Questions about the fix rigid/nve small command

Hi everyone
May I have a idea that whether this fix rigid/nve small command can be used to generate the finite-size clump with shape, as this image?
1730249746148

And secondly, I have tryied to use the attached command in lammps, but it seems that there is some wrong about the atom_style sphere, then I also tried to use the the atom_style “ellipsoid”, “line” and “tri”, however, they are not fine, in my read_data, i have a column for the molecule ID. I used all these recommended atom_style from fix rigid documentations, but they do not seem work. May I have different options?

atom_style sphere
fix 11 all rigid/nve/small molecule
compute t1 all temp/sphere
thermo_modify temp t1
fix 9 all press/berendsen x {pstart} {pend} {pdamp} y {pstart} {pend} {pdamp} z {pstart} {pend} ${pdamp}

also for the infile keywords, as suggested, if we hope to use the clump made of overlapped spheres, it seems that we have to provide attributes by ourselves, so is this file included in the read_data file, with the diameter, position, velocities,etc? or should we have the new file just for these attributes?

Many thanks for your help and suggestions for this.

Since you have overlapping particles, you would need to pre-compute mass and moments of inertia and read them with the infile keyword.

BTW: I would suggest to use fix rigid/small instead of fix rigid/nve/small.

There is nothing wrong with those atom styles. They function as documented. However, the documentation does not state that those styles contain a molecule ID. So you have to either use a hybrid atom style, e.g. hybrid sphere bond that includes moldule IDs, or add molecule IDs via fix property/atom.

Using fix press/berendsen is a bad idea (in general, but even more so for rigid objects). Better to use fix rigid/npt/small.

No.

That is what the documentation says.

Dear akohlmey

Many thanks for your help and response, it is really helpful. it just I still have some confusions.
would you mind share the idea why not include the “nve” in the fix rigid/small. I assum that we always need to use something like nve, for spherical particles, i remembered that we need use “fix 1 all nve/sphere” something like this.

for the fix press/berendsen, i used this to achieve the certain system stress level, if wew use the fix rigid/npt/small, can this achieve the same target?

fix rigid/nve/small is a more complex implementation. It basically follows the Nose-Hoover scheme without including a thermostat or a barostat. fix rigid/small is the original implementation. I always prefer simpler since simpler means less chances for bugs.

Please study the documentation.

1 Like