UT theoretical chemistry code forum

neb.dat:
0 0.000000 0.000000 0.007661 0
1 0.595791 0.054466 -0.383753 1
2 1.215026 0.278019 -0.016640 2
3 3.671896 0.281341 -1.341445 3
4 6.172996 0.194630 -0.122926 4
5 8.459793 -0.242267 -0.004356 5

output of nebef.pl

0 0.00000000 -649.51444400 0.00000000
1 0.04869700 -649.45997800 0.05446600
2 6.72448500 -649.23642500 0.27801900
3 0.92719500 -649.23310300 0.28134100
4 0.77632500 -649.31981400 0.19463000
5 0.00000000 -649.75671100 -0.24226700

I have converged the band with optimizer QM to EDIFFG=-0.3 ， and then switch to CG optimizer.

All related files were attached.

Can some one give me advice on my calculation?

My INCAR:

ISTART = 1
ICHARG = 1
ENCUT = 480
EDIFF = 1E-4
NELM = 10
EDIFFG = -0.05
#NPAR = 2
NSW = 50
IBRION = 3
#LORBIT = 10
NELMIN = 3
#NFREE =10
LREAL= Auto
ISMEAR = 0
LMAXMIX = 6
SIGMA = 0.2
ALGO = F
POTIM = 0 ! to test the POTIM tag in the OUTCAR
#MAXMIX = 50 ! a test for MAXMIX parameter. the default valume is -45.
ISPIN = 2
LDAU = .True.
LDAUL = -1 -1 3
LDAUU = 0 0 5
VOSKOWN = 1
#MAGMOM = 1*3 54*0 26*0
ICHAIN = 0
IOPT = 2
MAXMOVE = 0.05
#TIMESTEP = 0.04
FDSTEP = 2E-3
LCLIMB = .TRUE.
IMAGES = 4
SPRING = -5
#SMASS = 2

The same question i also always have. The IOPT=3 seems conservative and can decrease the force stepwisely. Once the aggressive optimizers are used (such as IOPT=1), the forces didnot drcrease any more and sometimes become very large.

That may be due to the interval between images are too large.