[color=red][b]Very weird problem[/b][/color] about K points convergence check

Queries about input and output files, running specific calculations, etc.


Moderators: Global Moderator, Moderator

Message
Author
vasp16888
Newbie
Newbie
Posts: 44
Joined: Fri Apr 23, 2010 3:09 am

[color=red][b]Very weird problem[/b][/color] about K points convergence check

#1 Post by vasp16888 » Fri Apr 23, 2010 11:25 pm

Dear vasp users:
I have been doing Fe12Ga4 system for weeks, but there are some pretty weird problems in the process of kpoints convergence check. There are 12 Fe atoms and 4 Ga atoms in the unit cell, the equilibrium lattice constant is around 5.745 Ang, cubic. see below:
INCAR is always the same:
=====================
SYSTEM = Fe12Ga4
ISTART = 0 # job : 0-new 1-cont 2-samecut
ICHARG = 2
ISPIN = 2 # spin : 1-off 2-on
# MAGMOM = 12 12 -3 -3 -3 -3

# Electronic Relaxation 1
ENCUT = 350 # cut-off energy
PREC = H # precision : medium, high low
NELM = 500 # maximal ELM
EDIFF = 1E-04 # stopping-criterion for ELM
NELMDL = -5 # No. of inconsistent ELM
NELMIN = 4 # minimal ELM

# Electronic relaxation 2 (details)
ALGO = FAST # algorithm
NSIM = 4 # parameter used in RMM-DIIS
LREAL = .FALSE. # space where projection will be done
VOSKOWN = 1 # gga related

AMIX = 0.05
BMIX = 0.0001
AMIX_MAG = 0.8
BMIX_MAG = 0.0001
LMAXMIX = 6
====================

KPOINTS varies:
=================
5*5*5
6*6*6
7*7*7
9*9*9
11*11*11
==================

POSCAR and POTCAR(PAW_PW91) doesn't change.

Here are the results:
kpoint volume total energy(eV) magnet(uB) pressure(kB)
5 189.610000 -111.678440 24.254700 3.780000
6 189.610000 -108.176590 -2.766800 -27.440000
7 189.610000 -111.681880 24.260700 10.310000
9 189.610000 -108.162300 -3.151700 -32.880000
11 189.610000 -108.493080 0.695500 -3.960000

QUESTIONS:
(1)Why there were such a big changes in total energy during the process of kpoints convergence check, I have never seen this before ??
(2)Magnetic moment is ridiculous??
(3)External pressure is weird, too. It should be linear. I don't why??


Looking forward to vasp users's reply, thank you in advance :)
<span class='smallblacktext'>[ Edited ]</span>
Last edited by vasp16888 on Fri Apr 23, 2010 11:25 pm, edited 1 time in total.
[align=center]
AB INITIO STUDY OF MAGNETIC MATERIALS
[/align]

vasp16888
Newbie
Newbie
Posts: 44
Joined: Fri Apr 23, 2010 3:09 am

[color=red][b]Very weird problem[/b][/color] about K points convergence check

#2 Post by vasp16888 » Mon Apr 26, 2010 6:46 am

Any suggestions will be greatly appreciated:)
Last edited by vasp16888 on Mon Apr 26, 2010 6:46 am, edited 1 time in total.
[align=center]
AB INITIO STUDY OF MAGNETIC MATERIALS
[/align]

Danny
Full Member
Full Member
Posts: 201
Joined: Thu Nov 02, 2006 4:35 pm
License Nr.: 5-532
Location: Ghent, Belgium
Contact:

[color=red][b]Very weird problem[/b][/color] about K points convergence check

#3 Post by Danny » Tue Apr 27, 2010 1:37 pm

(1-3) my guess would be that your calculations have not converged. Check if your calculations finished with 500 electronic steps, if so they are probably not converged.

Also set EDIFF to a higher accuracy than 0.001 eg: EDIFF=1.0E-7
Set LMAXMIX = 4 (you have no f electrons in neither Fe or Ga, only d-->=4)
Last edited by Danny on Tue Apr 27, 2010 1:37 pm, edited 1 time in total.

panda

[color=red][b]Very weird problem[/b][/color] about K points convergence check

#4 Post by panda » Tue Apr 27, 2010 5:17 pm

I typically set EDIFFG = -0.01 to achieve convergence of forces to 0.01 eV/atom or smaller. But I agree with Danny, I think the calculation is not converged. Have you tried setting auto kpoints? What about setting ISYM = 0, to turn off symmetry? These are a few things I would try off the top of my head.
Last edited by panda on Tue Apr 27, 2010 5:17 pm, edited 1 time in total.

Danny
Full Member
Full Member
Posts: 201
Joined: Thu Nov 02, 2006 4:35 pm
License Nr.: 5-532
Location: Ghent, Belgium
Contact:

[color=red][b]Very weird problem[/b][/color] about K points convergence check

#5 Post by Danny » Wed Apr 28, 2010 9:19 am

[quote="red"]EDIFFG = -0.01 [/color]to achieve convergence of forces to 0.01 eV/atom or smaller. But I agree with Danny, I think the calculation is not converged. Have you tried setting auto kpoints? What about setting ISYM = 0, to turn off symmetry? These are a few things I would try off the top of my head. [/quote]</span>
Last edited by Danny on Wed Apr 28, 2010 9:19 am, edited 1 time in total.

vasp16888
Newbie
Newbie
Posts: 44
Joined: Fri Apr 23, 2010 3:09 am

[color=red][b]Very weird problem[/b][/color] about K points convergence check

#6 Post by vasp16888 » Thu Apr 29, 2010 12:20 am

[quote author=1.0E-7
Set LMAXMIX = 4 (you have no f electrons in neither Fe or Ga, only d-->=4)[/quote]Previous calculation:[/size]

INCAR: doesn't change
=====================
SYSTEM = Fe12Ga4
ISTART = 0 # job : 0-new 1-cont 2-samecut
ICHARG = 2
ISPIN = 2 # spin : 1-off 2-on
# MAGMOM = 12 12 -3 -3 -3 -3

# Electronic Relaxation 1
ENCUT = 350 # cut-off energy
PREC = H # precision : medium, high low
NELM = 500 # maximal ELM
EDIFF = 1E-04 # stopping-criterion for ELM
NELMDL = -5 # No. of inconsistent ELM
NELMIN = 4 # minimal ELM

