Re: possible minimum box size

[Jean-Noël Grad]

Hi again,

Sorry, I forgot to mention that the error message "number of cells 1 is 
smaller than minimum 8" is typically emitted on 1 MPI rank when the 
system contains a bond whose cutoff is larger than half the max 
interaction range calculated with the formula in my last email (or 
larger that one quarter if a dihedral is present).

Best,
JN

On 2022-03-01 01:03, Jean-Noël Grad wrote:
> Dear Ahmad Reza,
>
> The minimum box size can be calculated with the following formula:
>
> ```python
> max_interaction_cutoff = min(system.box_l /
> system.cell_system.node_grid) - system.cell_system.skin *
> (system.cell_system.get_state()['n_nodes'] > 1)
> ```
>
> If your polymer has dihedrals, divide this value by 2. Run the
> attached MWE with 1 or 2 MPI ranks to see the effect of the LJ cutoff
> and skin.
>
> Best,
> JN
>
> On 2022-02-28 16:15, Ahmad Reza Motezakker wrote:
> > Dear EspressoMD users,
> >
> > I have a system including 2 polymer chains. In the warm up stage, I
> > face below error:
> >
> > Encountered errors during integrate: ERROR: number of cells 1 is
> > smaller than minimum 8 (interaction range too large or min_num_cells
> > too large)
> >
> > The polymers are made of 15 beads (sigma = 1). cut off radius is
> > 2**(1/6). I have played with box size and skin and cut off to see what
> > is happening but it only works with box size of 57. Could you please
> > help me with that? I want to decrease the box size to the smallest
> > possible size.
> >
> > Thank you very much,
> >
> > Ahmad Reza