Hello all,
I have started to learn simulations in lammps. I created a lattice of 4000 atoms but it created 4186 no of atoms, also it took wrong lattice spacing. And at last, it showed this error (ERROR: Lost atoms: original 4186 current 3927 (…/thermo.cpp:442)).
Someone, please help me in this. I am posting my input script below.
Thanks in advance
Input file for uniaxial tensile loading of single crystal aluminum
Mark Tschopp, November 2010
------------------------ INITIALIZATION ----------------------------
units metal
dimension 3
boundary f f f
atom_style atomic
variable latparam equal 3.60
----------------------- ATOM DEFINITION ----------------------------
lattice fcc {latparam}
region whole block 0 10 0 10 0 10
create_box 1 whole
lattice fcc {latparam} orient x 1 1 1 orient y -1 1 0 orient z -1 -1 2
create_atoms 1 region whole
------------------------ FORCE FIELDS ------------------------------
pair_style meam/c
pair_coeff * * library.meam Cu Cu.meam Cu
------------------------- SETTINGS ---------------------------------
compute peratom all pe/atom
1 correction boundary is p p p not f f f .
Input file for uniaxial tensile loading of single crystal aluminum
Mark Tschopp, November 2010
------------------------ INITIALIZATION ----------------------------
units metal
dimension 3
boundary p p p
atom_style atomic
variable latparam equal 3.60
----------------------- ATOM DEFINITION ----------------------------
lattice fcc {latparam}
region whole block 0 10 0 10 0 10
create_box 1 whole
lattice fcc {latparam} orient x 1 1 1 orient y -1 1 0 orient z -1 -1 2
create_atoms 1 region whole
------------------------ FORCE FIELDS ------------------------------
pair_style meam/c
pair_coeff * * library.meam Cu Cu.meam Cu
------------------------- SETTINGS ---------------------------------
compute peratom all pe/atom
The system you obtain is exactly what you asked Lammps to do. Did you realize that your lattice is rotated? If yes, how did you determine that you obtain the “wrong lattice spacing”? Julien
in addition to that info. with rotated lattices, you cannot use any multiple of the lattice spacing in that direction. in each direction x-, y-, and z, you have to make certain, that the surface is properly replicated across periodic boundaries. this has been discussed many times on this mailing list in the context of slab systems. for a bulk system, you have to fulfill one more constraint. also, note that you should check for close contacts, because for computed positions on rotated lattices numerical errors, can lead LAMMPS to not detect atoms that are periodic replicas, because you placed your box exactly on lattice boundaries. shifting the box by some small percentage (e.g. 0.1* lattice units), removes any ambiguity and can avoid such issues that can happen even if you have used the proper multiples of lattice spacings in all directions.
again, this has been discussed repeatedly here in multiple contexts and thus there should be previous discussions in the mailing list archives.
axel.