ISMEAR = 0 # smearing method : -5-tet -1-fermi 0-gaus 1-MP
SIGMA = 0.1 # smearing factor in eV
# NPAR = 1

# Electronic relaxation 2 (details)
ALGO = FAST # algorithm
NSIM = 4 # parameter used in RMM-DIIS
LREAL = .FALSE. # space where projection will be done
VOSKOWN = 1 # gga related

AMIX = 0.05
BMIX = 0.0001
AMIX_MAG = 0.8
BMIX_MAG = 0.0001
LMAXMIX = 6
====================

KPOINTS:
==================
Automatic generation
0
Monkhorst-Pack
5 5 5
0 0 0
===================
then just change 5*5*5 to 6*6*6, 7*7*7, 9*9*9, 11*11*11

OSZICAR:

k=5*5*5
---------------------------------------
RMM: 20 -0.111283325533E+03 0.19571E+00 -0.43035E-02 4655 0.919E-01 0.696E+00
RMM: 21 -0.111680611489E+03 -0.39729E+00 -0.11162E+00 4407 0.393E+00 0.929E-01
RMM: 22 -0.111678216825E+03 0.23947E-02 -0.11303E-01 5028 0.104E+00 0.730E-01
RMM: 23 -0.111672368245E+03 0.58486E-02 -0.77294E-02 4239 0.104E+00 0.168E+00
RMM: 24 -0.111647388568E+03 0.24980E-01 -0.28888E-02 4962 0.575E-01 0.229E+00
RMM: 25 -0.111679428227E+03 -0.32040E-01 -0.99002E-02 4128 0.120E+00 0.492E-01
RMM: 26 -0.111677431050E+03 0.19972E-02 -0.57489E-03 4952 0.365E-01 0.436E-01
RMM: 27 -0.111677816538E+03 -0.38549E-03 -0.44294E-03 4074 0.271E-01 0.349E-01
RMM: 28 -0.111678820209E+03 -0.10037E-02 -0.51031E-03 4140 0.257E-01 0.463E-01
RMM: 29 -0.111678383570E+03 0.43664E-03 -0.85971E-04 4138 0.992E-02 0.310E-01
RMM: 30 -0.111677786562E+03 0.59701E-03 -0.16376E-03 4099 0.152E-01 0.254E-01
RMM: 31 -0.111678457961E+03 -0.67140E-03 -0.10202E-03 4081 0.140E-01 0.117E-01
RMM: 32 -0.111678435335E+03 0.22626E-04 -0.59160E-04 3621 0.868E-02

1 F= -.11167844E+03 E0= -.11164138E+03 d E =-.741059E-01 mag= 24.2547
---------------------------------------

k=6*6*6
---------------------------------------
RMM: 148 -0.108168379365E+03 0.11597E-01 -0.85943E-04 3631 0.101E-01 0.331E+00
RMM: 149 -0.108119408803E+03 0.48971E-01 -0.15050E-02 3886 0.448E-01 0.311E+00
RMM: 150 -0.108080605796E+03 0.38803E-01 -0.14679E-02 4087 0.396E-01 0.302E+00
RMM: 151 -0.108034389434E+03 0.46216E-01 -0.21590E-02 4428 0.509E-01 0.308E+00
RMM: 152 -0.108042963915E+03 -0.85745E-02 -0.12768E-03 4921 0.152E-01 0.301E+00
RMM: 153 -0.108091374135E+03 -0.48410E-01 -0.14666E-02 3868 0.473E-01 0.302E+00
RMM: 154 -0.108130613200E+03 -0.39239E-01 -0.15737E-02 4170 0.409E-01 0.312E+00
RMM: 155 -0.108148708333E+03 -0.18095E-01 -0.53631E-03 4890 0.230E-01 0.320E+00
RMM: 156 -0.108197917340E+03 -0.49209E-01 -0.15338E-02 3963 0.453E-01 0.342E+00
RMM: 157 -0.108190501621E+03 0.74157E-02 -0.10928E-03 4758 0.131E-01 0.341E+00
RMM: 158 -0.108176620225E+03 0.13881E-01 -0.10254E-03 3780 0.141E-01 0.334E+00
RMM: 159 -0.108176586690E+03 0.33535E-04 -0.14454E-04 1813 0.393E-02

1 F= -.10817659E+03 E0= -.10809401E+03 d E =-.165154E+00 mag= -2.7668
---------------------------------------

k=7*7*7
---------------------------------------
RMM: 50 -0.111600164334E+03 0.55678E-01 -0.52679E-02 8451 0.859E-01 0.274E+00
RMM: 51 -0.111622788680E+03 -0.22624E-01 -0.38883E-03 9581 0.352E-01 0.230E+00
RMM: 52 -0.111632284138E+03 -0.94955E-02 -0.18463E-03 7773 0.194E-01 0.215E+00
RMM: 53 -0.111646531219E+03 -0.14247E-01 -0.44595E-03 8528 0.233E-01 0.187E+00
RMM: 54 -0.111664232618E+03 -0.17701E-01 -0.10736E-02 7901 0.343E-01 0.153E+00
RMM: 55 -0.111677710081E+03 -0.13477E-01 -0.12610E-02 8291 0.361E-01 0.123E+00
RMM: 56 -0.111684275888E+03 -0.65658E-02 -0.13826E-02 8572 0.386E-01 0.873E-01
RMM: 57 -0.111683896503E+03 0.37939E-03 -0.15160E-03 9829 0.122E-01 0.792E-01
RMM: 58 -0.111683084285E+03 0.81222E-03 -0.29419E-04 5816 0.812E-02 0.710E-01
RMM: 59 -0.111682754646E+03 0.32964E-03 -0.10115E-04 4083 0.407E-02 0.712E-01
RMM: 60 -0.111682051388E+03 0.70326E-03 -0.15781E-03 8032 0.155E-01 0.418E-01
RMM: 61 -0.111681932632E+03 0.11876E-03 -0.58174E-04 7962 0.749E-02 0.350E-01
RMM: 62 -0.111681884882E+03 0.47750E-04 -0.13478E-04 4182 0.416E-02

