Setting up run for a very long time and finally get bad termination

Dear everyone,

I want to build graphite and run it into equilibrium. I saw some experiences shared by the others and used VMD to generate graphite. In VMD, the 'Extension->Modeling->Nanotube Builder' is used here. I unchecked 'Bonds', so there is no bonds, angles, dihedrals etc.. The 'Edge length along x' and along y are both 10nm, and the 'Number of layers' is 10 layers. After click 'Generate Sheets', I use 'topo writelammpsdata graphite.data' to get the lammps data file.

Here is the heads of graphite.data:

LAMMPS data file. CGCMM style. atom_style full generated by VMD/TopoTools v1.5 on Tue Jan 02 10:51:53 CST 2018
39360 atoms
0 bonds
0 angles
0 dihedrals
0 impropers
1 atom types
0 bond types
0 angle types
0 dihedral types
0 improper types
-0.500000 0.500000 xlo xhi
-0.500000 0.500000 ylo yhi
-0.500000 0.500000 zlo zhi

# Pair Coeffs

in.graphite.equil (541 Bytes)

out.graphite.equil (1.16 KB)

Dear everyone,

I want to build graphite and run it into equilibrium. I saw some
experiences shared by the others and used VMD to generate graphite. In VMD,
the 'Extension->Modeling->Nanotube Builder' is used here. I unchecked
'Bonds', so there is no bonds, angles, dihedrals etc.. The 'Edge length
along x' and along y are both 10nm, and the 'Number of layers' is 10
layers. After click 'Generate Sheets', I use 'topo writelammpsdata
graphite.data' to get the lammps data file.

​but you didn't set a suitable simulation cell dimension. just look at the
header part of your data file and your output.
you have defined a 1x1x1 angstrom box. if you squeeze ~40000 atoms in such
a small volume, bad things are bound to happen.

also you are using an outdated version of topotools and an outdated version
of LAMMPS.​

​axel.​

Dear Axel,

 Thanks very much for your reply\! I have just moved to VMD version 1\.9\.3 instead\. It seems I got nearly the same data file with VMD/TopoTools v1\.7\.

 I did not find any other options in 'Nanotube Builder' to define the size of the simulation box\. I set the length of x and y directions to be 10nm and the number of layers to be 10\. I am sorry I did not find further instructions on the help page of Nanotube Builder \(http://www.ks.uiuc.edu/Research/vmd/plugins/nanotube/). I thought the size of the simulation box is set automatically by the plugin and it can be used in lammps directly before\.

 Could you please give me some instructions on how to modify the data file?  Maybe I do not use the 'Nanotube Builder' properly\. Or I should modify the data file manually?

 Thanks in advance\!

With My Best Regards,

Liu

Dear Axel,

    Thanks very much for your reply! I have just moved to VMD version
1.9.3 instead. It seems I got nearly the same data file with VMD/TopoTools
v1.7.

    I did not find any other options in 'Nanotube Builder' to define the
size of the simulation box. I set the length of x and y directions to be
10nm and the number of layers to be 10. I am sorry I did not find further
instructions on the help page of Nanotube Builder (
http://www.ks.uiuc.edu/Research/vmd/plugins/nanotube/). I thought the
size of the simulation box is set automatically by the plugin and it can be
used in lammps directly before.

​no, it doesn't set the box. you can use "pbc set" from the pbctools
plugin for it. or just use VMD Tcl script commands. there is no simple way
for this anyway, also please note, that the nanotube plugin does produce
isolated and not proper periodic geometries.

    Could you please give me some instructions on how to modify the data
file? Maybe I do not use the 'Nanotube Builder' properly. Or I should
modify the data file manually?

​you can use VMD scripting or you could manually edit the data file. either
should work.​

​axel.​

Dear Axel,

Thanks very much! I followed your instructions and tried to check the pbctools or VMD script, I found that it is a bit difficult for me to use them right now because I need some time to read and understand the manual. So firstly I tried to modify my data file manually.

As you pointed out in previous reply, the main problem is the size of the simulation cell is not correct. So I tried to modify it, since the size of graphite I choose in the nanotube plugin is 10nmx10nm and 10 layers, I changed the three lines in data file to the following lines and kept the other part unchanged.

0.00000 100.0000 xlo xhi
0.00000 100.0000 ylo yhi
0.00000 34.00000 zlo zhi

However, when I run it in LAMMPS with minimize command, the carbon atoms rearranged at all the boundaries, the result seems unreasonable. One more thing is that, I found that the distance between two neighboring layers is 3.35, however in some articles I found the value is 0.34nm. Could you please tell me why? BTW, I also tried ‘0.00000 33.50000 zlo zhi’, the results are similar, and the total layers even changed to 11. The initial and final state of the atoms can be found in the attachments.

Could you please give me some more detailed instructions on how to modify my data file? Do I just need to get the right box size or some more actions are needed?

With My Best Regards,

Liu

initial.png

final.png

final335.png

Dear Axel,

    Thanks very much! I followed your instructions and tried to check the
pbctools or VMD script, I found that it is a bit difficult for me to use
them right now because I need some time to read and understand the manual.
So firstly I tried to modify my data file manually.

    As you pointed out in previous reply, the main problem is the size of
the simulation cell is not correct. So I tried to modify it, since the size
of graphite I choose in the nanotube plugin is 10nmx10nm and 10 layers, I
changed the three lines in data file to the following lines and kept the
other part unchanged.

     0.00000 100.0000 xlo xhi
     0.00000 100.0000 ylo yhi
     0.00000 34.00000 zlo zhi

    However, when I run it in LAMMPS with minimize command, the carbon
atoms rearranged at all the boundaries, the result seems unreasonable. One
more thing is that, I found that the distance between two neighboring
layers is 3.35, however in some articles I found the value is 0.34nm. Could
you please tell me why? BTW, I also tried '0.00000 33.50000 zlo zhi', the
results are similar, and the total layers even changed to 11. The initial
and final state of the atoms can be found in the attachments.

    Could you please give me some more detailed instructions on how to
modify my data file? Do I just need to get the right box size or some more
actions are needed?

please note, that ​all of these are *not* specific questions about LAMMPS
(or VMD), but about how to do research and set up simulations. that makes
it a topic for a discussion with your adviser or tutor or senior colleagues
and off-topic for this mailing list.

i have alluded to it before, but perhaps need to point out again, that the
VMD nanotube​ plugin is *not* written to produce a periodic system with
proper matching periodic replication. what you get is a stack of graphene
sheets, which is similar to, but not a proper graphite crystal (and neither
does it promise to build a graphite crystal cell).

​axel.​

Dear Axel,

Thanks very much for all your helps! I will try to get help from maybe other forums which is suitable for this topic.

With My Best Regards,

Liu