# FENE bond cutoff length

Dear LAMMPS users,

What is the cutoff distance for bond_style fene, and how to calculate/define it?

The fene potential is written as

E = -0.5 * K * ln (1- (r/R0)^2) (LJ term eliminated)

Isn’t R0 the cutoff length? I have R0 set as 1.5 in my simulation, and distance between bonded atoms set as ~ 1.0 (distributed, but definitely below 1.5). However, while running a simulation, it always complains that the FENE bonds between some certain atoms are too long. Should I set the initial distance even smaller?

Plus, I’ve told by senior people in my group that the cutoff length for FENE is calculated as 0.97, however they were not clear how it was calculated… Later on I searched some literatures that saying the cutoff length for FENE is around 0.97. My questions is: shouldn’t the cutoff length vary along with the preset value of R0?

Thank you very much in advance!

Sunnia

Dear LAMMPS users,

What is the cutoff distance for bond_style fene, and how to calculate/define
it?

what kind of cutoff are is this in reference to? bonds don't have an
explicit cutoff in LAMMPS.

The fene potential is written as

E = -0.5 * K * ln (1- (r/R0)^2) (LJ term eliminated)

Isn't R0 the cutoff length? I have R0 set as 1.5 in my simulation, and

r0 is the length where the potential (obviously) diverges, i.e. the
maximum allowed bond length.
the only cutoff used is the cutoff in the minimum of the LJ potentials
at 2**(1/6) sigma.

distance between bonded atoms set as ~ 1.0 (distributed, but definitely
below 1.5). However, while running a simulation, it always complains that
the FENE bonds between some certain atoms are too long. Should I set the
initial distance even smaller?

there are all kinds of reasons for getting warnings about too long
bonds. it could be a large time step, too high temperature,
unreasonable potential parameters or masses and so on...

Plus, I've told by senior people in my group that the cutoff length for FENE
is calculated as 0.97, however they were not clear how it was calculated...
Later on I searched some literatures that saying the cutoff length for FENE
is around 0.97. My questions is: shouldn't the cutoff length vary along with
the preset value of R0?

you have to be more specific what kind of cutoff you are inquiring about.

axel.

As I recall ~0.97 is the equilibrium bond length
for FENE with R0 = 1.5. R0 is the maximum
stretch of the bond, where the energy equation
goes to infinity.

Steve

I have this question (FENE cutoff) mainly for two reasons:

1. as the simulation goes on, it crashes after a while reporting “FENE bond too long”. However, I thought it might be due to the initial configuration so I set the distance between bonded atoms as ~ 0.97 hoping it will help, but the same problem came again. Now I understand it might be the dynamics that has been set improperly.

2. I am using ‘fix bond/create’ command to create another FENE bond between certain types of atoms when they meet a specific criteria, but was able to put it into use. The programme always complains about the invalid cutoff set in fix bond/create commands, which I assume refers to the cutoff distance of FENE. Isn’t it? If not, what’s the criteria of setting R_min in fix bond/create command, so that the fene bond can be created correctly?

Best,

Sunnia

Thanks Steve! - I later on find some answers in archives that the maximum bond length of FENE is 1.5sigma and equilibrium as 0.97sigma for a default setting (R_0=1.5). So the FENE bond is bad when it is > 1.5 or >0.97? My understanding is the bond can be established when the distance between bonded atoms are <1.5, but it will reach 0.97 eventually. Is that correct?

Also, if I increase the R_0 in FENE, for example, set R_0 as 3, will the maximum distance become 3.0sigma and equilibrium become 0.973.0/1.5*sigma (basically be proportionally increased)?

Many thanks.

Sunnia

I have this question (FENE cutoff) mainly for two reasons:

1) as the simulation goes on, it crashes after a while reporting "FENE bond
too long". However, I thought it might be due to the initial configuration
so I set the distance between bonded atoms as ~ 0.97 hoping it will help,
but the same problem came again. Now I understand it might be the dynamics
that has been set improperly.

2) I am using 'fix bond/create' command to create another FENE bond between
certain types of atoms when they meet a specific criteria, but was able to
put it into use. The programme always complains about the invalid cutoff set
in fix bond/create commands, which I assume refers to the cutoff distance of

no it doesn't. for fix bond/create you look at atoms that are not yet
bonded by traversing the neighbor list for suitable pairs of atoms.
the

FENE. Isn't it? If not, what's the criteria of setting R_min in fix

the condition is given in the fix bond/create documentation:

Rmin must be <= the pairwise cutoff distance between itype and jtype
atoms, as defined by the pair_style command.

bond/create command, so that the fene bond can be created correctly?

axel.

Thanks Steve! - I later on find some answers in archives that the maximum
bond length of FENE is 1.5*sigma and equilibrium as 0.97*sigma for a default
setting (R_0=1.5). So the FENE bond is bad when it is > 1.5 or >0.97? My

= 1.5

understanding is the bond can be established when the distance between
bonded atoms are <1.5, but it will reach 0.97 eventually. Is that correct?

the FENE potential will be applied, whenever a bond between two atoms
is defined, always.

Also, if I increase the R_0 in FENE, for example, set R_0 as 3, will the
maximum distance become 3.0*sigma and equilibrium become 0.97*3.0/1.5*sigma
(basically be proportionally increased)?

your question is not clearly defined. R_0 is given in distance units,
which may be 1 sigma when using reduced units, but that is independent
form the LJ sigma used in the FENE potential. the average or typical
bond length may grow a little due to the shift in the R_0, but it will
still be primarily determined by the LJ repulsion. please note, that
this is also affected by temperature.

axel.