errors in LOCPOT?

Problems running VASP: crashes, internal errors, "wrong" results.


Moderators: Global Moderator, Moderator

Post Reply
Message
Author
t.ossowski
Newbie
Newbie
Posts: 6
Joined: Mon Nov 09, 2009 4:08 pm

errors in LOCPOT?

#1 Post by t.ossowski » Wed Nov 13, 2013 10:35 am

I have problems with LOCPOT files. I'm trying to find work function of magnetic iron surfaces. I performed calculations (using VASP 5.2.12) and used vtotav.f program to find average potential along Z-direction. It looks very strange! When I look at VLINE average potential along Z-direction has oscillations and strange steps near vacuum level. When I look into LOCPOT found that there is more points than NGX*NGY*NGZ and there is additional part with something what looks like additional potential??? My LOCPOT looks like follows (Of course I present only part of them):

####################################################################################
Fe_110
2.832601555000000
0.707107 -0.500000 0.000000
0.707107 0.500000 0.000000
0.000000 0.000000 8.485281
Fe
5
Direct
0.000000 0.000000 0.082978
0.500000 0.500000 0.166183
0.000000 0.000000 0.250000
0.500000 0.500000 0.333817
0.000000 0.000000 0.417022

24 24 240
0.36647110705E+00 0.56175002151E+00 0.69477258713E+00 0.96350469192E+00 -.88985603364E+00
-.82476657706E+00 0.12726712759E+01 -.48267885063E+00 0.57189547349E+00 0.89569886989E+00
............. many lines with numbers ................
0.20966385291E+00 0.16888388758E+01 0.35356464206E+00 -.60458038520E+00 -.50725877579E+00
0.400000000000E+01 0.400000000000E+01 0.400000000000E+01 0.400000000000E+01 0.400000000000E+01
24 24 240
0.15698279108E+01 0.14325949678E+01 0.13394073541E+01 0.10962926365E+01 0.33848922491E+00
0.53608218933E+00 0.73773469491E+00 0.38997511493E+00 0.69606846664E+00 0.65481543515E+00
............many lines with numbers..................
0.96059896910E+00 0.89514435109E+00 0.10577636853E+01 0.10238703994E+01 0.96383384907E+00
0.93463656573E+00 0.14370606306E+01 0.10765162897E+01 0.78265517930E+00 0.58589100602E+00
#####################################################################################

First problem is that in first part of LOCPOT the number of values is larger than 24*24*240 (additional line with 0.400000 ....).
Second problem: what is in this additional part after this line with "0.400000000000E+01 0.400000000000E+01 0.400000000000E+01 0.400000000000E+01 0.400000000000E+01"? I thought that in LOCPOT is only electrostatic potential? Could someone tell me what is going on? Maybe I did something wrong or my VASP compilation is wrong? I did caclulations for cells with more atomic layers and for much larger vacuum and the problem is the same.

Best regards
Tomasz
Last edited by t.ossowski on Wed Nov 13, 2013 10:35 am, edited 1 time in total.

t.ossowski
Newbie
Newbie
Posts: 6
Joined: Mon Nov 09, 2009 4:08 pm

errors in LOCPOT?

#2 Post by t.ossowski » Fri Nov 15, 2013 12:10 pm

I did calculations for more dense (2 times) FFT grid and now there are no oscillations but still there is strange shape of average potential at the boundary between sureface and vacuum :( I used PAW PBE potential (I have not metioned this before).

Of course still I have qeustion why in LOCPOT there are two parts and why there is additional line at the end of the first part?
Last edited by t.ossowski on Fri Nov 15, 2013 12:10 pm, edited 1 time in total.

t.ossowski
Newbie
Newbie
Posts: 6
Joined: Mon Nov 09, 2009 4:08 pm

errors in LOCPOT?

#3 Post by t.ossowski » Mon Nov 25, 2013 12:16 pm

I did calculations using vasp 4.6.36 and the same files as for 5.2.12. VLINE has nice shape - no oscillations, jumps and discontinuities as it is in VASP 5. In LOCPOT from vasp4 the same as from VASP 5 there are two parts (I really do not know why?) but it looks that both parts in VASP 4 are the same (except one additional linie in first part) whereas for VASP 5 they differ.
Last edited by t.ossowski on Mon Nov 25, 2013 12:16 pm, edited 1 time in total.

Post Reply