Deleting random atoms

Hello again LAMMPS users,

I am performing some tests over a bulk conformed by 64,000 atoms, I tried to incorporate a vacancy defect to see the resulting behavior, for this I read that the command delete atoms can delete a specific fraction of atoms randomly, it goes like
delete_atoms porosity region fraction seed
the problem is that by the time of using it the atoms are not erased or it erase a really low number of atoms, I have tried many combinations of “fractions” and “seeds”, even some that should generate an error by example set fraction to 10 (supposedly represents 1000%), at the end I have only managed to delete 16 atom (roughly a 0.0025%). Next is the starting piece of my scrip

----------- INITIALIZATION ------------

units metal
dimension 3
boundary p p p
atom_style atomic
atom_modify map array

------- INPUTS ---------

modify this part for the bulk size

a,b,c (x,y,z)=box size

variable a equal 40
variable b equal 20
variable c equal 20

---------- Create Atoms -----------

#Ni

lattice custom 1.0 a1 2.999 0 0 a2 0 2.999 0 a3 0.0 0.0 2.999 &
basis 0.0 0 0.0 orient x 1 0 1 orient y 0 1 0 orient z -1 0 1
region supercell prism 0 4.22 0.0 2.999 0.0 4.22 0 0 0 units box
create_box 2 supercell
create_atoms 1 region supercell

#Ti
lattice custom 1.0 a1 2.999 0 0 a2 0 2.999 0 a3 0.0 0.0 2.999 &
basis 0.5 0.5 0.5 orient x 1 0 1 orient y 0 1 0 orient z -1 0 1
create_atoms 2 region supercell

replicate $a $b $c

---------------- Defining porosity 1=100% -------------

delete_atoms porosity supercell 0.01 45348

----------- POTENTIAL --------------------

As always thank you for your help and any hint of why this happens is very welcome

Mario Muralles

Hello again LAMMPS users,
I am performing some tests over a bulk conformed by 64,000 atoms, I tried to
incorporate a vacancy defect to see the resulting behavior, for this I read
that the command delete atoms can delete a specific fraction of atoms
randomly, it goes like
delete_atoms porosity region fraction seed
the problem is that by the time of using it the atoms are not erased or it
erase a really low number of atoms, I have tried many combinations of
"fractions" and "seeds", even some that should generate an error by example
set fraction to 10 (supposedly represents 1000%), at the end I have only
managed to delete 16 atom (roughly a 0.0025%). Next is the starting piece of
my scrip

the secret is in the command replicate.
you replicate your atoms 40*20*20=16000 times. however, for your
deletion you use only the region describing the original cell
so even if you delete *all* atoms in that region, you would only
delete 0.00625% of the total atoms.

axel