Dear users,
I have 2 kinds of atoms: atoms that are in a NVE integrator and atoms that are not. However, the atoms that are outside the integrator still interact with the atoms that are in the FIX nve.
What I want is for a certain group of atoms to behave as if they were not there (that is until the time comes to activate them)
It seems that not including them in a fix is not good enough – they still attract other atoms.
Thanks
Daniel
……………………………………………………………………………………………………………
Daniel Mulvihill
Government of Ireland Research Fellow,
Department of Mechanical, Aeronautical, and Biomedical Enginering,
University of Limerick
Dear users,
I have 2 kinds of atoms: atoms that are in a NVE integrator and atoms that
are not. However, the atoms that are outside the integrator still interact
with the atoms that are in the FIX nve.
What I want is for a certain group of atoms to behave as if they were not
there (that is until the time comes to activate them)
It seems that not including them in a fix is not good enough – they still
attract other atoms.
fixes do not *define* interactions, they modify the system
you make atoms not interact you need to either set their interaction
energy to zero (specifically for all kinds of bonded interactions). in
the case of non-bonded interactions you can also exclude them from the
neighbor list.
axel.
Axel,
Thanks for this.
So without any fixes such as nve, nvt etc. atoms can still have
temperatures, velocities and can interact with each other?
So there is already some kind of ensemble? I had thought that things
only happened when you included them in an ensemble such as NVE, NVT
etc.
Could you suggest the commands needed to set the interaction energy to
zero for a certain group of atoms? Would that group of atoms then
behave as if they weren't there? And can I later turn interactions back
on?
This misunderstanding has come about since I had thought that anything
outside of a defined ensemble (fix nve etc) played no part .
Daniel
Axel,
Thanks for this.
So without any fixes such as nve, nvt etc. atoms can still have
temperatures, velocities and can interact with each other?
do you have to repeat *every* question? that is what i just said.
So there is already some kind of ensemble? I had thought that things
only happened when you included them in an ensemble such as NVE, NVT
etc.
this has nothing to do with ensemble. further move an NVE/NVT/NPT/...
fix does *not* "generate" an ensemble.
Could you suggest the commands needed to set the interaction energy to
zero for a certain group of atoms? Would that group of atoms then
behave as if they weren't there? And can I later turn interactions back
on?
i told you what to do. if that is not sufficient, you have to go back,
read the documentation, get a better understanding of force fields and
(try to) figure it out. you cannot do proper MD simulations without
having a basic understanding of what you are doing.
This misunderstanding has come about since I had thought that anything
outside of a defined ensemble (fix nve etc) played no part .
it would help yourself and everybody on this list, if you wound stop
assuming and guessing and start improving your general understanding
of MD, force fields, statistical mechanics and what is written in the
LAMMPS documentation.
LAMMPS is a piece of software and like most of those very single
minded. it does exactly what you tell it to do; no more, no less. it
has no concept of what you *mean* it to do. assume that you are
talking to a single minded person, that only understands what you say,
if you say it in a very specific way. so you have to figure out how to
express your wishes.
axel.
Thanks Axel.
Sorry about the repetition!
Your contributions to this forum are a great benefit to learners like
me.
Daniel