Dear lammps users
I have a question about the fix wall/harmonic command after read the manual.
I have my own wall which consist of two layers atoms and it is defined as one group in my input file. Manual says “the fix wall/harmonic command will bound the simulation domain on one or more of its faces with a flat wall that interacts with atoms in the defined group by generating forces on atoms in the direction perpendicular to the wall”.
My first question is what does the “wall” in manual mean? My understanding is it is just some ghost atoms with no mass and it will interact with real wall atoms with mass defined by users. The function of this command is just to fix the real user defined wall at ,for instance, FCC crystal surface. Is that right?
If I use the command in my system, the two layer atoms in the group will act the with “wall” with a spring separately (two lays are within the cuttoff distance) in the direction perpendicular to the “wall”. But there should be no interaction between the atoms in the group. If I want to assign the L-J potential to the atoms in the group. I still should use pair_style and pair_coeff command.
Is my understanding right?
Fan Li
Dear lammps users
I have a question about the fix wall/harmonic command after read the
manual.
I have my own wall which consist of two layers atoms and it is defined as
one group in my input file. Manual says "the fix wall/harmonic command will
bound the simulation domain on one or more of its faces with a flat wall
that interacts with atoms in the defined group by generating forces on atoms
in the direction perpendicular to the wall".
My first question is what does the "wall" in manual mean? My understanding
is it is just some ghost atoms with no mass and it will interact with real
wall atoms with mass defined by users. The function of this command is just
to fix the real user defined wall at ,for instance, FCC crystal surface. Is
that right?
no. fix wall/harmonic is simply a one-sided repulsive harmonic potential.
if you have a slab system with the non-periodic dimension in Z and place
a zhi wall/harmonic at z=+10 then any atom with z > 10.0 will
experience a a force in z pushing it back to z < 10.0 with U=k *
(z-10.0)
If I use the command in my system, the two layer atoms in the group will act
the with "wall" with a spring separately (two lays are within the cuttoff
distance) in the direction perpendicular to the "wall". But there should be
no interaction between the atoms in the group. If I want to assign the L-J
potential to the atoms in the group. I still should use pair_style and
pair_coeff command.
Is my understanding right?
if you build a wall from explicit atoms, you don't need fix wall/*
the purpose of fix wall either to keep atoms from leaving a box or to
have a less detailed, but much more computational efficient
interaction than a wall with explicit particles.
a lj96 wall can be derived by integrating over a semi-infinite block
of 12-6 lj particles.
axel.
Hi Axel
Thanks. I understand.
Fan Li
------------------ Original ------------------