Can LAMMPS/LIGGGTHS be used for simulations with residual contact behaviour and for fragmentation?

Dear all,
I need to simulate a granular system where the contacts show an initial an a residual behaviour. I mean, when the contact is first created, it behaves as a frictional and cohesive contact. But, if cohesion is broken for some reason, and the particles touch each other again, then this contact should only behave as a frictional. The cohesion does not appear when the contact is not new. This is, in just a few words, a way to simulate some pavements where the breaking of a cohesive contact is not reversible.
I would like to know of this or something similar can or cannot be done by means of LAMMPS/LIGGGHTS. I could think of an extension to the source code, in terms of a table of old contacts to select the actual contact properties, but my lack of knowledge in parallel programming avoids me to try it inside the LAMMPS/LIGGGHTS.

​I am currently starting to use LIGGGHTS 3.0.0 (but struggling with some installations problems, which are not the topic of the current email), both in Linux and Mac Os X. Thanks in advance for your guidance.​

To the best of my knowledge, this is not currently possible. The level of difficulty of allowing LAMMPS or LIGGGHTS to implement and use such a potential I cannot say. That meaning this probably confronts some details in the LAMMPS architecture.

In short, this won’t be a plug and play problem.

The Peridynamics package does something like this.

Particles are initially bonded, but when they break

the bond, it does not reform if the particles get
close together again. So in principle that concept
could be combined with granular interactions. But
you’d have to create a new pair style to do it.


Thanks. I will check it.

Another thought. There are 2 mechanisms in LAMMPS
for breaking bonds between a pair of atoms. Fix bond/break
and bond quartic - see their doc pages. Neither may
be exactly what you want, but it would be simple to

write another bond potential (e.g. harmonic) that allows
a bond to “break” based on some distance or energy criterion,
using the same logic as in bond quartic. Once the bond
is “broken” it will never re-form if the 2 atoms happen to
get close together again.