DFPT

Moderators: mverstra, aromero

DFPT

Postby wlsong » Sat May 25, 2019 11:37 am

hello,thank you for your attension!my error as follows:
getshell : BUG -
The calculated weights do not solve the linear system for all k-points.
Action: contact ABINIT group (please attach the output of `abinit -b`)


--- !BUG
src_file: m_getshell.F90
src_line: 653
mpi_rank: 0
message: |
There is a problem with the finite difference expression of the ddk
...

Action: contact ABINIT group (please attach the output of `abinit -b`)
wlsong
 
Posts: 6
Joined: Mon Apr 01, 2019 4:17 am

Re: DFPT

Postby ebousquet » Tue Jun 04, 2019 11:08 am

Dear wlsong,
You should be more "talkative" if you want a response, what is your calculation exactly? Potentially, your input-output? With just the error message, it'll be hard to help you.
Best wishes,
Eric
ebousquet
 
Posts: 303
Joined: Tue Apr 19, 2011 11:13 am
Location: University of Liege, Belgium

Re: DFPT

Postby wlsong » Mon Jun 10, 2019 4:06 am

I am caculating a poscar containing C.N,O H.when i was caculating its response ,its err as my lastest poster.
# Definition of lattice parameters
#********************************
acell 4.6785467407E+00 4.6785465952E+00 5.4240539143E+01
rprim 9.9979825533E-01 1.3096122241E-02 -1.5229583441E-02
-4.8855756831E-01 8.7239877279E-01 1.5227727480E-02
-1.5804714401E-02 9.1226315354E-03 9.9983348043E-01





#Definition of atomic positions and types
#*************************************
natom 8
ntypat 4
znucl 5 7 1 9
typat 1 1 1 2 2 2 3 4
xred 6.6427145839E-01 3.3573183651E-01 3.8655144545E-01
-2.4285448185E-02 2.4241144023E-02 4.5731699263E-01
3.3682932037E-01 6.6321527484E-01 3.1539849813E-01
9.6619097835E-03 -9.6643284553E-03 4.0570763200E-01
3.3328032397E-01 6.6667186760E-01 4.7505895302E-01
6.9329998641E-01 3.0674082204E-01 3.3515841280E-01
2.9821439204E-01 7.0172586775E-01 5.0978669868E-01
3.5539423281E-01 6.4467060348E-01 2.6720778517E-01



ixc 3

#Parameters of the SCF cycles
#****************************
nstep 200


#Plane wave basis and k-point grid
#*********************************
ecut 20
ecutsm 0.5
dilatmx 1.05
ngkpt 13 13 1
nshiftk 1
shiftk 0.5 0.5 0.5




ndtset 5 jdtset 1 2 3 4 5

#DATASET1 : scf calculation: GS WF in the BZ
#********************************************
prtden1 1
kptopt1 1
toldfe1 1.0d-15

#DATASET2 : non scf calculation: GS WF in the whole BZ
#*****************************************************
getden2 1
kptopt2 2
iscf2 -2
getwfk2 1
tolwfr2 1.0d-20
nband2 20


#DATASET3 : ddk on the MP grid of k-points
#*****************************************
getwfk3 2
rfdir3 1 1 1
rfelfd3 2
tolwfr3 1.0d-20
nband3 20
kptopt3 2

#DATASET4 : ddE on the MP grid of k-points
#*****************************************
prtden4 1
getwfk4 2
getddk4 3
rfdir4 1 1 1
rfphon4 1
rfatpol4 1 8
tolvrs4 1.0d-15
rfelfd4 3
nband4 20
kptopt4 2
prepanl4 1

#DATASET5 : 3DTE calculation
#***************************
getden5 1
get1den5 4
getwfk5 2
get1wf5 4
nband5 20
kptopt5 2
optdriver5 5
d3e_pert1_elfd5 1
d3e_pert1_phon5 1
d3e_pert1_atpol5 1 8
d3e_pert1_dir5 1 1 1
d3e_pert2_elfd5 1
d3e_pert2_dir5 1 1 1
d3e_pert3_elfd5 1
d3e_pert3_dir5 1 1 1

nbdbuf 0


## After modifying the following section, one might need to regenerate the pickle database with runtests.py -r
#%%<BEGIN TEST_INFO>
#%% [setup]
#%% executable = abinit
#%% test_chain = tnlo_2.in, tnlo_3.in, tnlo_4.in, tnlo_5.in
#%% [files]
#%% files_to_test =
#%% tnlo_2.out, tolnlines= 9, tolabs= 9.000e-07, tolrel= 6.000e-05
#%% psp_files = 13al.pspnc, 33as.pspnc
#%% [paral_info]
#%% max_nprocs = 2
#%% [extra_info]
#%% authors = M. Veithen
#%% keywords = NC, DFPT, NONLINEAR
#%% description =
#%% Linear and nonlinear response calculation for AlAs
#%% Perturbations: electric fields & atomic displacements
#%%<END TEST_INFO>
wlsong
 
Posts: 6
Joined: Mon Apr 01, 2019 4:17 am

Re: DFPT

Postby baguetl » Tue Jun 18, 2019 10:21 am

Dear wlsong,

I am not an expert on Berry phase computations, but I used it a little some years ago.

1) Have you tried kptopt = 3 (instead of 2) ? By the way, I see in your input file that ngkpt = 13 13 1, is the use of one kpoint in the third axis intended?

2) You can use an other implementation of the non-linear response, following the workflow of the test v8/t81 for instance. You can get more information about that method reading the input variable description of "usepead" in the documentation. The convergence with respect to the number of kpoint could be faster, and you can use kptopt = 2 safely.

Best,
Lucas
baguetl
 
Posts: 1
Joined: Mon Oct 05, 2015 9:06 am


Return to Response calculations

Who is online

Users browsing this forum: No registered users and 3 guests

cron