 0.25 0.25

#Gives the number of band, explicitely (do not take the default)
nband 24

#Exchange-correlation functional

ixc 11 # GGA, Perdew-Burke-Ernzerhof GGA functional

#Definition of the planewave basis set

ecut 500 eV # Maximal kinetic energy cut-off, in eV
enunit 2
#Definition of the k-point grid
ngkpt 8 8 8
# prtkpt 0
nshiftk 1 # Use one copy of grid only (default)
shiftk 0.0 0.0 0.0 # This gives the usual fcc Monkhorst-Pack grid
# 0.5 0.5 0.5
# 0.5 0.5 0.5
# 0.5 0.5 0.5

#Definition of the SCF procedure
iscf 5 # Self-consistent calculation, using algorithm 5
nstep 5000 # Maximal number of SCF cycles
diemac 1.0d0 # Although this is not mandatory, it is worth to
# precondition the SCF cycle. The model dielectric
# function used as the standard preconditioner
# is described in the "dielng" input variable section.
# add to conserve old < 6.7.2 behavior for calculating forces at each SCF step
optforces 1
# rfasr 1
tsmear .04
#%%<BEGIN TEST_INFO>
#%% [setup]
#%% executable = abinit
#%% test_chain = trf2_1.in, trf2_3.in, trf2_4.in, trf2_5.in, trf2_6.in, trf2_7.in
#%% [files]
#%% files_to_test =
#%% trf2_1.out, tolnlines= 13, tolabs= 4.294e-05, tolrel= 5.000e-04, fld_options=-medium
#%% psp_files = 13al.981214.fhi, 33as.pspnc
#%% [paral_info]
#%% max_nprocs = 1
#%% [extra_info]
#%% author =
#%% keywords = NC, DFPT
#%% description =
#%%<END TEST_INFO>

From your input file it looks like you're trying to model a high pressure phase of nitrogen, that you are treating as metallic (you have tsmear defined, you have nband = 24 though I suspect there are only 20 valenece electrons in the system--5 for each nitrogen, 4 nitrogen in the cell). So, if my guesses above are correct, you should not be doing the ddk or rfelfd calculations, they have no meaning in a metallic system. You should only do the phonon calculations.

Thank you for your response Professor. But I am working on Nitrogen when it is an insulator. But at some point during my trials, the code suggested that I increase the number of bands. That is why I used nband=24 instead of my original value of 20. I shall try again with nband=20 and not specifying tsmear. Will that be a correct direction of managing parameters..?
Uma

Yes, but I don't understand why it is 20 and not 10--what is your valence space for the nitrogen pseudopotentials? I'm guessing 2s2-2p3, so 5 electrons per nitrogen. You have four nitrogens in the cell, so 20 electrons, so you should have nband = 10. Or am I missing something?

Dear Professor,
With nband=10 this is the error message I received:

scfcge : ERROR -
Potential-based CG line minimization not converged after 13 restarts.
Action : read the eventual warnings about lack of convergence.
Some might be relevant. Otherwise, raise nband.

Nonetheless, the ddk calculation is only meant to work with no unoccupied bands. I would suggest two things: first have you computed the band structure? It would be worthwhile to see what the band gap appears to be based on the xc and pseudopotentials you are using. Perhaps the band gap is too small and that's where the problems come from. Secondly, do you really need the electric field calculation? If you only need the phonons (and can live without the lo-to spliting) you can simply not do the ddk and rfelfd calcs and only do the phonons.

Dear Professor,
Yes. I need only the phonon frequencies of Nitrogen as an insulator.I commented datasets 2 and 3. So I did not include d/dk calculations and the electric field perturbation. Also nband=10. diemac =1. This is the error message I get.

Subroutine Unknown:0:ERROR

scfcge : ERROR -
Potential-based CG line minimization not converged after 13 restarts.
Action : read the eventual warnings about lack of convergence.
Some might be relevant. Otherwise, raise nband.

leave_new : decision taken to exit ...

Any suggestion please..? I will plot the band structure if it will solve this problem. I have the results of this calculation using VASP. So I know it to be an insulator.

Uma