Hello,
I’m trying to run a system consisting of graphene walls and water molecules, but whenever I run the script, it shows an error of ‘‘lost atoms’’ I had tried the suggested neighbor bin and neighbor modify commands, but still it doesn’t seem to work. I had once run the system with a very small time step and observed that the the molecules fly away from the system. Here is my script file attached. Anny suggestions would be extremely helpful.
Thanks and Regards,
Deepak
gw1.in (5.84 KB)
gw.data (394 KB)
Hello,
I'm trying to run a system consisting of graphene walls and water molecules,
but whenever I run the script, it shows an error of ''lost atoms'' I had
tried the suggested neighbor bin and neighbor modify commands, but still it
doesn't seem to work. I had once run the system with a very small time step
and observed that the the molecules fly away from the system. Here is my
script file attached. Anny suggestions would be extremely helpful.
your input script is basically unreadable, but - as it is said in the
lammps manual - lost atoms happen, if you have a bad model (i.e. bad
physics), bad input parameters, or a bad starting geometry. the fact
that you observe molecules flying away, confirms that. all of these
are problems that you can solve yourself by systematically debugging
your input and building your system piece by piece and confirming at
each step, that it is working correctly. nobody here has the time to
do it for you.
axel.
For some reason, it looks like I cannot download the input file, but just as a recommendation, try to do an energy minimisation before starting the dynamics and check if you lose any atom then.
Regards,
Pablo
For some reason, it looks like I cannot download the input file, but just as
a recommendation, try to do an energy minimisation before starting the
dynamics and check if you lose any atom then.
if you feel like debugging a near unreadable input, here is the copy i
downloaded attached in a different format. the input is far too
convoluted to see immediately what is wrong, but my guess is that
there are several problems.
axel
lammpsuserslostatomserrorspcemodelsimulation.zip (83 KB)
i have to agree with you alex, that input isn’t quite readable.
anyway, deepak, if you are still with this issue, maybe you can try
crippling your input file to a bare minimum that reproduces the issue
and send it back to the list! although probably while you do that, you
might find your issue yourself.
pablo
Hello,
I have sorted it out! I had followed the method what Axel had told, i.e. to debug the script line by line and found that there was an issue with the force fields used. Thanks everyone, by the way sorry for presenting such a script.
I had one more question, I would like to delete some water molecules and create a spherical void. Can it be done using the the delete atoms and delete bond commands or are there any alternative methods to do this?
Thanks and Regards
Deepak
Hello,
I have sorted it out! I had followed the method what Axel had told, i.e. to
debug the script line by line and found that there was an issue with the
force fields used. Thanks everyone, by the way sorry for presenting such a
script.
well, there were other issues, too. being syntactically correct,
doesn't mean that those lines were good choices. what stood out the
most was the (ab)use of temp/rescale. this is a bad, bad thing to use
in an equilibration. its effect is the opposite of what you want at
this point and it just gives you a false sense of things being ok.
I had one more question, I would like to delete some water molecules and
create a spherical void. Can it be done using the the delete atoms and
delete bond commands or are there any alternative methods to do this?
yes, delete_atoms should be able to do that, but you have to look
carefully at the docs how to do it right, since you also need to
delete the associated bonds and angles and need to make sure, that you
delete complete molecules.
i personally prefer to do this in pre-processing (that is one of the
uses of VMD and topotools that i have).
mind you, it may not be the best idea to do this at the beginning and
before your system is properly equilibrated. you may actually have a
little bubble forming by itself in the process. if you want to have
control over this *and* make sure you have the void at the right
location, you could insert a spherical indenter with a tiny radius
into your system and slowly grow its radius.
that will be less drastic and will survive equilibration.
axel.