converge slowly

[lusommmg]

Hi,
I find that my results converge very slowly, can you give me some advice? Here is the INCAR and the results of conducting the vef.pl on one image.
INCAR:
General
SYSTEM = transition matal
PREC = Normal
ISTART = 0
ICHARG = 0
IMAGES = 8
SPRING = -5
ISYM = 0

Electronic minimization
ENCUT = 500
ENAUG = 613.6
ISIF = 3
NELM = 60
NELMIN = 4
NELMDL = -12
EDIFF = 1E-6
MAXMIX = 80
LREAL = .FALSE.
NBANDS = 16

Ionic relaxation
IALGO = 48
NSW = 400
IBRION = 3
POTIM = 0.1
EDIFFG = -0.05
BMIX = 0.8
#GGA = 91
LCLIMB = .TRUE.

DOS related parameters
ISMEAR = 1
SIGMA = 0.2

Parallel related
LPLANE = .TRUE.
NPAR = 8
LSCALU = .FALSE.
NSIM = 4

the results of conducting vef.pl on images 3:
0 0.15923600 -30.795598 0
1 0.16022800 -30.795596 2e-06
2 0.16194500 -30.795663 -6.5e-05
3 0.16343100 -30.795786 -0.000188
4 0.16378600 -30.795957 -0.000359
5 0.16372400 -30.796123 -0.000525
6 0.16332200 -30.796277 -0.000679
7 0.16303400 -30.796418 -0.00082
8 0.16265100 -30.796546 -0.000948
9 0.16234600 -30.796658 -0.00106
10 0.16204400 -30.796757 -0.001159
11 0.16177100 -30.796845 -0.001247
12 0.16160000 -30.796921 -0.001323
13 0.16119600 -30.796988 -0.00139
14 0.16106300 -30.797044 -0.001446
15 0.16078900 -30.797093 -0.001495
16 0.16047900 -30.797133 -0.001535
17 0.16029900 -30.797164 -0.001566
18 0.16009300 -30.797190 -0.001592
19 0.15992200 -30.797209 -0.001611
20 0.15975700 -30.797223 -0.001625
21 0.15959600 -30.797232 -0.001634
22 0.15945900 -30.797236 -0.001638
23 0.15961500 -30.797236 -0.001638
24 0.15982100 -30.797232 -0.001634
25 0.15994500 -30.797227 -0.001629
26 0.16007400 -30.797219 -0.001621
27 0.16018000 -30.797207 -0.001609
28 0.16026400 -30.797194 -0.001596
29 0.16033900 -30.797178 -0.00158
30 0.16040000 -30.797161 -0.001563
31 0.16045200 -30.797141 -0.001543
32 0.16049300 -30.797121 -0.001523
33 0.16052500 -30.797099 -0.001501
34 0.16054900 -30.797076 -0.001478
35 0.16057100 -30.797053 -0.001455
36 0.16058700 -30.797029 -0.001431
37 0.16059800 -30.797004 -0.001406
38 0.16060500 -30.796979 -0.001381
39 0.16059400 -30.796953 -0.001355
40 0.16059000 -30.796928 -0.00133
41 0.16058800 -30.796902 -0.001304
42 0.16058900 -30.796877 -0.001279
43 0.16058900 -30.796851 -0.001253
44 0.16059800 -30.796825 -0.001227
45 0.16059800 -30.796799 -0.001201
46 0.16058400 -30.796773 -0.001175
47 0.16057700 -30.796747 -0.001149
48 0.16057200 -30.796721 -0.001123
49 0.16056200 -30.796695 -0.001097
50 0.16054600 -30.796669 -0.001071
51 0.16053300 -30.796643 -0.001045
52 0.16051400 -30.796618 -0.00102
53 0.16049400 -30.796592 -0.000994
54 0.16047200 -30.796568 -0.00097
55 0.16045600 -30.796543 -0.000945
56 0.16043700 -30.796519 -0.000921
57 0.16041800 -30.796495 -0.000897
58 0.16037700 -30.796472 -0.000874
59 0.16073900 -30.796476 -0.000878
60 0.16033400 -30.796425 -0.000827
61 0.16023100 -30.796294 -0.000696
62 0.16014700 -30.796312 -0.000714
63 0.16012600 -30.796357 -0.000759
64 0.16009300 -30.796331 -0.000733
65 0.15976700 -30.796299 -0.000701
66 0.16000100 -30.796261 -0.000663
67 0.16009600 -30.796265 -0.000667
68 0.16017700 -30.796247 -0.000649
69 0.16018200 -30.796233 -0.000635
70 0.16012700 -30.796218 -0.00062
71 0.16014000 -30.796201 -0.000603
72 0.16014200 -30.796184 -0.000586
73 0.16012600 -30.796168 -0.00057
74 0.16010200 -30.796151 -0.000553
75 0.16008800 -30.796135 -0.000537
76 0.16006400 -30.796119 -0.000521
77 0.16004800 -30.796103 -0.000505
78 0.16002700 -30.796088 -0.00049
Thanks!

[dano]

I would suggest that you try the lbfgs optimizer.

IBRION=3
POTIM=0.0

IOPT=1
LGLOBAL=.TRUE
INVCURV=0.05

For a system with strong bonds like oxygen-oxygen you will have to reduce the initial scaling of the inverse hessian (INVCURV parameter) to a more conservative value ~0.01

You can find details about other optimizer at:
http://theory.cm.utexas.edu/vtsttools/optimizers/

[graeme]

It's also possible that there is something else going wrong with your NEB calculation, but we can't tell from the force and energy on the one image.