Page 1 of 1

DIMER question

Posted: Mon Jan 12, 2009 7:45 am
by barryhyc
Hi, everyone
I am using VASP with DIMER method. I encounter some difficulities which I cannot solve it by myself. So I would
be very appreciated if someone tell me how to do.

I start a DIMER task with a MODECAR which produced by modemake.pl script. But unfortunately, it cannot converged
within 60 electronic steps. Up to now, there are 16 ions steps all not converged. So I try to do a DIMER without
a MODECAR, and also let NELM=100, but after 100 electronic steps, VASP exits with the following error:

forrtl: No such file or directory
forrtl: No such file or directory
forrtl: severe (29): file not found, unit 210, file /localhome/huang/221/221pd/dehydrogenation/dimer-2-layer-1/MODECAR
Image PC Routine Line Source
vasp 00000000010C6762 Unknown Unknown Unknown
vasp 00000000010C5962 Unknown Unknown Unknown
vasp 000000000108E4DE Unknown Unknown Unknown
vasp 000000000104E8E9 Unknown Unknown Unknown
vasp 000000000104E1D7 Unknown Unknown Unknown
vasp 000000000105D625 Unknown Unknown Unknown
vasp 0000000000562811 Unknown Unknown Unknown
vasp 000000000059DF70 Unknown Unknown Unknown
vasp 0000000000438F81 Unknown Unknown Unknown
vasp 0000000000412BE2 Unknown Unknown Unknown
libc.so.6 000000351D71C4BB Unknown forrtl: severe (29): file not found, unit 210, file /localhome/huang/221/221pd/dehydrogenation/dimer-2-layer-1/MODECAR
Image PC Routine Line Source
vasp 00000000010C6762 Unknown Unknown Unknown
vasp 00000000010C5962 Unknown Unknown Unknown
vasp 000000000108E4DE Unknown Unknown Unknown
vasp 000000000104E8E9 Unknown Unknown Unknown
vasp 000000000104E1D7 Unknown Unknown Unknown
vasp 000000000105D625 Unknown Unknown Unknown
vasp 0000000000562811 Unknown Unknown Unknown
vasp 000000000059DF70 Unknown Unknown Unknown
vasp 0000000000438F81 Unknown Unknown Unknown
vasp 0000000000412BE2 Unknown Unknown Unknown
libc.so.6 000000351D71C4BB Unknown forrtl: severe (29): file not found, unit 210, file /localhome/huang/221/221pd/dehydrogenation/dimer-2-layer-1/MODECAR
Image PC Routine Line Source
vasp 00000000010C6762 Unknown Unknown Unknown
vasp 00000000010C5962 Unknown Unknown Unknown
vasp 000000000108E4DE Unknown Unknown Unknown
vasp 000000000104E8E9 Unknown Unknown Unknown
vasp 000000000104E1D7 Unknown Unknown Unknown
vasp 000000000105D625 Unknown Unknown Unknown
vasp 0000000000562811 Unknown Unknown Unknown
vasp 000000000059DF70 Unknown Unknown Unknown
vasp 0000000000438F81 Unknown Unknown Unknown
vasp 0000000000412BE2 Unknown Unknown Unknown
libc.so.6 000000351D71C4BB Unknown Unknown Unknown
vasp 0000000000412B2A Unknown Unknown Unknown
Unknown Unknown
vasp 0000000000412B2A Unknown Unknown Unknown
Unknown Unknown
vasp 0000000000412B2A Unknown Unknown Unknown

[2]+ Exit 137 ./bat (wd: /localhome/huang/221/221pd/dehydrogenation)
(wd now: /localhome/huang/221/221pd/dehydrogenation/dimer-2-layer-1)

So my questions are:
1. How to use DIMER method effectively? It must continue after regular NEB?
2. Why I met this error, is the compilation question?

Thanks, I am looking forward to your reply and discussion.

Re: DIMER question

Posted: Wed Jan 14, 2009 5:12 am
by graeme
The electronic convergence problem is probably not due to the dimer at all. Try a regular vasp calculation using the same POSCAR file, and you should get the same output for the first electronic steps. There are several reasons that can cause the electronic structure to converge poorly and some solutions in the vasp manual. Try looking for information about the AMIX and BMIX tags, MAGMOM if your system is spin polarized, or IALGO to try a different electronic optimizer.

