ABINIT Discussion Forums

Hello
dear AbInit admin,,

Im a new user of the wonderful abinit code.
Lately I tried to carry out calculations of electron-phonon calculations thus, response function.
I followed exactly the respective tutorials and I was able successfully calculate for two cubic cells/ FCC and BCC.
Then when I try with other system, ej. Trigonal space group (164). bravv lattice = 4 , mind: Hexagonal"
I Performed in one run (jdtset): GS + [WF_Gamma + RF_Gamma] + [(WF_k+q) + (RF_K + q)] + ... ...
for all atoms and I have tried explicitly to all directions with rfdir 111 or specifying the perturbation (ej. 100 010 001).
mrggddb + mrggkk wrks normally without problem, ok
When "anaddb" is called, (to resolve the dynamic array) stop suddenly!
error:
#################################################################################
symdm9 : the bloks found in the DDB are characterized
by the following wavevectors :
0.000000000000D+00 0.000000000000D+00 0.000000000000D+00 0.100000000000D+01
0.000000000000D+00 0.000000000000D+00 0.500000000000D+00 0.100000000000D+01
0.500000000000D+00 0.000000000000D+00 0.000000000000D+00 0.100000000000D+01
symdm9 : ERROR -
Informations are missing in the DDB.
The dynamical matrix number 6 cannot be built,
since no blok with wavevector 5.000000E-01 0.000000E+00 5.000000E-01
has been found.
Action : add the required blok in the DDB, or modify your input file.
leave_new : decision taken to exit ...
###############################################################################

I tried for abinit versions. 5.6.3,/ 5.6.4,/ and 5.7.4./

Do not know what is happening.
a priori I think the problem is related to one symmetry operation that does not recognize anaddb rutine out of a cubic cell.
would greatly appreciate any clues or comments, if I am wrong in something please let me know.

cheers,
José A. FLORES L.
Université Lyon 1
FRANCE

##part of my In init file and four file to call anaddb
########################## JAFL june elec_phon app-574-004
ndtset 13
jdtset 1 2 3 4 5 6 7 8 9 10 11 12 13
#14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29
###################################### DATASET 1 : make ground state wavefunctions and density
iscf 3
kptopt1 3
tolwfr1 1.0E-14
prtden1 1
nline1 8 #precondition def=4,now expedite the convergence of higher-lying bands ok
#############################################################################################
# DATASET 2: construct the wavefunctions at k+q (same as at k,
# but kept for similarity to the other qpoints).
iscf2 -3 #Non-SC
tolwfr2 1.0E-14 #
getden2 1 #get density from exactly 1-last set
kptopt2 3
nqpt2 1 #rely on other but this is preperaning a non symmetry RF
qpt2 0.0 0.0 0.0 #default value

# DATASET 3,4,5: perturbation at q=Gamma along reduced direction 1,2,3 no! along
rfatpol3 1 2 rfdir3 1 0 0 rfphon3 1 tolvrs3 1.0E-5 getwfk3 1 getwfq3 2 kptopt3 3 nqpt3 1 qpt3 0.0 0.0 0.0
rfatpol4 1 2 rfdir4 0 1 0 rfphon4 1 tolvrs4 1.0E-5 getwfk4 1 getwfq4 2 kptopt4 3 nqpt4 1 qpt4 0.0 0.0 0.0
rfatpol5 1 2 rfdir5 0 0 1 rfphon5 1 tolvrs5 1.0E-5 getwfk5 1 getwfq5 2 kptopt5 3 nqpt5 1 qpt5 0.0 0.0 0.0
#####################################################################################################################
# DATASET 4: construct the wavefunctions at k+q
iscf6 -3 #Non-SC
tolwfr6 1.0E-14 #pqremeter convergence eaual to used for GAMMA
getden6 1 #get density from exactly 1-last set
kptopt6 3 #NON symmetry build
nqpt6 1 #rely on other but this is preperaning a non symmetry RF
qpt6 0.0 0.0 1/2 #q-vector A
rfatpol7 1 2 rfdir7 1 0 0 rfphon7 1 tolvrs7 1.0E-5 getwfk7 1 getwfq7 6 kptopt7 3 nqpt7 1 qpt7 0.0 0.0 0.5
rfatpol8 1 2 rfdir8 0 1 0 rfphon8 1 tolvrs8 1.0E-5 getwfk8 1 getwfq8 6 kptopt8 3 nqpt8 1 qpt8 0.0 0.0 0.5
rfatpol9 1 2 rfdir9 0 0 1 rfphon9 1 tolvrs9 1.0E-5 getwfk9 1 getwfq9 6 kptopt9 3 nqpt9 1 qpt9 0.0 0.0 0.5
#####################################################################################################################
# qpoint M
# DATASET 10: k+q+q
iscf10 -3
tolwfr10 1.0E-14
getden10 1
kptopt10 3
nqpt10 1
qpt10 1/2 0.0 0.0
rfatpol11 1 2 rfdir11 1 0 0 rfphon11 1 tolvrs11 1.0E-5 getwfk11 1 getwfq11 10 kptopt11 3 nqpt11 1 qpt11 0.5 0.0 0.0
rfatpol12 1 2 rfdir12 0 1 0 rfphon12 1 tolvrs12 1.0E-5 getwfk12 1 getwfq12 10 kptopt12 3 nqpt12 1 qpt12 0.5 0.0 0.0
rfatpol13 1 2 rfdir13 0 0 1 rfphon13 1 tolvrs13 1.0E-5 getwfk13 1 getwfq13 10 kptopt13 3 nqpt13 1 qpt13 0.5 0.0 0.0