1 F= -.11168188E+03 E0= -.11165035E+03 d E =-.630775E-01 mag= 24.2607
---------------------------------------

k=9*9*9
---------------------------------------
RMM: 168 -0.108150331115E+03 0.18368E-03 -0.27694E-04 10430 0.614E-02 0.311E+00
RMM: 169 -0.108145891344E+03 0.44398E-02 -0.10736E-04 6339 0.486E-02 0.309E+00
RMM: 170 -0.108163107448E+03 -0.17216E-01 -0.17037E-03 13208 0.166E-01 0.315E+00
RMM: 171 -0.108197147719E+03 -0.34040E-01 -0.71713E-03 13598 0.305E-01 0.328E+00
RMM: 172 -0.108166342860E+03 0.30805E-01 -0.33914E-03 14416 0.264E-01 0.319E+00
RMM: 173 -0.108145846633E+03 0.20496E-01 -0.35905E-03 14551 0.203E-01 0.311E+00
RMM: 174 -0.108111552167E+03 0.34294E-01 -0.90862E-03 14525 0.337E-01 0.303E+00
RMM: 175 -0.108096758737E+03 0.14793E-01 -0.38420E-03 16583 0.186E-01 0.299E+00
RMM: 176 -0.108170828566E+03 -0.74070E-01 -0.31247E-02 13617 0.676E-01 0.315E+00
RMM: 177 -0.108172925563E+03 -0.20970E-02 -0.34569E-03 16950 0.186E-01 0.320E+00
RMM: 178 -0.108162356535E+03 0.10569E-01 -0.49575E-04 11499 0.841E-02 0.317E+00
RMM: 179 -0.108162302471E+03 0.54064E-04 -0.60437E-05 5700 0.329E-02

1 F= -.10816230E+03 E0= -.10808193E+03 d E =-.160739E+00 mag= -3.1517
---------------------------------------

k=11*11*11
---------------------------------------
RMM: 47 -0.108993242351E+03 -0.12171E+00 -0.21586E-02 22670 0.546E-01 0.748E+00
RMM: 48 -0.109028387527E+03 -0.35145E-01 -0.47797E-03 26248 0.231E-01 0.794E+00
RMM: 49 -0.109022832047E+03 0.55555E-02 -0.54860E-04 23240 0.114E-01 0.789E+00
RMM: 50 -0.108985209560E+03 0.37622E-01 -0.15730E-03 21041 0.164E-01 0.752E+00
RMM: 51 -0.108925327631E+03 0.59882E-01 -0.71428E-03 21420 0.317E-01 0.676E+00
RMM: 52 -0.108820407862E+03 0.10492E+00 -0.18451E-02 21698 0.512E-01 0.564E+00
RMM: 53 -0.108752344791E+03 0.68063E-01 -0.16929E-02 23137 0.438E-01 0.495E+00
RMM: 54 -0.108684708222E+03 0.67637E-01 -0.20206E-02 24270 0.513E-01 0.443E+00
RMM: 55 -0.108580607577E+03 0.10410E+00 -0.49742E-02 22537 0.798E-01 0.410E+00
RMM: 56 -0.108494211707E+03 0.86396E-01 -0.44857E-02 24090 0.696E-01 0.405E+00
RMM: 57 -0.108493065299E+03 0.11464E-02 -0.48060E-03 30974 0.205E-01 0.400E+00
RMM: 58 -0.108493078108E+03 -0.12809E-04 -0.28536E-04 14274 0.743E-02

1 F= -.10849308E+03 E0= -.10841875E+03 d E =-.148666E+00 mag= 0.6955
---------------------------------------

POSCAR and POTCAR(PAW_PW91) doesn't change.

Here are the results:
kpoint volume total energy(eV) magnet(uB) pressure(kB)
--------------------------------------------------------
5 189.610000 -111.678440 24.254700 3.780000
6 189.610000 -108.176590 -2.766800 -27.440000
7 189.610000 -111.681880 24.260700 10.310000
9 189.610000 -108.162300 -3.151700 -32.880000
11 189.610000 -108.493080 0.695500 -3.960000
--------------------------------------------------------
Last edited by vasp16888 on Thu Apr 29, 2010 12:20 am, edited 1 time in total.
[align=center]
AB INITIO STUDY OF MAGNETIC MATERIALS
[/align]

vasp16888
Newbie
Newbie
Posts: 44
Joined: Fri Apr 23, 2010 3:09 am

[color=red][b]Very weird problem[/b][/color] about K points convergence check

#7 Post by vasp16888 » Thu Apr 29, 2010 12:38 am

[quote="panda"]I typically set EDIFFG = -0.01 to achieve convergence of forces to 0.01 eV/atom or smaller. But I agree with Danny, I think the calculation is not converged. Have you tried setting auto kpoints? What about setting ISYM = 0, to turn off symmetry? These are a few things I would try off the top of my head. [/quote]


After adopting your suggestions, here are the details:

Latest Calculation

INCAR: doesn't change, the different from previous INCAR are written in red.
=====================
SYSTEM = Fe12Ga4
ISTART = 0 # job : 0-new 1-cont 2-samecut
ICHARG = 2
ISPIN = 2 # spin : 1-off 2-on
# MAGMOM = 12 12 -3 -3 -3 -3

# Electronic Relaxation 1
ENCUT = 350 # cut-off energy
PREC = H # precision : medium, high low
NELM = 500 # maximal ELM
EDIFF = 1E-07 # stopping-criterion for ELM
NELMDL = -5 # No. of inconsistent ELM
NELMIN = 4 # minimal ELM

ISMEAR = 0 # smearing method : -5-tet -1-fermi 0-gaus 1-MP
SIGMA = 0.1 # smearing factor in eV
NPAR = 6
LPLANE = .TRUE.


# Electronic relaxation 2 (details)
ALGO = FAST # algorithm
NSIM = 4 # parameter used in RMM-DIIS
LREAL = .FALSE. # space where projection will be done
VOSKOWN = 1 # gga related

AMIX = 0.05
BMIX = 0.0001
AMIX_MAG = 0.8
BMIX_MAG = 0.0001
LMAXMIX = 4
====================

KPOINTS:
==================
Automatic generation
0
Monkhorst-Pack
5 5 5
0 0 0
===================
then just change 5*5*5 to 6*6*6, 7*7*7, 9*9*9, 11*11*11