The dimer subroutine is not entered until after the first force/energy evaluation is run, which is why it should not effect the initial electronic structure convergence.

You must have a modecar to run the dimer.

The dimer can be used to finely converge a single image to a saddle from an NEB calculation. It can also be used to search for saddles from minima, or any place on the potential surface. The former is a more straight-forward application. The latter is more expensive and can lead to irrelevant saddles, but it can also be used to find unexpected reaction mechanisms.

Re: DIMER question

Posted: Wed Jan 14, 2009 6:54 am
by barryhyc
Thany you for your reply. I try to use example of H diffusion in a frozen Ir lattice (downloaded from
website http://theory.cm.utexas.edu/vtsttools/) as a benchmark, but I cannot get the same result.
It also donnot converge after 60 electronic steps, so I suspect whether it is the compilation probelm.
This is my DIMCAR:
Step Force Torque Energy Curvature Angle
1 6.00303 62.54346 -43.65348 -4.35175 51.96997
1 6.00303 29.66141 -43.65348 -30.30397 28.72234
1 6.00303 43.47671 -43.65348 -29.62354 22.03665
1 6.00303 43.95336 -43.65348 -36.35973 32.77807
2 5.96294 62.10673 -42.93899 7.41361 47.81462
2 5.96294 41.27629 -42.93899 -0.21401 50.02325
2 5.96294 52.04768 -42.93899 -15.22481 63.97874
2 5.96294 30.37417 -42.93899 -45.71445 47.09034
3 9.29362 50.33743 -42.11095 -39.44532 39.66616
3 9.29362 49.35017 -42.11095 -46.38845 15.85466
3 9.29362 37.96595 -42.11095 -19.45687 16.87820
3 9.29362 63.88947 -42.11095 -9.39955 29.57697
4 6.63616 55.00641 -42.72418 -13.63088 49.31484
4 6.63616 47.34260 -42.72418 -20.79042 53.31710
4 6.63616 44.61408 -42.72418 -7.00383 36.38111
4 6.63616 52.42354 -42.72418 -29.27678 40.73667
5 6.12472 58.14640 -42.90530 16.37393 49.95035
5 6.12472 27.44705 -42.90530 -39.67825 45.06249
5 6.12472 28.63904 -42.90530 -46.47566 43.13384
5 6.12472 37.36798 -42.90530 -35.21709 28.73560
6 8.41172 85.36661 -42.25029 54.46575 51.26931
6 8.41172 53.50163 -42.25029 9.50417 51.54881
6 8.41172 39.19520 -42.25029 -17.23088 55.37392
6 8.41172 15.99630 -42.25029 -2.68529 37.49634
7 7.06283 82.18965 -42.79868 -46.79451 43.00328
7 7.06283 19.18125 -42.79868 -43.55311 10.04628
7 7.06283 25.02933 -42.79868 -65.14327 18.95049
7 7.06283 25.48192 -42.79868 -38.12766 8.81324
8 5.14574 24.66614 -43.19446 34.59774 53.52603
8 5.14574 41.57356 -43.19446 -26.00020 45.50149
8 5.14574 18.81300 -43.19446 -42.94956 44.32579
( I killed it)
This is DIMCAR from your website
Step Force Torque Energy Curvature Angle
1 3.26055 1.42551 -38.08043 -3.67928 7.32459
1 3.26055 0.47200 -38.08043 -3.89665 0.48560
2 0.65134 0.32662 -38.17792 -2.88463 1.05846
3 0.18678 0.22588 -38.19421 -2.88144 0.23646
4 0.20597 0.11406 -38.19315 -2.82977 0.14029
5 0.02071 0.58961 -38.19131 -3.08022 0.61217
6 0.01329 0.06599 -38.19131 -3.05717 0.14982
7 0.00990 0.11718 -38.19130 -3.07347 0.26844
8 0.00208 0.11183 -38.19130 -3.04074 0.14479

I cannot understand what you say "You must have a modecar to run the dimer". The manual of DIMER say " It is
strongly recommended that you generate a MODECAR file". This means we also can not offer MODECAR, used a random
direction. So I am puzzled about the DIMER.

So what I shall do? Please give me some advice. Thank you very much.