Hello everyone,
I wanted to model bonds between two atoms such that they behave like strings, i.e., there maximum separation is fixed, however, there is no force due to the bonded potential if the separation is less than the given separation.
I modelled it initially using FENE force model with parameters
K very close to zero
Ro as my distance
epsilon and sigma exactly zero
However, in this case the I got FENE bond too long.
Is there any other way to model it or to correct this FENE model.
Thanks
Sonu
Hello everyone,
I wanted to model bonds between two atoms such that they behave like
strings, i.e., there maximum separation is fixed, however, there is no
force due to the bonded potential if the separation is less than the given
separation.
I modelled it initially using FENE force model with parameters
K very close to zero
Ro as my distance
epsilon and sigma exactly zero
this doesn't make any sense. this way you have no more repulsive
interaction of the bonded atoms, so they would be able to move "through"
each other.
of course, you can model any shape of bond potential using bond style
table. however, that kind of "reverse hard sphere" interaction, that you
want to implement, is very difficult to do: a steep potential requires a
short time step and an infinitely steep potential is thus not possible to
do. a soft restoring potential on the other hand, is easily done with using
a harmonic style where the part for r < r0 is set to 0.
axel.
Actually I was using fene to model Granular sphere particles to attach to one another with thread like structure(I used atom_style hybrid sphere bonds). That is why they wouldn’t have passed through each other but would have behaved as hard spheres connect to one another. The reason I couldn’t use harmonic was because it deviates a lot from the typical granular chain behaviour.
I think I can use bond_style table to do my job if I use a force large enough after the separation greater than the length of the thread. So since the force is not infinite, I can simulate my structure using an acceptable timestep.
I think this should work. Thanks for your help.
Actually I was using fene to model Granular sphere particles to attach to
one another with thread like structure(I used atom_style hybrid sphere
bonds). That is why they wouldn't have passed through each other but would
have behaved as hard spheres connect to one another.
that is not correct in multiple ways. when you add bonds to your system,
the neighbor list code will remove the corresponding pairs from the
neighbor list, unless you adjust the special_bonds settings accordingly.
now the bond style fene will re-introduce a non-bonded repulsive term in
the bond force, however that one is not a hard sphere interaction, but a
soft lennard-jones, which in your example was disabled, because you were
setting epsilon to zero.
axel.
The reason I couldn't use harmonic was because it deviates a lot from the