Strange relation between simulation box size and computed data

​please always reply to the mailing list and not only individual people.​

Thanks a lot for your response Aidan and Axel!!

Maybe it could be a way to modify the Voronoi method for approximate
calculation of the cluster volume, or at least of it's change.

i don't see how. ​voronoi tessellation is a method for *dividing* a given
volume, not for determining a volume.​

​axel.​

Thanks a lot for your response Aidan and Axel!!

Maybe it could be a way to modify the Voronoi method for approximate calculation of the cluster volume, or at least of it’s change.

With all respect,
Ruslan

The voronoi package in lammps is built over the voro++ library. That library has features to deal with a case like the one you are describing: http://math.lbl.gov/voro++/examples/irregular/
Note however that the solution described there is purely geomettrical and other techniques such a probe particle approaches as Axel mentioned may be more suitable to construct volume descriptors that suit better your needs.
Carlos

Thanks a lot for your suggestions Carlos!

It may be a good solution, however I first time hear about probe particle
approach technique and have no idea what it stands for. Could you, or
someone else, give me a link?

With all respect,
Ruslan

Thanks a lot for your suggestions Carlos!

It may be a good solution, however I first time hear about probe particle
approach technique and have no idea what it stands for. Could you, or
someone else, give me a link?

​this method is widely used in biomolecular simulations to assess the
solvent accessible surface are, for example. of course, that surface - by
definition - also defines an effective volume of the biomolecule.

https://en.wikipedia.org/wiki/Accessible_surface_area

please note that the dependency of the resulting surface (or volume) on the
size of the probe particle highlights the inherent arbitrariness of such
definitions.

axel.​

The particle probe approach results may end up being very similar to the one I linked because “I believe” in the Voro++ you can initialize the WZ cell of each particle to be a sphere. Take my words with a grain of salt as I have never done it but just repeated for you what the docs state.
Carlos