LAMMPS WWW Site - LAMMPS Documentation - LAMMPS Mailing List Archives
Re: [lammps-users] Voronoi/atom issue
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [lammps-users] Voronoi/atom issue

From: Julien Guénolé <julien.guenole@...4839...>
Date: Fri, 10 Nov 2017 14:19:32 +0100

That is indeed odd to get a volume twice as the simulation box... he volume should be anyway equal or smaller than the simulation box.

But in general, one should not use the Voronoi volume when atoms belong to a surface. In such a case, it is impossible to build a "voronoi cell" around the atom, and voro++ (used in Lammps) will stop the volume at the potential cutoff or ghost atoms distance (don't remember exactly). So, the "volume" of the surface atom will be anyway wrong. In a more global point of view, the volume of a surface is an ill defined concept.

Also, I do not see any accurate way to compute the volume of an atomic sphere within Lammps...

On 09/11/17 14:05, Stefan Paquay wrote:
This seems weird indeed. Can you dump the per atom volume and see how the distribution is? Maybe that will help in assessing what goes wrong.

On Nov 8, 2017 11:34 PM, "Eric Murphy" <murphyericjames@...24...> wrote:
Answering because nobody else has.

I cannot give an idea as to why it's >2x as big (more worryingly its not constant) - you should get the volume of the box if everything's done correctly.  The voronoi tesselation doesn't know anything about where the cluster ends and the void begins.

In the mean time, try the scientific method, test it on some simpler systems where things should behave better (non-clustered) and see what you get.

Eric Murphy
Graduate Research Assistant, Mechanical Engineering
Iowa State University

On Wed, Nov 8, 2017 at 12:01 PM, Rajdeep Behera <rajdeepbehera87@...24...> wrote:
Hi all,
I want to calculate the volume of a spherical nano-particle resides inside of a cubical simulation box (see the attached figure) utilizing Voronoi tessellation. My input commands are given below,

units      metal
dimension  3
boundary   p p p
atom_style  atomic

lattice      fcc 4.247
region       whole  block 0 50 0 50 0 50 units box
create_box   2 whole
region       Domain sphere 25 25 25 10 units box 
lattice      fcc 4.247
create_atoms 2 region Domain

region         Core sphere 25 25 25 6 units box 
group          Core region Core 

pair_style       eam/alloy
pair_coeff      * * AlGa.eam Al Ga
neighbor        2.0 bin
neigh_modify    delay 10 check yes

minimize        1e-15 1e-15 1000 100000

compute         v Core voronoi/atom only_group
compute         tv Core  reduce sum c_v[1]

dump            1 all custom 100 dat.test.*.cfg id type xs ys zs vx vy vz

thermo_style    custom step time temp vol c_tv

thermo          10
timestep        0.001

fix             relax all nvt temp 2.5 2.5 0.001

run 1000
write_data      Eql-NVT-1 

But, interestingly the volume of the nanoparticle is computed almost double of the simulation box. Here is my output,

Step Time Temp Volume c_tv 
      64            0            0                       125000    265676.44 
      70        0.006    50.253752             125000    265675.37 
      80        0.016 1.4260517e-010       125000    265682.03 
      90        0.026 3.2938985e-006       125000    265682.01 
     100        0.036  0.040481345          125000    265679.68 

Am I missing anything? 

Please make your comment/suggestion in this regard.



Check out the vibrant tech community on one of the world's most
engaging tech sites,!
lammps-users mailing list

Check out the vibrant tech community on one of the world's most
engaging tech sites,!
lammps-users mailing list

Check out the vibrant tech community on one of the world's most
engaging tech sites,!

lammps-users mailing list

Dr. Julien GUÉNOLÉ
Postdoctoral researcher
Department of Materials Science & Engineering
Institute 1: General Material Properties
Friedrich-Alexander-University Erlangen-Nürnberg
Martensstr. 5, 91058 Erlangen, Germany
-----------------------  Office 3.40
Phone [office] (+49) 91 31 85 27 486
Email          julien.guenole@...4839...
Twitter        @nanouayeur