This thread is locked.Only browsing is available.
Top Page > Browsing
Error occurs when NEB calculation runs
Date: 2021/07/21 10:42
Name: Ninomiya

Dear OpenMX developers.

Hi, I have a trouble.

The NEB calculation is performed with Atoms.Number = 161.

We considered the case of 57 oxygen atoms moving.
========
(initial)
57 O 0.26305223526938 0.06680827415674 0.16377421555169 3.0 3.0
(final)
57 O 0.10923829425217 0.09779455947938 0.01173312110049 3.0 3.0
========

However, when we looked at the calculation results, we found that the energy values were very large. (neb.ene file)
========
# 1st column: index of images, where 0 and MD.NEB.Number.Images+1 are the terminals
# 2nd column: Total energy (Hartree) of each image
# 3rd column: distance (Bohr) between neighbors
# 4th column: distance (Bohr) from the image of the index 0
# 5th column: x distance
#
0 -4954.27312597 0.00000000 0.00000000 -11.12678864
1 -96262788022.69296265 18981421613159041499922432.00000000 18981421613159041499922432.00000000 81360357919.32688904
2 -117687575528.02560425 69320177017854817235632128.00000000 88301598631013850145619968.00000000 164472584468.32949829
3 -105771745866.57998657 90220368344328616705785856.00000000 178521966975342449671536640.00000000 119753186234.34959412
4 -4954.27348220 44289392119073775055536128.00000000 222811359094416224727072768.00000000 2.07958699
========

The structure after structural relaxation was also in a very broken state. (# file)
========
(NEB2) 57 O 132005894.21298195421696 820052018.22722733020782 340473157.58966547250748 3.0 3.0
========

The input file was set up as follows.
========
System.CurrrentDirectory ./
System.Name 1
level.of.stdout 1
level.of.fileout 1

Species.Number 2
<Definition.of.Atomic.Species
O O6.0-s2p2d1 O_PBE19
Ti Ti7.0-s3p2d1 Ti_PBE19
Definition.of.Atomic.Species>


Atoms.Number 161
Atoms.SpeciesAndCoordinates.Unit FRAC
<Atoms.SpeciesAndCoordinates
Atoms.SpeciesAndCoordinates>

Atoms.UnitVectors.Unit Ang

<Atoms.UnitVectors
13.778999000000001 0.000000000000000 0.000000000000000
0.000000000000000 13.778999000000001 0.000000000000000
0.000000000000000 0.000000000000000 8.877001000000000
Atoms.UnitVectors>

scf.XcType GGA-PBE
scf.SpinPolarization off
scf.ElectronicTemperature 300.0
scf.energycutoff 220.0
scf.maxIter 100
scf.EigenvalueSolver band
scf.Kgrid 1 1 2
scf.Mixing.Type rmm-diisk

scf.Init.Mixing.Weight 0.01
scf.Min.Mixing.Weight 0.01
scf.Max.Mixing.Weight 0.2
scf.Mixing.History 40
scf.Mixing.StartPulay 20
scf.Mixing.EveryPulay 1
scf.Kerker.factor 5.0

scf.criterion 1.0e-6
scf.maxIter 100

MD.Type NEB # Nomd|Opt|DIIS|NVE|NVT_VS|NVT_NH
MD.NEB.Spring.Const 0.1
MD.Opt.DIIS.History 3 # default=7
MD.Opt.StartDIIS 5 # default=5
MD.maxIter 100 # default=1
MD.TimeStep 1.0 # default=0.5 (fs)
MD.Opt.criterion 0.0003 # default=1.0e-4 (Hartree/bohr)

MD.NEB.Number.Images 3 # default=10

<NEB.Atoms.SpeciesAndCoordinates
NEB.Atoms.SpeciesAndCoordinates>
========

I can' t figure out what is wrong with the input file.
The NEB calculation was done successfully when the number of atoms was 47. Is it difficult to calculate NEB when using a large model?
Please let me know how to avoid this problem.

Best regards,
Ninomiya,
メンテ
Page: [1]

Re: Error occurs when NEB calculation runs ( No.3 )
Date: 2021/07/11 20:22
Name: Naoya Yamaguchi

Hi,

Can you try using 'AU' instead of 'FRAC' for the unit of atomic coordinates?
http://www.openmx-square.org/openmx_man3.9jp/node162.html
http://www.openmx-square.org/openmx_man3.9/node162.html

Regards,
Naoya Yamaguchi
メンテ
Re: Error occurs when NEB calculation runs ( No.4 )
Date: 2021/07/21 10:42
Name: Ninomiya

Dear Prof.Yamaguchi

Thank you very much for your kind support.
The problem is solved.

Best regards,
Ninomiya,
メンテ
Re: Error occurs when NEB calculation runs ( No.5 )
Date: 2021/07/21 10:43
Name: Ninomiya

Dear OpenMX developers.