OSZICAR:

k=5*5*5
---------------------------------------
RMM: 27 -0.111677870873E+03 -0.56617E-03 -0.44877E-03 4510 0.268E-01 0.341E-01
RMM: 28 -0.111678836605E+03 -0.96573E-03 -0.47833E-03 4563 0.240E-01 0.463E-01
RMM: 29 -0.111678352588E+03 0.48402E-03 -0.88602E-04 5251 0.100E-01 0.312E-01
RMM: 30 -0.111677776630E+03 0.57596E-03 -0.15351E-03 4545 0.150E-01 0.257E-01
RMM: 31 -0.111678438154E+03 -0.66152E-03 -0.95702E-04 4706 0.135E-01 0.117E-01
RMM: 32 -0.111678436196E+03 0.19582E-05 -0.65725E-04 4810 0.931E-02 0.667E-02
RMM: 33 -0.111678428530E+03 0.76663E-05 -0.51030E-05 5495 0.235E-02 0.486E-02
RMM: 34 -0.111678437531E+03 -0.90010E-05 -0.23981E-05 4627 0.206E-02 0.343E-02
RMM: 35 -0.111678441079E+03 -0.35487E-05 -0.10101E-04 4466 0.375E-02 0.188E-02
RMM: 36 -0.111678445701E+03 -0.46217E-05 -0.64139E-06 4796 0.136E-02 0.482E-03
RMM: 37 -0.111678445979E+03 -0.27762E-06 -0.79699E-07 3762 0.437E-03 0.455E-03
RMM: 38 -0.111678445887E+03 0.91250E-07 -0.52504E-07 3686 0.322E-03

1 F= -.11167845E+03 E0= -.11164111E+03 d E =-.746690E-01 mag= 24.2901
---------------------------------------

k=6*6*6
---------------------------------------
RMM: 83 -0.111677595217E+03 0.42252E-03 -0.54862E-04 5095 0.898E-02 0.452E-01
RMM: 84 -0.111677880745E+03 -0.28553E-03 -0.11310E-04 5096 0.400E-02 0.429E-01
RMM: 85 -0.111678805103E+03 -0.92436E-03 -0.52792E-04 4404 0.820E-02 0.287E-01
RMM: 86 -0.111679456576E+03 -0.65147E-03 -0.98743E-04 4346 0.108E-01 0.895E-02
RMM: 87 -0.111679520992E+03 -0.64416E-04 -0.23460E-04 5864 0.415E-02 0.682E-02
RMM: 88 -0.111679571324E+03 -0.50331E-04 -0.22876E-04 4754 0.548E-02 0.908E-03
RMM: 89 -0.111679570072E+03 0.12516E-05 -0.17183E-05 5421 0.150E-02 0.467E-03
RMM: 90 -0.111679568995E+03 0.10768E-05 -0.17361E-06 4095 0.643E-03 0.313E-03
RMM: 91 -0.111679568303E+03 0.69228E-06 -0.15764E-06 3916 0.477E-03 0.399E-03
RMM: 92 -0.111679568032E+03 0.27037E-06 -0.12976E-07 2533 0.185E-03 0.228E-03
RMM: 93 -0.111679568239E+03 -0.20684E-06 -0.34602E-07 2626 0.274E-03 0.105E-03
RMM: 94 -0.111679568326E+03 -0.87055E-07 -0.26372E-07 2285 0.182E-03

1 F= -.11167957E+03 E0= -.11165156E+03 d E =-.560229E-01 mag= 24.2457
---------------------------------------

k=7*7*7
---------------------------------------
RMM: 485 -0.895735851093E+02 0.55599E+01 -0.12895E+00 9385 0.360E+00 0.484E+01
RMM: 486 -0.894469503523E+02 0.12663E+00 -0.11172E-01 11408 0.123E+00 0.487E+01
RMM: 487 -0.894310030744E+02 0.15947E-01 -0.68213E-03 10703 0.388E-01 0.486E+01
RMM: 488 -0.880398503903E+02 0.13912E+01 -0.41753E-02 8953 0.777E-01 0.507E+01
RMM: 489 -0.883093739127E+02 -0.26952E+00 -0.28870E-03 11130 0.250E-01 0.503E+01
RMM: 490 -0.920308273057E+02 -0.37215E+01 -0.28494E-01 8469 0.206E+00 0.442E+01
RMM: 491 -0.987837561096E+02 -0.67529E+01 -0.16794E+00 8583 0.473E+00 0.313E+01
RMM: 492 -0.989537565027E+02 -0.17000E+00 -0.17078E-01 11416 0.112E+00 0.313E+01
RMM: 493 -0.994996015846E+02 -0.54585E+00 -0.38561E-02 9398 0.804E-01 0.302E+01
RMM: 494 -0.982127354766E+02 0.12869E+01 -0.78167E-02 8827 0.117E+00 0.330E+01
RMM: 495 -0.982202479215E+02 -0.75124E-02 -0.75002E-03 10899 0.305E-01 0.330E+01
RMM: 496 -0.806420291262E+02 0.17578E+02 -0.70336E+00 8642 0.101E+01 0.607E+01
RMM: 497 -0.885772051589E+02 -0.79352E+01 -0.66115E-01 9982 0.465E+00 0.498E+01
RMM: 498 -0.852941125640E+02 0.32831E+01 -0.23283E-01 10401 0.183E+00 0.542E+01
RMM: 499 -0.892206185422E+02 -0.39265E+01 -0.23979E-01 9126 0.192E+00 0.483E+01
RMM: 500 -0.874389343391E+02 0.17817E+01 -0.48200E-02 9745 0.112E+00

1 F= -.87438934E+02 E0= -.87379679E+02 d E =-.118510E+00 mag= -23.7551
I don't think it will converge even in more scf steps
---------------------------------------

k=9*9*9
---------------------------------------
RMM: 486 -0.995129104111E+02 0.19360E+01 -0.19486E-01 16526 0.196E+00 0.300E+01
RMM: 487 -0.100553321393E+03 -0.10404E+01 -0.44764E-02 16909 0.101E+00 0.278E+01
RMM: 488 -0.100382046224E+03 0.17128E+00 -0.38642E-03 17752 0.328E-01 0.282E+01
RMM: 489 -0.100262336510E+03 0.11971E+00 -0.21120E-03 16046 0.183E-01 0.284E+01
RMM: 490 -0.100304038906E+03 -0.41702E-01 -0.39043E-04 16978 0.873E-02 0.283E+01
RMM: 491 -0.103708661762E+03 -0.34046E+01 -0.11546E+00 14718 0.401E+00 0.209E+01
RMM: 492 -0.103064846528E+03 0.64382E+00 -0.76118E-02 18514 0.157E+00 0.226E+01
RMM: 493 -0.102755783911E+03 0.30906E+00 -0.28556E-02 14712 0.820E-01 0.233E+01
RMM: 494 -0.102709544044E+03 0.46240E-01 -0.18482E-03 18857 0.190E-01 0.234E+01
RMM: 495 -0.103095625297E+03 -0.38608E+00 -0.12141E-02 15701 0.405E-01 0.225E+01
RMM: 496 -0.103348079853E+03 -0.25245E+00 -0.13657E-02 17334 0.375E-01 0.219E+01
RMM: 497 -0.103323241256E+03 0.24839E-01 -0.61902E-04 19119 0.957E-02 0.220E+01
RMM: 498 -0.103471235867E+03 -0.14799E+00 -0.26991E-03 15704 0.189E-01 0.217E+01
RMM: 499 -0.103223978084E+03 0.24726E+00 -0.10009E-02 15458 0.372E-01 0.223E+01
RMM: 500 -0.102945349557E+03 0.27863E+00 -0.12641E-02 14983 0.382E-01

