Dear LAMMPS,
I am currently trying to model a ferric chloride cell using LAMMPS. I have obtained the atomic coordinates and inter atomic potential (Born-Mayer-Huggins) from literature. However, when I simulate it, there are no bonds formed between Fe and Cl atoms. This is when I use “1.0” as the lattice constant. However, when I use “0.5” as the lattice constant, the bonds will be formed between atoms. However, then the cell will be shrinked to half of the actual size of the cell. I am not sure whether is there anything wrong with the program that I wrote. I am a beginner in LAMMPS and any help/suggestion would be really helpful and highly appreciated.
(I have attached my input file for reference)
Thank you
Yours Sincerely,
Devinda
in.fecl3.txt (2.69 KB)
Dear LAMMPS,
I am currently trying to model a ferric chloride cell using LAMMPS. I have
obtained the atomic coordinates and inter atomic potential
(Born-Mayer-Huggins) from literature. However, when I simulate it, there are
no bonds formed between Fe and Cl atoms. This is when I use “1.0” as the
lattice constant. However, when I use “0.5” as the lattice constant, the
bonds will be formed between atoms. However, then the cell will be shrinked
to half of the actual size of the cell. I am not sure whether is there
anything wrong with the program that I wrote. I am a beginner in LAMMPS and
any help/suggestion would be really helpful and highly appreciated.
how do you determine whether a "bond" is formed?
axel.
Hi Dr. Axel,
Thanks a lot for the reply. I use VMD software to view the output file.
Thanks
Best Regards,
Devinda
Hi Dr. Axel,
Thanks a lot for the reply. I use VMD software to view the output file.
...and you trust a simple program like VMD to know any chemistry???
the "bonds" that VMD is showing are the result of a simple minded
guess based on atom diameters. this works so-so for most bio
molecules, where you have hydrogens that can form only one bond
and the rest of the atoms are about the same size.
so whatever VMD will show you for your system is likely to be bogus.
axel.
Thank you for the advice sir. I will keep that in mind and proceed further. Really appreciate your help
Best Regards
Devinda
Hello Axel,
On a similar note i have a question about dump file VMD reads, I am unsure if i am posting to the wrong forum but
i need to know how the LAMMPS extension of VMD calculates the bond connectivity. I just want to import some
snapshots from dump file to tecplot and wanted to know if VMD uses a distance search for atom connectivity or
do the atom indexes in dump file give the connectivity?
Thanks,
Rohit.
Hello Axel,
On a similar note i have a question about dump file VMD reads, I am unsure
if i am posting to the wrong forum but
i need to know how the LAMMPS extension of VMD calculates the bond
connectivity. I just want to import some
snapshots from dump file to tecplot and wanted to know if VMD uses
a distance search for atom connectivity or
do the atom indexes in dump file give the connectivity?
a visualization tool like VMD can only show you
information that it knows about. for the rest it has
to either not show it or make something up (guess).
a lammps dump file does not contain any
connectivity information, so VMD uses a
distance search. even worse, since VMD
has no clue about the elements in your
dump, it will have to make them up as well.
VMD will use something better, if the file
format provides it.
what you can do, for example, is to import
a data file via the topotools plugin, guess
elements from masses and whatever else
you want to add to your system (e.g. residue
names) and the write out the result as a
.psf format file. if you then load the .psf
file first (it contains only structure information,
no coordinates) and then import the dump
file(s) on top of that, you'll have the bonding
information from your data file preserved and
also all other additional information that
can be stored in a .psf file (mass, charge, atom
type/name, residue, segment).
axel.
Thanks for the suggestions Axel. I ran the example scripts given in the topotools plugin page and understood the process
you explained. I have a question,
It says in the dump atom documentation that the indexes of the atom may change over time if the atoms migrate in between processors
Doesn’t it mean that the atom connectivity in original .psf file will not hold true to the dump file at a later stage if the indexes change?
VMD has self sufficient options when using the ‘graphic representation’ of the atoms as ‘bonds’ I can simply specify the cut-off distance of a bond
between two atoms and it shows the bonding nature (here in a bundle of SWCNTs) accurately even as the atom coordinates change.
I just wanted to know if the indexes hold good even as the positions of the atoms change so that i can just copy the connectivity list in the .psf file
or i can post process the connectivity myself if it is the better way.
Thanks,
Rohit.
Thanks for the suggestions Axel. I ran the example scripts given in the
topotools plugin page and understood the process
you explained. I have a question,
It says in the dump atom documentation that the indexes of the atom may
change over time if the atoms migrate in between processors
no. this in the dump style "local" documentation. local style
dumps have not been tested and are likely not working in VMD
Doesn't it mean that the atom connectivity in original \.psf file will
not hold true to the dump file at a later stage if the indexes change?
for an atom style dump the indices are the global atom ids
and those don't change, so the psf remains consistent.
VMD has self sufficient options when using the 'graphic representation' of
the atoms as 'bonds' I can simply specify the cut-off distance of a bond
between two atoms and it shows the bonding nature (here in a bundle of
SWCNTs) accurately even as the atom coordinates change.
the Bonds representation draws the information from the
topology/structure information that is either presented to
VMD or whatever it guesses, if it is not present. only "dynamic bonds"
does a recompute based on a (time consuming) distance
search whenever the visualization is updated, e.g. by
moving to the next frame.
I just wanted to know if the indexes hold good even as the positions of the
atoms change so that i can just copy the connectivity list in the .psf file
or i can post process the connectivity myself if it is the better way.
i don't understand what you are asking. connectivity in the
psf file is stored globally in VMD and will persist, same is
the identify of atom indices. so for a classical force field with
traditional bonded interactions, this will just work. things
get different, when using a model that allows for bonds to
be broken and formed. this requires extra scripting or
careful combining of static and dynamic representations.
axel.
Axel,
I deal with SWCNTs under mechanical loading where the chances of bond breaking are very high and bond forming very low but not impossible.I use VMD to visualize the bonding nature of the structure, I wanted to import some snapshots from an intermediate stage of the simulation (which
may contain broken bonds) from the dump file to TECPLOT. As i see now, i will need to script myself some dynamic distance search for the atoms
to get connectivity for the ‘state of atoms’ i consider (snapshot) OR better yet stick with VMD 
I think a better idea would be to extract the bonding information(connectivity) from the AIREBO potential model (used here), which i believe would require
tinkering the source code which i am not ready for, yet.
Thanks for the help!!
Rohit.