I tried using 'Ang', but it did not compute well.

Do I need to specify the atomic coordinates in AU instead of Ang?
http://www.openmx-square.org/openmx_man3.9jp/node162.html
http://www.openmx-square.org/openmx_man3.9/node162.html

Or is there a problem with some other parameter?

Best regards,
Ninomiya,

・neb.ene file
========
0 -4954.27304645 0.00000000 0.00000000 2.65221040
1 -4910.75558652 37159.25746197 37159.25746197 2.03915168
2 -4901.56311992 32641.18331847 69800.44078045 2.27328272
3 -4906.23951962 33402.69641407 103203.13719451 -0.03117329
4 -4954.27325706 50505.23397560 153708.37117012 2.07958700
========


・input file
========
System.CurrrentDirectory ./
System.Name 1-2
level.of.stdout 1
level.of.fileout 1

DATA.PATH /system/apps/rhel8/cpu/OpenMX/intel2020u4/3.9/openmx3.9/work/DFT_DATA19

Species.Number 2
<Definition.of.Atomic.Species
O O6.0-s2p2d1 O_PBE19
Ti Ti7.0-s3p2d1 Ti_PBE19
Definition.of.Atomic.Species>


Atoms.Number 161
Atoms.SpeciesAndCoordinates.Unit Ang
<Atoms.SpeciesAndCoordinates
Atoms.SpeciesAndCoordinates>

Atoms.UnitVectors.Unit Ang

<Atoms.UnitVectors
13.778999000000001 0.000000000000000 0.000000000000000
0.000000000000000 13.778999000000001 0.000000000000000
0.000000000000000 0.000000000000000 8.877001000000000
Atoms.UnitVectors>

scf.XcType GGA-PBE
scf.SpinPolarization off
scf.ElectronicTemperature 300.0
scf.energycutoff 220.0
scf.maxIter 100
scf.EigenvalueSolver band
scf.Kgrid 1 1 2
scf.Mixing.Type rmm-diisk

scf.Init.Mixing.Weight 0.01
scf.Min.Mixing.Weight 0.01
scf.Max.Mixing.Weight 0.2
scf.Mixing.History 40
scf.Mixing.StartPulay 20
scf.Mixing.EveryPulay 1
scf.Kerker.factor 5.0

scf.criterion 1.0e-6
scf.maxIter 100

MD.Type NEB # Nomd|Opt|DIIS|NVE|NVT_VS|NVT_NH
MD.NEB.Spring.Const 0.1
MD.Opt.DIIS.History 3 # default=7
MD.Opt.StartDIIS 5 # default=5
MD.maxIter 100 # default=1
MD.TimeStep 1.0 # default=0.5 (fs)
MD.Opt.criterion 0.0003 # default=1.0e-4 (Hartree/bohr)

MD.NEB.Number.Images 3 # default=10

<NEB.Atoms.SpeciesAndCoordinates
NEB.Atoms.SpeciesAndCoordinates>
========
メンテ
Re: Error occurs when NEB calculation runs ( No.8 )
Date: 2021/07/20 10:48
Name: Naoya Yamaguchi

Dear Ninomiya-san,

>Do I need to specify the atomic coordinates in AU instead of Ang?
Why I suggested `AU` instead of `Ang` is that I found a bug in version 3.9, through my practical calculations with `Ang`, that unit conversion between `Ang` and `AU` for atomic coordinates was wrong. And, this problem is serious when one restart the NEB calculations.
I share my modified code of `neb.c`.
https://gist.github.com/Ncmexp2717/607372fa850d3d83b922612dd216e1ba
If it causes your problem, you can solve it by recompiling OpenMX after overwriting `neb.c`.

Regards,
Naoya Yamaguchi
メンテ
Re: Error occurs when NEB calculation runs ( No.9 )
Date: 2021/07/20 13:06
Name: T. Ozaki

Dear Yamaguchi-san,

Thank you for letting us know the possible bug in neb.c.
Indeed, a part of transforming coordinates should be properly modified.

I have checked your code and found that your correction for "if (coordinates_unit==0)" is okay,
while that for "if (coordinates_unit==2)" is incomplete.

I will correct the part of "if (coordinates_unit==2)".

Thank you for your cooperation.

Regards,

TO
メンテ
Re: Error occurs when NEB calculation runs ( No.10 )
Date: 2021/07/20 18:33
Name: Naoya Yamaguchi

Dear Ozaki-sensei,

Thank you for checking my modification.
I hope that `FRAC` mode will be available, especially in NEB calculations of bulk systems.

Regards,
Naoya Yamaguchi
メンテ
Re: Error occurs when NEB calculation runs ( No.11 )
Date: 2021/07/21 10:42
Name: Ninomiya

Dear Prof.Yamaguchi

Thank you very much for your kind support.
The problem is solved.

Best regards,
Ninomiya,
メンテ

Page: [1]