[lammps-users] bug??? on "pair_style hybrid"

Hello,

I keep trying “pair_style hybrid” with dpd and coul/cut (or coul/long).
My data file contains 1 polyeletrolyte chain and 50 counterions.
When I run a simulation (nve/nvt/npt), I obtain final configuration (box size) which is 10- or so-fold larger than initial configuration.
In other words, distance between particles expands significantly and polyelectrolyte chain is pushed to the edge of the box.
If Coulombian interactions are appropriately considered, I think box size does not significantly change.
Could you please check if this is a bug?

Thank you.

Hello,
I keep trying "pair_style hybrid" with dpd and coul/cut (or coul/long).
My data file contains 1 polyeletrolyte chain and 50 counterions.
When I run a simulation (nve/nvt/npt), I obtain final configuration (box
size) which is 10- or so-fold larger than initial configuration.
In other words, distance between particles expands significantly and
polyelectrolyte chain is pushed to the edge of the box.
If Coulombian interactions are appropriately considered, I think box size
does not significantly change.
Could you please check if this is a bug?

_how_ should anybody be able to check something
from such a vague and confusing description?

how can your box size change with nve or nvt?

there is no "edge" in periodic boundary conditions.

if you suspect a bug and want somebody to fix it:
please provide some form of _proof_ and don't
speculate or guess.

thanks,
    axel.

p.s.: the vast majority of people suspecting a "bug",
simple have made errors in their input or are confused
about how to use one or more features correctly.

It doesn’t change during simulation run.
There’s missing atoms or bond atoms, so I increased box size to avoid it.
Error messages are like:
ERROR on proc 1: Bond atoms 66 67 missing on proc 1 at step 970

As a result, the final box size that doesn’t give an error becomes much larger in the end.

It doesn't change during simulation run.
There's missing atoms or bond atoms, so I increased box size to avoid it.

avoiding an error does not mean you solved the problem.

Error messages are like:
ERROR on proc 1: Bond atoms 66 67 missing on proc 1 at step 970
As a result, the final box size that doesn't give an error becomes much
larger in the end.

that usually happens when atoms move too fast.
the potential reasons for that are manyfold (too large
time step, too steep potential, too infrequent neighborlist
update, too smal neighborlist skin, bad choice of potential,
bad choice of starting configuration, and so on).

unless you understand why you get a specific error,
it is an invitation for trouble to continue by avoiding it.

GI-GO!

axel.

Thank you very much.
I think I solved the problem by removing initial velocity for all particles.