Dear Axel,
Thank you very much for your help and suggestions. Actually, i need to do
that for a project and i have to learn it.
then go to the person that assigned that project to you and ask that
person to be properly trained. this mailing list cannot do this for
you. ...and also let me alert you, that there is more to learn about
MD and how to do it correctly, than just writing a syntactically
correct input. you cannot learn this while doing your project. since
your project seems to be complex, it is *necessary* to first train
with simpler system. failing to do so, *will* take more time since you
will be making bad choices and will thus will have to repeat
calculations or will get bad and unpublishable results.
With your help, i changed my input to :
atom_style atomic
read_data ./data.out extra/atom/types 1
read_data ./SphereUnit.xyz add append offset 2 2 2 2 2
pair_style hybrid lj/cut 10 tersoff tersoff
pair_coeff * * tersoff 1 BN.tersoff B N NULL
pair_coeff * * tersoff 2 Si.tersoff Si NULL NULL
pair_coeff 1 3 lj/cut 0.8 1.3
pair_coeff 2 3 lj/cut 0.8 1.3
pair_coeff 3 3 none
I think it's much better now. I just have a question, how can i calculate
it is still wrong. you are now overwriting the B parameters with Si parameters.
there should be no pair_coeff none setting for this kind of system.
the value of lj/cut? the values that i wrote (0.8 and 1.3) are just for
test, i could not find anything on lammps web site that help me to how to
calculate the value of lj/cut, from which formula?
they are not computed, they need to be parameterized. you need to look
them up in the published literature, if they exist, or learn how to
parameterize those yourself (which is a *massive* undertaking for a
beginner) or collaborate with somebody that knows how to parameterize
interactions like these. this question and your inability to
understand that your input is incorrect, just confirms my assessment,
that you are far too inexperienced to do a simulation of such a
complex system. please go to your adviser/supervisor and request to be
properly trained.
your issues are primarily not LAMMPS issues, but problems of not
understanding MD and classical force fields well enough altogether.
please understand, that it is not the purpose of this mailing list or
the LAMMPS manual to provide this kind of understanding, but rather it
is a prerequisite to be able to make sense of the kind of (technical)
help we can provide and understand the choices required for writing
LAMMPS simulation input scripts.
Thank you very much, and i am so sorry if my questions are so basic to you.
as mentioned before, you are mostly harming yourself. if you want be
successful at performing good and meaningful MD simulations, you will
need to know more about the basics of MD simulations. no amount of
apologies will save you from that.
it is ok to state that you are a beginner and ask for help. however,
it is not going to be OK (at least not with me) if you keep continuing
with your project without following the advice being given and first
obtaining better training in simulation and statistical mechanics
basics. you'll be wasting your and - more importantly - our time.
axel.