Relaxation of ions with EFIELD_PEAD

[jq]

Dear admin,

I would like to investigate the lattice constant dependence on the applied electric field strength. I have few questions on the calculation.

Q1:
I got the warning of "One or more components of EFIELD_PEAD is too large for comfort, in all probability you are too near to the onset of Zener tunneling."

Would my calculation be realistic even if I try to reduce my KPOINTS to 2 x 2 x 2 in order to get rid of the warning?

Q2:
How can I determine the optimized structural positions of my ions in the presence of finite electric field? From the OUTCAR file obtained, I notice the calculations have been performed following the three distinct blocks of SCF iterations mentioned in the manual.

But after the electric field is switched on, and the field-polarized groundstate is calculated (marked with DMP), the calculation continues and I got SCF calculations (marked with DAV).

Part of my OUTCAR file:
....
....
DMP: 59 -0.635178055661E+02 -0.67230E-08 -0.70928E-08 1200 0.848E-08 0.178E-07
p_tot=( -0.174E-02 -0.101E-02 0.439E+01 )
dp_tot=( 0.348E-04 0.168E-04 0.393E-01 ) diag[e(oo)]=( --- --- 6.68262 )
gam= 0.519 g(H,U,f)= 0.822E-08 0.000E+00 0.000E+00 ort(H,U,f) = 0.171E-07 0.000E+00 0.000E+00
DMP: 60 -0.635178055726E+02 -0.64920E-08 -0.68301E-08 1200 0.822E-08 0.171E-07
final diagonalization
p_tot=( -0.174E-02 -0.101E-02 0.439E+01 )
dp_tot=( 0.348E-04 0.168E-04 0.393E-01 ) diag[e(oo)]=( --- --- 6.68484 )
1 F= -.63517806E+02 E0= -.63517806E+02 d E =-.635178E+02 mag= 0.0000
curvature: 0.00 expect dE= 0.000E+00 dE for cont linesearch 0.000E+00
trial: gam= 0.00000 g(F)= 0.790E-05 g(S)= 0.820E+00 ort = 0.000E+00 (trialstep = 0.100E+01)
search vector abs. value= 0.820E+00
bond charge predicted
N E dE d eps ncg rms rms(c)
DAV: 1 -0.674328529488E+02 -0.39588E+01 -0.95265E+01 5408 0.906E+01 0.147E+01
DAV: 2 -0.635834740423E+02 0.38494E+01 -0.15891E+01 5712 0.292E+01 0.359E+00
DAV: 3 -0.635771708744E+02 0.63032E-02 -0.94236E-01 6224 0.923E+00 0.181E+00
DAV: 4 -0.635277123865E+02 0.49458E-01 -0.14491E-01 6512 0.620E+00 0.982E-01
DAV: 5 -0.635076605256E+02 0.20052E-01 -0.66978E-02 6816 0.170E+00 0.296E-01
DAV: 6 -0.635081547596E+02 -0.49423E-03 -0.98950E-03 6800 0.865E-01 0.121E-01
DAV: 7 -0.635082995377E+02 -0.14478E-03 -0.14721E-03 7328 0.378E-01 0.836E-02
DAV: 8 -0.635083564700E+02 -0.56932E-04 -0.15072E-04 7952 0.149E-01 0.497E-02
DAV: 9 -0.635084289621E+02 -0.72492E-04 -0.19125E-04 5440 0.109E-01 0.170E-02
DAV: 10 -0.635084161363E+02 0.12826E-04 -0.29849E-05 6640 0.405E-02 0.842E-03
DAV: 11 -0.635084136071E+02 0.25292E-05 -0.32202E-06 6096 0.133E-02 0.453E-03
DAV: 12 -0.635084102435E+02 0.33636E-05 -0.12822E-06 6736 0.676E-03 0.380E-03
DAV: 13 -0.635084093586E+02 0.88482E-06 -0.25165E-07 4040 0.426E-03 0.335E-03
DAV: 14 -0.635084089259E+02 0.43277E-06 -0.11042E-07 3408 0.373E-03 0.331E-03
DAV: 15 -0.635084084040E+02 0.52186E-06 -0.35245E-08 2848 0.230E-03 0.327E-03
DAV: 16 -0.635084080735E+02 0.33053E-06 -0.16956E-08 2720 0.811E-04 0.327E-03
DAV: 17 -0.635084078484E+02 0.22504E-06 -0.46525E-09 2704 0.654E-04 0.327E-03
DAV: 18 -0.635084077590E+02 0.89497E-07 -0.24220E-09 2640 0.341E-04
2 F= -.63508408E+02 E0= -.63508408E+02 d E =0.939781E-02 mag= 0.0000
trial-energy change: 0.009398 1 .order -0.067560 -0.819613 0.684492
step: 0.4685(harm= 0.5449) dis= 0.00013 next Energy= -63.701877 (dE=-0.184E+00)
bond charge predicted
....
....

I am not sure the obtained CONTCAR file will give us the optimized structure in the presence of electric field effect. Thank you.

My INCAR file contains the settings as followed:
#
PREC = normal
ENCUT = 500
EDIFF = 1E-7

# !suppress output
LWAVE = .FALSE.

# !Apply electric field
EFIELD_PEAD = 0.0 0.0 0.01

# !Relaxation
IBRION = 2
ISIF = 3
ISMEAR = 0
NSW = 100
EDIFFG = -0.005

# !For insulating system integration scheme
ISMEAR = 0
SIGMA = 0.05

# !Spin-polarized calculation
ISPIN = 2

# !Omit Spin-orbit coupling
LSORBIT = .FALSE.

# !G-type AFM
MAGMOM = 6*0 5.0 -5.0 0 0

# !Hubbard model +U calculation
LDAU = .TRUE.
LDAUTYPE = 1
LDAUL = -1 2 -1
LDAUU = 0 4 0
LDAUJ = 0 1 0
LDAUPRINT = 2
LMAXMIX = 4

[eariel]

Hi jq,
Did you find the the method to relax the ions under electric field?
I am tryng te same, adding LCALCOL=.TRUE. However VASP aborts after the self-consistency of the first step.
Eduardo