1 F= -.10294535E+03 E0= -.10285887E+03 d E =-.172962E+00 mag= 8.6635
I don't think it will converge even in more scf steps
---------------------------------------

k=11*11*11
---------------------------------------
RMM: 287 -0.108250460630E+03 -0.14956E-01 -0.78590E-03 32279 0.249E-01 0.351E+00
RMM: 288 -0.108247328098E+03 0.31325E-02 -0.29016E-04 29229 0.797E-02 0.351E+00
RMM: 289 -0.108223982134E+03 0.23346E-01 -0.26142E-03 24299 0.191E-01 0.342E+00
RMM: 290 -0.108202861301E+03 0.21121E-01 -0.33591E-03 24524 0.197E-01 0.332E+00
RMM: 291 -0.108188624321E+03 0.14237E-01 -0.21556E-03 27902 0.153E-01 0.325E+00
RMM: 292 -0.108178789020E+03 0.98353E-02 -0.10314E-03 27143 0.106E-01 0.321E+00
RMM: 293 -0.108145291881E+03 0.33497E-01 -0.73141E-03 24199 0.309E-01 0.309E+00
RMM: 294 -0.108106869426E+03 0.38422E-01 -0.11760E-02 24537 0.373E-01 0.300E+00
RMM: 295 -0.108026794737E+03 0.80075E-01 -0.49126E-02 24953 0.781E-01 0.314E+00
RMM: 296 -0.107595703835E+03 0.43109E+00 -0.53531E-01 23975 0.267E+00 0.678E+00
RMM: 297 -0.104458892802E+03 0.31368E+01 -0.34842E+00 23889 0.675E+00 0.198E+01
RMM: 298 -0.916916459094E+02 0.12767E+02 -0.10606E+01 24200 0.117E+01 0.452E+01
RMM: 299 -0.956210593731E+02 -0.39294E+01 -0.49218E-01 29458 0.354E+00 0.375E+01
RMM: 300 0.662913397589E+02 0.16191E+03 -0.14646E+02 27035 0.444E+01 0.995E+01
RMM: 301 0.212725816682E+04 0.20610E+04 -0.89854E+02 28938 0.102E+02 0.182E+02
RMM: 302 0.103331901608E+04 -0.10939E+04 -0.23575E+01 28566 0.265E+01 0.150E+02
RMM: 303 -0.443102484515E+02 -0.10776E+04 -0.45769E+02 29685 0.642E+01 0.118E+02
RMM: 304 0.202799938084E+05 0.20324E+05 -0.28860E+03 25549 0.480E+02 0.445E+02
RMM: 305 0.399097248906E+05 0.19630E+05 0.75559E+03 36766 0.127E+03 0.848E+02
RMM: 306 0.238530753437E+06 0.19862E+06 -0.13244E+04 34101 0.987E+02 0.330E+03
RMM: 307 0.372936232654E+05 -0.20124E+06 -0.24999E+04 31531 0.217E+03 0.652E+02

It oscillate so severely that I stop running it, and I don't think it will converge, eigher.
---------------------------------------

POSCAR and POTCAR(PAW_PW91) doesn't change.

Here are the results:
kpoint volume total energy(eV) magnet(uB) pressure(kB)
--------------------------------------------------------
5 189.610000 -111.678450 24.290100 3.380000
6 189.610000 -111.679570 24.245700 3.110000
7 189.610000 -87.438934 -23.755100 218.720000
9 189.610000 -102.945350 8.663500 42.700000
11......................
since it didn't converge when k=7*7*7, 9*9*9 and 11*11*1, the total energy and the external pressure is ridiculous
--------------------------------------------------------

I was totally confused about this so wierd convergence problem of kpoints.

Any further suggestions will be greatly appreciated:)
<span class='smallblacktext'>[ Edited Thu Apr 29 2010, 02:43AM ]</span>
Last edited by vasp16888 on Thu Apr 29, 2010 12:38 am, edited 1 time in total.
[align=center]
AB INITIO STUDY OF MAGNETIC MATERIALS
[/align]

vasp16888
Newbie
Newbie
Posts: 44
Joined: Fri Apr 23, 2010 3:09 am

[color=red][b]Very weird problem[/b][/color] about K points convergence check

#8 Post by vasp16888 » Thu Apr 29, 2010 1:00 am

[quote=" -0.01 to achieve convergence of forces to 0.01 eV/atom or smaller. But I agree with Danny, I think the calculation is not converged. Have you tried setting auto kpoints? What about setting ISYM = 0, to turn off symmetry? These are a few things I would try off the top of my head. [/quote"]ISYM = 0 calculation [/size]

INCAR: doesn't change, the different from previous INCAR are written in red.
=====================
SYSTEM = Fe12Ga4
ISTART = 0 # job : 0-new 1-cont 2-samecut
ICHARG = 2
ISPIN = 2 # spin : 1-off 2-on
# MAGMOM = 12 12 -3 -3 -3 -3

# Electronic Relaxation 1
ENCUT = 350 # cut-off energy
PREC = H # precision : medium, high low
NELM = 500 # maximal ELM
EDIFF = 1E-07 # stopping-criterion for ELM
NELMDL = -5 # No. of inconsistent ELM
NELMIN = 4 # minimal ELM
ISYM = 0 # symmetry : 0-no 1,2-yes


ISMEAR = 0 # smearing method : -5-tet -1-fermi 0-gaus 1-MP
SIGMA = 0.1 # smearing factor in eV
NPAR = 6
LPLANE = .TRUE.