# STRUCTURE EXPERIMENTAL-compltment relaxé-
###################important check and converge
ngkpt 4 4 4
# le kinetic enrgie permet avoir resultats pour manger vite
ecut 250.0 ev #HA
####################
acell 4.0805 4.0805 5.352088 angstrom
rprim 1.0 0.00000000 0.0
-0.5 sqrt(0.75) 0.0
0.0 0.00000000 1.0
#########################################################################################################################
xred 0.0 0.0 0.0
1/3 2/3 0.5614991969919704
2/3 1/3 0.4385008030080296
#Definition of the atom types
natom 3
ntypat 2
typat 1 2 2
znucl 56 14
nband 18
occopt 7
tsmear 0.001
nstep 100
#Definition of the k-point grig here to make the difference: according to reduced the shift to voilâ
nshiftk 1
shiftk 0.0 0.0 0.0
#Exchange-correlation functional
ixc 11 # PBE-GGA
################################################################################################################FIN

anaddb file/

elphflag 1.0
# Phonon dispersion will be calculated/ Fourier interpolation
nqpath 3
qpath
0.00000000E+00 0.00000000E+00 0.00000000E+00 #G
0.00000000E+00 0.00000000E+00 5.00000000E-01 #A
5.00000000E-01 0.00000000E+00 0.00000000E+00 #M
mustar 0.2
# Minimalistic qpoint grid
ngqpt 2 2 2
asr 2
dipdip 1
brav 4
nqshft 1
q1shft 0.0 0.0 0.0
ifcflag 1
ifcana 1
natifc 3
atifc 1 2 3
eivec 1
#Wavevector list number 1 (Reduced coordinates and normalization factor)
nph1l 1
qph1l
0.00000000E+00 0.00000000E+00 0.00000000E+00 1
#This line added when defaults were changed (v5.3) to keep the previous, old behaviour
symdynmat 0

you are missing perturbations wrt certain q-points (anaddb even tells you which one). As you have less symmetry than in FCC/BCC, the q you give - 0 0 0 0.5 0 0 0 0 0.5 - are not enough.

You need to add the corresponding datasets for (at least) 0.5 0 0.5. To get a list of the irreducible q for your system you should run a ground state calculation with kptopt 1 and the ngkpt set to the q-point grid that you want, then copy over the k-points as qpt for the response function calculation.

If you check the latest (v6.0) tutorial input file, there is a much simpler way to make abinit do all the needed perturbations (prtgkk 1 and prepgkk 1 with rfatpol 1 3 and rfdir 1 1 1). With your present file you will also be missing perturbations wrt the last atom, which are needed for the el-ph calculation.

matthieu

Dear proffesor Matthieu,

Thank you very much for your response.
I follow the changes as you told me and I have no problem
and also included more high-symmetry points and ngkpt 1 in GS,
indeed following instructions given in the tutorial 6v

again thank you very much

ciao José

Why do you use only ngkpt 1?

Note that for the moment some parts of the electron phonon code are still isotropic, and will presume the system is cubic. You may not get all that you want out of it, but we're working on it.

matthieu

Dear admin.
I used only ngkpt 1 in my GS file, then perform RF+q with kptoptx 3, rfidir 111 (and explicit perturbation), rfatpol, prepgkk, etc..

I know some parts of the code for "telphon" still in progress.in that way I'm doing convergence test necessary to see the approach. cmparing as possible with experimental parameters. Im keeping in mind/ different K-points (and construction), energy, high sym. points, (excluding including) regarding the evolution of dispersion curves (BDS), and also including the two possibles integration methods for a-squareF; Tetrahedron and Gaussian. telphint 0 . 1
Maybe I 've forgotten something, or is out of my understand.
so, in that case, i will be grateful if you could tell me. (advice)

José Flores

Dear admin of AbInit,

I have one more question, again, my GS calculation is only using irreducible zone, then; the response function using the entire B. zone,
all this with the specific Qwavectors. everything looks good, but, when I try to calculate for exemple 4x4x4 qmesh, calculating the perturbation and the GKK of all this in the irreducible zone 13 points for my system including gamma and not shifted/.

call anaddb. :

" elphon : calling mkqptequiv to set up the FS qpoint set
mkqptequiv : making rankFSkpt and invrankFSkpt
mkqptequiv : Error : looks like no kpoint equiv to k+q !!! "

Im trying for exemple Q2x2X2 ok: ; but 4X4X4 not, Im checked the perturbations and are complete in the gkk and DDB is ok also for the points in the irreducible for 4X4X4....
I really appreciate if you can give me some hint or clue of what im doing wrong,
initially I was thinking that some points in the Gst calculation for each perturbation (K+q), have different mesh and not correspond to other calculation of perturbation.

thanks in advance.

ps. I prefer not open an other subject, because is very punctual and probably I did somthing wrong.

JAFL

which k-mesh are you using. The algorithm is quite simple, so it's probably right: your q-grid is not a subgrid of the k-grid. See the elphon tutorial as well.

matthieu