ABINIT Discussion Forums

Dear ABINIT Users,

I have used the conducti code for calculations at gamma. Is it possible to calculate the conductivity at say q=(0.25 0.25 0.25) using the Conducti code? If yes, what changes are needed to the v3/t78.in other than changing the qpt and ensuring that the qpt is included in the k-grid?

Thanks,
Vahid

Let me make the above question more specific. When I use the following input file for calculating the ddk response function at q=(0.25 0.25 0.25) in the 1,0,0 direction (using ABINIT-6.12.2),

ndtset 3
# - Options for output
enunit 2
prtvol 1
# - Parameters governing the convergence
ecut 6.
nband 20
occopt 3
tsmear 0.0002
# - Options for the choice of calculations
ixc 1
nstep 30
# SC run
iscf1 3
toldfe1 1.d-10
nqpt1 0
prtden1 1
# NSC run
iscf2 -2
tolwfr2 1.0d-22
nqpt2 0
getden2 1
getwfk2 1
# ddk response function
iscf3 -3
tolwfr3 1.0d-22
nqpt3 1
qpt3 0.25 0.25 0.25
rfdir3 1 0 0
rfelfd3 2
getwfk3 2
# - Definition of special k-points
kptopt 1
ngkpt 4 4 4
nshiftk 1
shiftk 0 0 0
istwfk 36*1
# - Definition of the unit cell
acell 3*7.6557
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
amu 26.98
# - Symmetries of the lattice
nsym 1
# This line added when defaults were changed (v5.3) to keep the previous, old behaviour
iscf 5
# add to conserve old < 6.7.2 behavior for calculating forces at each SCF step
optforces 1
--------------------END OF INPUT FILE----------------------------

I get the following error pointing to the last i2:

At line 1070 of file respfn.F90
Fortran runtime error: Expected INTEGER for item 17 in formatted transfer, got CHARACTER
(a,a,a,3es16.6,a,a,a,i2,a,i2,a,i2,a,i2,a,a,a)

The k-grid is chosen such that the above q point is included. The ecut and nband are low so as to speed up the calculations. The above input runs fine if in the third dataset, q=0,0,0. I am wondering if it is possible to run the above analysis for non-zero wavevectors. The log file is attached.

Any suggestion or hint will be appreciated.
Vahid