# Electronic relaxation 2 (details)
ALGO = FAST # algorithm
NSIM = 4 # parameter used in RMM-DIIS
LREAL = .FALSE. # space where projection will be done
VOSKOWN = 1 # gga related

AMIX = 0.05
BMIX = 0.0001
AMIX_MAG = 0.8
BMIX_MAG = 0.0001
LMAXMIX = 4
====================

KPOINTS:
==================
Automatic generation
0
Monkhorst-Pack
9 9 9
0 0 0
===================

OSZICAR:
--------------------------------------------
RMM: 223 -0.107959180338E+03 -0.11332E-01 -0.27890E-03172334 0.213E-01 0.390E+00
RMM: 224 -0.107974301711E+03 -0.15121E-01 -0.22454E-03186177 0.167E-01 0.385E+00
RMM: 225 -0.108024790648E+03 -0.50489E-01 -0.19489E-02168190 0.496E-01 0.337E+00
RMM: 226 -0.108094471987E+03 -0.69681E-01 -0.85328E-02173548 0.103E+00 0.317E+00
RMM: 227 -0.108093885889E+03 0.58610E-03 -0.74237E-03216733 0.338E-01 0.322E+00
RMM: 228 -0.108162098408E+03 -0.68213E-01 -0.38230E-02164470 0.709E-01 0.315E+00
RMM: 229 -0.108119265669E+03 0.42833E-01 -0.11829E-02188374 0.487E-01 0.308E+00
RMM: 230 -0.108071417685E+03 0.47848E-01 -0.82481E-02175240 0.984E-01 0.306E+00
RMM: 231 -0.107947444428E+03 0.12397E+00 -0.16308E-01172475 0.152E+00 0.399E+00
RMM: 232 -0.107917786330E+03 0.29658E-01 -0.30131E-02227474 0.580E-01 0.412E+00
RMM: 233 -0.107865330399E+03 0.52456E-01 -0.17932E-02175330 0.504E-01 0.455E+00
RMM: 234 -0.107883703106E+03 -0.18373E-01 -0.22844E-03205831 0.227E-01 0.441E+00
RMM: 235 -0.107869472274E+03 0.14231E-01 -0.49968E-03181145 0.256E-01 0.451E+00
RMM: 236 -0.107679535395E+03 0.18994E+00 -0.17596E-01162965 0.149E+00 0.608E+00
RMM: 237 -0.107795880657E+03 -0.11635E+00 -0.55219E-02194203 0.984E-01 0.512E+00
RMM: 238 -0.107776080293E+03 0.19800E-01 -0.34938E-02173692 0.724E-01 0.520E+00
RMM: 239 -0.107871835981E+03 -0.95756E-01 -0.67238E-02177357 0.935E-01 0.427E+00
RMM: 240 -0.107805910608E+03 0.65925E-01 -0.17389E-02195769 0.567E-01 0.486E+00
RMM: 241 -0.107857358027E+03 -0.51447E-01 -0.31090E-02188902 0.628E-01 0.439E+00
RMM: 242 -0.107847634980E+03 0.97230E-02 -0.10842E-02231030 0.326E-01 0.451E+00
RMM: 243 -0.107848865020E+03 -0.12300E-02 -0.21536E-03188188 0.166E-01 0.453E+00
RMM: 244 -0.107845232159E+03 0.36329E-02 -0.28480E-04224754 0.627E-02 0.456E+00
RMM: 245 -0.107873765608E+03 -0.28533E-01 -0.35797E-03164907 0.216E-01 0.437E+00
RMM: 246 -0.108053842087E+03 -0.18008E+00 -0.19848E-01159157 0.159E+00 0.318E+00
RMM: 247 -0.108059977810E+03 -0.61357E-02 -0.22967E-02204384 0.669E-01 0.332E+00
RMM: 248 -0.108077626429E+03 -0.17649E-01 -0.83450E-03202329 0.380E-01 0.325E+00
RMM: 249 -0.108077835931E+03 -0.20950E-03 -0.98745E-04226087 0.956E-02 0.327E+00
RMM: 250 -0.108073873150E+03 0.39628E-02 -0.10175E-03174521 0.118E-01 0.323E+00
RMM: 251 -0.108116495900E+03 -0.42623E-01 -0.31043E-02165586 0.621E-01 0.319E+00

It oscillate all the time, I don't see any convergence phenomenon.
--------------------------------------------

100% confused:(

Any suggestion after seeing this wierd result? Thanks:)
Last edited by vasp16888 on Thu Apr 29, 2010 1:00 am, edited 1 time in total.
[align=center]
AB INITIO STUDY OF MAGNETIC MATERIALS
[/align]

Danny
Full Member
Full Member
Posts: 201
Joined: Thu Nov 02, 2006 4:35 pm
License Nr.: 5-532
Location: Ghent, Belgium
Contact:

[color=red][b]Very weird problem[/b][/color] about K points convergence check

#9 Post by Danny » Fri Apr 30, 2010 8:20 am

Here are the results:
kpoint volume total energy(eV) magnet(uB) pressure(kB)
--------------------------------------------------------
5 189.610000 -111.678450 24.290100 3.380000
6 189.610000 -111.679570 24.245700 3.110000
7 189.610000 -87.438934 -23.755100 218.720000
9 189.610000 -102.945350 8.663500 42.700000
11......................
interesting.
* why not use a k-point grid without Gamma (even k-point numbers like 4x4x4,6x6x6,8x8x8)
* other option try to use the Davidson algo instead of RMM.

Danny
Last edited by Danny on Fri Apr 30, 2010 8:20 am, edited 1 time in total.

vasp16888
Newbie
Newbie
Posts: 44
Joined: Fri Apr 23, 2010 3:09 am

[color=red][b]Very weird problem[/b][/color] about K points convergence check

#10 Post by vasp16888 » Fri Apr 30, 2010 6:09 pm

ok, I am gonna do it now, thanks. I will keep you posted as soon as I get the result:)
Last edited by vasp16888 on Fri Apr 30, 2010 6:09 pm, edited 1 time in total.
[align=center]
AB INITIO STUDY OF MAGNETIC MATERIALS
[/align]

vasp16888
Newbie
Newbie
Posts: 44
Joined: Fri Apr 23, 2010 3:09 am

[color=red][b]Very weird problem[/b][/color] about K points convergence check

#11 Post by vasp16888 » Sat May 01, 2010 2:37 am

[quote="red"]So my questions are:[/color]
1. If I wanna optimize the structure, and lattice constant change from 5.70, 5.71,5.72,.....,to 5.80Ang. In this 11 calculations, some of them only converge with RMM, but some of them can only converge with DAV. So can I use the data calculated by these two algorithms to find the minimum total energy point and take this point as the optimized structure ???

