[lammps-users] numer of bonds too large in reaxFF

LAMMPS LAMMPS Mailing List Mirror
1

Hi,

I am using reaxFF in LAMMPS to perform some MD simulations. I have the following error after some picoseconds:

Too many bonds on atom. Increase MBONDDEF

I have read previous emails to the list about it. It happens when the number of bonds exceeds 20 for a given atom. However, if I integrate the g(r) of my system, I reach 20 atoms for distances never lower than 10 angstroms, which is reasonable. Also, by visual inspection I don't see any unrealistic cluster of atoms. Which is the cut off distance for the bond analysis? How could I overcome this problem?

Many thanks in advance

2

The simple answer is to edit lib/reax/reax_defs.h to set that limit
higher and re-compile the library and LAMMPS. If you're asking
why you have more than 20, I don't think it is uncommon, as others
have run into this limit issue before.

Steve