I am trying to run a simulation with nanotubes and argon molecules. When i had only one nanotube ,rigid,my simulation was workin fine.I created a bundle and it starts running and at some poitn a receive the error:
ERROR on proc 0: Too many atom sorting bins
ERROR on proc 8: Too many atom sorting bins
ERROR on proc 16: Too many atom sorting bins etc.
I changed from bin to nsq but the error remains the same.
This error likely comes from the simulation box becoming huge. That
is bad and you need to figure out why it is happening. You can
always turn off atom sorting (atom_modify command), but that
won't fix the root problem.
Steve
I thought aboutthis but i use either nve with thermostat and barostat(optionaly ) or nvt .So my simulation volume shouldn't grow.right?
I thought aboutthis but i use either nve with thermostat and barostat(optionaly ) or nvt .So my simulation volume shouldn't grow.right?
you also get this kind of error message if your forces become too large
to be represented by a double precision floating point number (so you
get NaN). this can happen, when atoms are very close, which i turn can
happen when you have too high velocities and/or a too large a time step
and/or overlapping atoms.
a.
Is your box size fixed, or is it shrink-wrapped to the atom
positions? I'm guessing the latter, and your box is becoming
huge. That error is generated when you have more than
2 billion bins.
Steve
The box is size fixed and i use a nve or nvt integrator.
In this case, you must be getting the error on the first timestep?
How big is your box and how long is your neighbor cutoff?
Steve