2. What's difference in principle between these two algorithms ? Why for some case RMM work, and for some case DAV work ?? (I always think RMM is more accurate than DAV.)

Looking forward to your reply:)
Incidentally, Happy Labour's Day:)
Last edited by vasp16888 on Sat May 01, 2010 2:37 am, edited 1 time in total.
[align=center]
AB INITIO STUDY OF MAGNETIC MATERIALS
[/align]

alex
Hero Member
Hero Member
Posts: 585
Joined: Tue Nov 16, 2004 2:21 pm
License Nr.: 5-67
Location: Germany

[color=red][b]Very weird problem[/b][/color] about K points convergence check

#12 Post by alex » Mon May 03, 2010 7:18 am

Hi,

1) Davidson algorithm (DAV, ALGO=N) is much more stable than RMM, but slower. Read the manual for that. So I doubt that you have convergence at all, if DAV does _not_ work.
2) Again, it does not hurt to take a look at the manual and read the appropriate sections. And: there is _no_ difference in accuracy, just stability and speed (and, of course, the algorithm).

alex
Last edited by alex on Mon May 03, 2010 7:18 am, edited 1 time in total.

vasp16888
Newbie
Newbie
Posts: 44
Joined: Fri Apr 23, 2010 3:09 am

[color=red][b]Very weird problem[/b][/color] about K points convergence check

#13 Post by vasp16888 » Tue May 04, 2010 12:05 am

[quote author=N) is much more stable than RMM, but slower. Read the manual for that. So I doubt that you have convergence at all, if DAV does _not_ work.
2) Again, it does not hurt to take a look at the manual and read the appropriate sections. And: there is _no_ difference in accuracy, just stability and speed (and, of course, the algorithm).

alex[/quote]For instance: [/size]
(1)About Fe12Ga4 system, when lattice constant is 5.7828:
all other parameter are the same, I just change the algorithm parameter ALGO.
when ALGO = FAST (RMM), it converges:
=============================
RMM: 30 -0.111661824649E+03 -0.46908E-04 -0.88900E-05 9121 0.389E-02 0.327E-02
RMM: 31 -0.111661818381E+03 0.62679E-05 -0.15566E-04 8885 0.467E-02 0.489E-02
RMM: 32 -0.111661840084E+03 -0.21703E-04 -0.13604E-05 10484 0.208E-02 0.299E-02
RMM: 33 -0.111661849427E+03 -0.93426E-05 -0.22888E-05 8085 0.199E-02 0.820E-03
RMM: 34 -0.111661849845E+03 -0.41864E-06 -0.78416E-06 10117 0.812E-03 0.262E-03
RMM: 35 -0.111661849949E+03 -0.10341E-06 -0.21505E-06 8922 0.595E-03 0.249E-03
RMM: 36 -0.111661849887E+03 0.61922E-07 -0.10309E-06 9508 0.378E-03 0.533E-03
RMM: 37 -0.111661850650E+03 -0.76273E-06 -0.38764E-07 6305 0.285E-03 0.223E-03
RMM: 38 -0.111661850662E+03 -0.12819E-07 -0.11357E-07 4600 0.169E-03

1 F= -.11166185E+03 E0= -.11162470E+03 d E =-.742983E-01 mag= 24.8187
========================================

when ALGO = NORMAL (DAV), it's hard to converge:
========================================
DAV: 352 -0.110068351764E+03 -0.12231E-01 -0.23508E-03 10618 0.301E-01 0.298E+00
DAV: 353 -0.110056477506E+03 0.11874E-01 -0.13494E-03 10354 0.324E-01 0.303E+00
DAV: 354 -0.110055540556E+03 0.93695E-03 -0.10729E-04 9270 0.440E-02 0.303E+00
DAV: 355 -0.110055849875E+03 -0.30932E-03 -0.72698E-06 9808 0.239E-02 0.304E+00
DAV: 356 -0.109905744653E+03 0.15011E+00 -0.11491E-01 11080 0.286E+00 0.486E+00
DAV: 357 -0.109649701444E+03 0.25604E+00 -0.15554E-01 11356 0.290E+00 0.766E+00
DAV: 358 -0.109127803586E+03 0.52190E+00 -0.27123E-01 11304 0.391E+00 0.990E+00
DAV: 359 -0.109129547948E+03 -0.17444E-02 -0.24117E-02 7738 0.883E-01 0.990E+00
DAV: 360 -0.109131592615E+03 -0.20447E-02 -0.27560E-04 10226 0.101E-01 0.989E+00
DAV: 361 -0.108781758120E+03 0.34983E+00 -0.21268E-01 11298 0.361E+00 0.944E+00
DAV: 362 -0.108708112188E+03 0.73646E-01 -0.55307E-02 10322 0.136E+00 0.990E+00
DAV: 363 -0.108744935716E+03 -0.36824E-01 -0.33922E-02 10338 0.148E+00 0.933E+00
DAV: 364 -0.109480974536E+03 -0.73604E+00 -0.70462E-01 10876 0.632E+00 0.873E+00
DAV: 365 -0.109378035793E+03 0.10294E+00 -0.56169E-02 9894 0.180E+00 0.948E+00
DAV: 366 -0.109314832773E+03 0.63203E-01 -0.88979E-03 10654 0.631E-01 0.965E+00
DAV: 367 -0.109361657955E+03 -0.46825E-01 -0.26336E-03 9184 0.444E-01 0.948E+00
372,1
======================================

