[lammps-users] neighbour list overflow

Dear all

I want to simulate the situation
that there are the extremely large particle and many small particles.
The interaction range of the large particle is the same as the system
size.
Therefore the large particle interacts with all other small particles.

When I performed the simulation, I got the error message as follows:
Neighbor list overflow, boost neigh_modify one or page.

Then, please tell me
what I should write the input script.

Sincerely yours.

Dear all

dear nakamura-san,

best greetings from sunny philadelphia.

I want to simulate the situation
that there are the extremely large particle and many small particles.
The interaction range of the large particle is the same as the system
size.
Therefore the large particle interacts with all other small particles.

When I performed the simulation, I got the error message as follows:
Neighbor list overflow, boost neigh_modify one or page.

Then, please tell me
what I should write the input script.

as far as i can tell, there is not an easy way around this from some
changes in the input script. the domain decomposition that LAMMPS uses,
does not expect particles this large. we have recently dealt with a
similar problem in our group and found the best solution is to
implement overly large particles as a "fix" and not through the regular
pair potentials.

feel free to contact me privately, if you need some help with this.
we can probably work something out.

please send my best regards to ikeshoji-san and wataru.

cheers,
   axel.

The neighbor multi command may help you do what you
want. It is designed for systems with widely varying
particle sizes and/or cutoffs. See the CPC 2008 paper
on the LAMMPS WWW site under Citations for details
of what that command does behind the scenes.

Steve