### Force convergence problem

Posted:

**Mon Nov 05, 2018 2:38 am**Dear Admin,

My CI-NEB calculations are not converging. I am trying to moving one H atom on the SiC surface into the subsurface. I used 7 images in the process and with one Gamma point calculation.The optimizer I used is IOPT=2. But after 1000 steps, the NEB is still not converged. It shows that the forces on each image is still very large, even more than 1 eV/A. Here is my INCAR:

general: geometry optimation-Calculation

Start Parameters for this run

ISTART = 0

ICHARG = 2

INIWAV = 1

PREC = Accurate

ISMEAR = 0

SIGMA = 0.05

ALGO = Fast

NSIM = 4

LPLANE = .TRUE.

NPAR = 4

ISYM = 0

Electronic minimisation

ENCUT = 500

NELM = 100

NELMIN = 4

LREAL = A

EDIFF = 1E-5

LWAVE = .FALSE.

LCHARG = .FALSE.

Ionic relaxation

ISIF = 2

NSW = 1000

EDIFFG = -0.01

ISPIN = 2

LVDW = .TRUE.

IBRION = 3

POTIM = 0

ICHAIN = 0

IMAGES = 7

SPRING = -5

LCLIMB = .TRUE.

IOPT = 2

-----------------------------------

I also tried to use IOPT=3, with TIMESTEP = 0.1, it is still not converged.

Thank you very much for your help!

My CI-NEB calculations are not converging. I am trying to moving one H atom on the SiC surface into the subsurface. I used 7 images in the process and with one Gamma point calculation.The optimizer I used is IOPT=2. But after 1000 steps, the NEB is still not converged. It shows that the forces on each image is still very large, even more than 1 eV/A. Here is my INCAR:

general: geometry optimation-Calculation

Start Parameters for this run

ISTART = 0

ICHARG = 2

INIWAV = 1

PREC = Accurate

ISMEAR = 0

SIGMA = 0.05

ALGO = Fast

NSIM = 4

LPLANE = .TRUE.

NPAR = 4

ISYM = 0

Electronic minimisation

ENCUT = 500

NELM = 100

NELMIN = 4

LREAL = A

EDIFF = 1E-5

LWAVE = .FALSE.

LCHARG = .FALSE.

Ionic relaxation

ISIF = 2

NSW = 1000

EDIFFG = -0.01

ISPIN = 2

LVDW = .TRUE.

IBRION = 3

POTIM = 0

ICHAIN = 0

IMAGES = 7

SPRING = -5

LCLIMB = .TRUE.

IOPT = 2

-----------------------------------

I also tried to use IOPT=3, with TIMESTEP = 0.1, it is still not converged.

Thank you very much for your help!