non zero net momentum

Dear LAMMPS users
i am tryin to simulate oxygen adsorption on graphene, here is my input file:
dimension 3
units real
boundary p p f
atom_style charge

ffield.reax (25.2 KB)

data.mix (20.3 KB)

Dear LAMMPS users
i am tryin to simulate oxygen adsorption on graphene, here is my input file:


i run this input but i observe that graphene sheet is moving in z direction
during simulation and after some steps is reached zlo and then graphene
structure is destroyed.

i used fix momentom to prevent this happens but afew days ago Ray said :
"System drifting is a symptom indicating a problem wrong with your setup,
structure, force field, etc. I suggest you fix the problem instead of
suppress the symptom."

but until now i can not solve this problem. please give me more suggestion.i
also attach data file and ffield.reax.

first of all. the person you need to discuss this with is your adviser
and your colleagues. furthermore, you have to address this problem
differently. i am certain that your adviser will have some
suggestions. the mailing list is not a replacement for proper MD

in any case. fix momentum is not a bad thing because ray says so, but
because you should realize what is happening with a system like yours
if you enforce momentum conservation. *does* that kind of a system
have a conserved momentum?? why not?

it is very difficult to understand, why a graphene sheet cannot remain
in place, if there is no interaction from the outside and you don't
have an initial momentum. at least without a thermostat (i.e. plain
fix nve) and no oxygen atoms/molecules injected it has to remain in

but then when you insert the oxygen, you might see some effect. that
may be the result of one of three things: too few carbon atoms
relative to the oxygen, bad placement of the oxygen atoms, cutoff

if you can run the graphene by itself without a drift, then there are
easy ways to make it remain in place, e.g. use a position tether via
fix spring, or do not time integrate one of the carbon atoms.