Regarding lost atoms

Dear user,

I am learning to use LAMMPS in my research now,

i was trying to make concentric cylinder with different material Ag and Cu but some atoms got lost during Simulation. Script file as given below,
Please help me out:

units metal
dimension 3
boundary p p p

atom_style atomic
neighbor 2 bin
neigh_modify delay 5

create geometry

lattice fcc 4.079 #Ag

region bigbox block 0 50 0 100 -10 10
create_box 2 bigbox

region inner cylinder y 25 0 2.5 30 60
region outer cylinder y 25 0 5 30 60
region hollow intersect 2 inner outer

create_atoms 1 region inner
create_atoms 1 region outer

group all region hollow

mass 1 107.8 #Ag

pair_style eam
pair_coeff * * Ag_u3.eam

delete_atoms region inner

lattice fcc 3.597 #Cu

region composite cylinder y 25 0 2.5 30 60

create_atoms 1 region composite

group all region composite

mass 1 63.456

pair_style eam
pair_coeff * * Cu_u3.eam

fix 1 all nve
fix 3 all temp/rescale 5 100 100 0.01 1.0

timestep 0.003
thermo 1000

dump 2 all image 1000 image.*.cfg type type &
zoom 1.6 adiam 1.5

dump snap all cfg 1000 dump.config.*.cfg mass type xs ys zs

run 2000

ERROR: Lost atoms: original 9577 current 9476 (…/thermo.cpp:395)

Dear user,

I am learning to use LAMMPS in my research now,

i was trying to make concentric cylinder with different material Ag and Cu
but some atoms got lost during Simulation. Script file as given below,
Please help me out:

units metal
dimension 3
boundary p p p

atom_style atomic
neighbor 2 bin
neigh_modify delay 5

# create geometry
lattice fcc 4.079 #Ag

region bigbox block 0 50 0 100 -10 10
create_box 2 bigbox

region inner cylinder y 25 0 2.5 30 60
region outer cylinder y 25 0 5 30 60
region hollow intersect 2 inner outer

create_atoms 1 region inner
create_atoms 1 region outer

group all region hollow

mass 1 107.8 #Ag

pair_style eam
pair_coeff * * Ag_u3.eam

[...]

pair_style eam
pair_coeff * * Cu_u3.eam

This cannot work. You need to find a EAM potential file that is
parameterized for both Ag and Cu and use pair_style only once. The other
option is to use pair hybrid to combine the two eam potentials you used but
you still need the Ag-Cu cross terms - hence not recommended.

Ray