(2)About Fe12Ga4 system, when lattice constant is 5.7628:
all other parameter are the same, I just change the algorithm parameter ALGO.
when ALGO = FAST (RMM), it's hard to converges:
======================================
RMM: 91 -0.111737880440E+03 0.54933E-02 -0.55702E-04 8518 0.831E-02 0.589E+00
RMM: 92 -0.111767932200E+03 -0.30052E-01 -0.23926E-03 8731 0.177E-01 0.581E+00
RMM: 93 -0.111748460263E+03 0.19472E-01 -0.18676E-03 9210 0.186E-01 0.604E+00
RMM: 94 -0.111728744659E+03 0.19716E-01 -0.36640E-03 8791 0.203E-01 0.635E+00
RMM: 95 -0.111726550395E+03 0.21943E-02 -0.38338E-04 10691 0.745E-02 0.625E+00
RMM: 96 -0.111304266079E+03 0.42228E+00 -0.30561E-01 8693 0.197E+00 0.742E+00
RMM: 97 -0.111257027332E+03 0.47239E-01 -0.54559E-02 11053 0.735E-01 0.716E+00
RMM: 98 -0.111095178796E+03 0.16185E+00 -0.52761E-02 9469 0.823E-01 0.749E+00
RMM: 99 -0.111227546409E+03 -0.13237E+00 -0.16719E-02 9661 0.605E-01 0.699E+00
RMM: 100 -0.111268650606E+03 -0.41104E-01 -0.48503E-03 9858 0.247E-01 0.686E+00
RMM: 101 -0.110972071521E+03 0.29658E+00 -0.11195E-01 8591 0.125E+00 0.777E+00
RMM: 102 -0.110994386796E+03 -0.22315E-01 -0.94400E-03 10732 0.359E-01 0.770E+00
RMM: 103 -0.111090900904E+03 -0.96514E-01 -0.11315E-02 8391 0.432E-01 0.755E+00
RMM: 104 -0.111109841076E+03 -0.18940E-01 -0.21929E-03 11098 0.140E-01 0.752E+00
RMM: 105 -0.111103710473E+03 0.61306E-02 -0.23257E-04 9284 0.685E-02 0.752E+00
RMM: 106 -0.110835535793E+03 0.26817E+00 -0.86114E-02 8760 0.107E+00 0.734E+00
RMM: 107 -0.110875045725E+03 -0.39510E-01 -0.60845E-03 10727 0.352E-01 0.743E+00
RMM: 108 -0.110879730055E+03 -0.46843E-02 -0.62198E-04 9757 0.127E-01 0.744E+00
RMM: 109 -0.110819764711E+03 0.59965E-01 -0.39986E-03 9075 0.223E-01 0.725E+00
RMM: 110 -0.110858075466E+03 -0.38311E-01 -0.33324E-03 8804 0.260E-01 0.744E+00
RMM: 111 -0.111304724653E+03 -0.44665E+00 -0.28163E-01 8561 0.188E+00 0.670E+00
======================================
when ALGO = NORMAL (DAV), it converges:
======================================
DAV: 25 -0.111680084507E+03 -0.25939E-01 -0.43845E-02 11208 0.195E+00 0.418E-01
DAV: 26 -0.111674293219E+03 0.57913E-02 -0.75801E-03 10920 0.847E-01 0.786E-01
DAV: 27 -0.111680926867E+03 -0.66336E-02 -0.11057E-02 11208 0.989E-01 0.264E-01
DAV: 28 -0.111680707788E+03 0.21908E-03 -0.87955E-04 9408 0.183E-01 0.272E-01
DAV: 29 -0.111681478436E+03 -0.77065E-03 -0.19426E-03 11856 0.422E-01 0.613E-02
DAV: 30 -0.111681537915E+03 -0.59480E-04 -0.37487E-04 10044 0.109E-01 0.256E-02
DAV: 31 -0.111681452146E+03 0.85770E-04 -0.18413E-04 11676 0.112E-01 0.851E-02
DAV: 32 -0.111681478832E+03 -0.26687E-04 -0.31170E-05 9552 0.524E-02 0.653E-02
DAV: 33 -0.111681484607E+03 -0.57743E-05 -0.27274E-05 9936 0.512E-02 0.463E-02
DAV: 34 -0.111681504745E+03 -0.20138E-04 -0.62435E-05 11244 0.718E-02 0.727E-03
DAV: 35 -0.111681504458E+03 0.28712E-06 -0.67145E-06 7932 0.143E-02 0.560E-03
DAV: 36 -0.111681504369E+03 0.88604E-07 -0.18698E-07 5880 0.321E-03

1 F= -.11168150E+03 E0= -.11164743E+03 d E =-.681412E-01 mag= 24.4970
=======================================

The lattice constant just change 0.02 Ang, and for 5.7628 Ang, DAV converges, but RMM doesn't, but for 5.7828 Ang, RMM converges, but DAV doesn't.

So please tell me where is the problem ? Thanks:)
Last edited by vasp16888 on Tue May 04, 2010 12:05 am, edited 1 time in total.
[align=center]
AB INITIO STUDY OF MAGNETIC MATERIALS
[/align]

alex
Hero Member
Hero Member
Posts: 585
Joined: Tue Nov 16, 2004 2:21 pm
License Nr.: 5-67
Location: Germany

[color=red][b]Very weird problem[/b][/color] about K points convergence check

#14 Post by alex » Tue May 04, 2010 6:46 am

I'd guess your wavefunction is a difficult case and hard to become converged. In such cases following approaches might help (one after the other):
- Start with a high-spin calculation (meaning count all up-down electrons) and try to become it converged.
- Play with the mixing tags according to the manual. There are some general rules, but you have to try several sets for optimal convergence

cheers,

alex
Last edited by alex on Tue May 04, 2010 6:46 am, edited 1 time in total.

vasp16888
Newbie
Newbie
Posts: 44
Joined: Fri Apr 23, 2010 3:09 am

[color=red][b]Very weird problem[/b][/color] about K points convergence check

#15 Post by vasp16888 » Tue May 04, 2010 10:49 pm

[quote="blue"](1)you said:[/color]
Start with a high-spin calculation (meaning count all up-down electrons) and try to become it converged.
can you explain it specifically for this Fe12Ga4 system.

(2)about mixing tags, i read the mannual where I find one of the point may be useful, and it says:
Search for the last orrurance of eigenvalues of (default mixing * dielectric matrix) in the OUTCAR file. The optimal parameters are then given by:
AMIX AMIX(as used in Pulay run)* smallest eigenvalue
AMIN=μ 2*SQRT(smallest eigenvalue/ largest eigenvalue)
and they also say try Broyden's and Pulay's mixing method, etc....
Sorry for bother you, but would you mind if giving me some tips on how to change the mixing tags. Thanks very much.
:)
Last edited by vasp16888 on Tue May 04, 2010 10:49 pm, edited 1 time in total.
[align=center]
AB INITIO STUDY OF MAGNETIC MATERIALS
[/align]

Post Reply