how to define symrel values for LDA+U calculations.

External optional components of Abinit (BigDFT, Wannier90...) and codes using/providing data from/to Abinit (AtomPAW, ONCVPSP, EXC, DP, Yambo...)

Moderators: ebousquet, bxu

Locked
abgs01
Posts: 7
Joined: Tue Dec 10, 2019 12:58 pm

how to define symrel values for LDA+U calculations.

Post by abgs01 » Thu Dec 12, 2019 12:35 pm

Dear abinit user,
I am trying to performe LDA+U calculation of BaNiF4. BaMnF4 with space group Cm c 2_1 (# 36) has two primitive cell in conventional cell. Firstly, I run tudet_1.in and tudet_1.files. Then I used symrel values of tudet_1.out in tudet_2.in. when I run tudet_2.files for input file tudet_2.in But it give me error messages as "Largest error (above) is so large (0.01) that either input atomic coordinates (xred) are wrong or space group symmetry data is wrong. Action: correct your input file."

can you review my in put files.
how can ı correct my input files.

I attached tudet_1.in, tudet_2.in, tudet_1.out, tudet_2.out and tudet_2.log files.

thank you in advance
Attachments
tudet_1.in
(3.78 KiB) Downloaded 483 times
tudet_1.out
(127.2 KiB) Downloaded 450 times
tudet_2.in
(3.81 KiB) Downloaded 460 times
tudet_2.log
(85.59 KiB) Downloaded 435 times
tudet_2.out
(45.29 KiB) Downloaded 449 times

abgs01
Posts: 7
Joined: Tue Dec 10, 2019 12:58 pm

Re: how to define symrel values for LDA+U calculations.

Post by abgs01 » Thu Dec 19, 2019 1:10 pm

dear abinit users,

do you have any idea about my above problem. how can ı overcame this problem.

thank you in advance

ebousquet
Posts: 469
Joined: Tue Apr 19, 2011 11:13 am
Location: University of Liege, Belgium

Re: how to define symrel values for LDA+U calculations.

Post by ebousquet » Thu Dec 19, 2019 8:46 pm

Dear abgs01,
I'm not sure of what you want to do exactly, but to me you should leave abinit calculate the symmetry operation automatically in all your cases. Because you are using non-primitive cells and chkprim=0, abinit does not print the space group but you can check it with Abinit or qAgate that use spglib or you can use a crystallo software suite (like ISOTROPY suite) to verify the space group of your system.
What supposed to be the difference between your two inputs?
Best wishes,
Eric

abgs01
Posts: 7
Joined: Tue Dec 10, 2019 12:58 pm

Re: how to define symrel values for LDA+U calculations.

Post by abgs01 » Wed Jun 10, 2020 4:17 pm

Dear Prof. Eric,

First of all, Thank you for response.

I want to determine U values for BaMnF4. There are two primitive cells in my conventional cell(BaMnF4: space group Cm c 2_1 (# 36)). I reduced unit cell from conventional cell with 24 atoms to primitive cell with 12 atoms. Now there are 12 atoms in unit cell of BaMnF4. I worked tutorial on "How to determine U for DFT+U in ABINIT". In this tutorial, there are 2 atoms in unit cell for Fe. So We get following values, namely for 2 atoms (ii=1, nat=2 (2 atoms)) (ii=2, nat=16 (16 atoms)), (ii=3, nat=54 (54 atoms)), (ii=4, nat=128 (128 atoms)) and (ii=5, nat=250 (250 atoms))


URES ii nat r_max U(J)[eV] U_ASA[eV] U_inf[eV]
URES 1 2 4.69390 4.74555 3.67983 3.20150
URES 2 16 9.38770 8.77694 6.80588 5.92122
URES 3 54 14.08160 9.17082 7.11130 6.18694
URES 4 128 18.77540 9.25647 7.17772 6.24472
URES 5 250 23.46930 9.28509 7.19991 6.26403

however there are 12 atoms in my unit cell, and ı get the same numbers of atoms too. At the same time, as seen below, the calculated U values for BaMnF4 are very large and strange. Because there are 12 atoms in unit cell, i think that it should be at least nat=12 atoms for ii=1, Accordingly, i think that it shoul be nat=96 atoms for ii=2 (2x2x2 super cell) and nat=324 atoms for ii=3 (3x3x3 super cell). The calculated U values for BaMnF4 with 12 atoms are given as following in tudet_2.out. I have also attached files in this calculation. can you check please, what is wrong in my input files.

---------- calculate U, (J) start ----------

All pawujat ok and equal to 1
All macro_uj ok and equal to 1
Matrix response chi_0 is singular
Probably too many symmetries kept
URES ii nat r_max U(J)[eV] U_ASA[eV] U_inf[eV]
URES 1 2 5.83410************************************
URES 2 16 23.82820 67.36377 51.84954 43.30240
URES 3 54 29.90320 70.50823 54.26981 45.32370
URES 4 128 47.65640 71.27692 54.86147 45.81783
URES 5 250 53.65260 71.55077 55.07225 45.99387

---------- calculate U, (J) end --------------

Another problem is that symrel values were used in tudet_2.in. The symrel values are calculated in tudet_1.out. I used these symrel values in tudet_2.in too.

Can i use the calculated symrel values in tudet_1.out for tudet_2.in, or how can ı calculate the used symrel values in tudet_2.in.


Best wishes
Attachments
tudet_1.in
(3.13 KiB) Downloaded 428 times
tudet_1.out
(89.16 KiB) Downloaded 448 times
tudet_2.in
(3.55 KiB) Downloaded 442 times
tudet_2.out
(111.2 KiB) Downloaded 425 times
tudet_2.log
(611.27 KiB) Downloaded 434 times

abgs01
Posts: 7
Joined: Tue Dec 10, 2019 12:58 pm

Re: how to define symrel values for LDA+U calculations.

Post by abgs01 » Mon Jun 29, 2020 12:21 pm

Dear abinit users

is there anyone who can help? or is there anybody calculating U value before?

Best wishes

Locked