reading initial configuration and simulation box boundaries

Hello LAMMPS users,

I have a question about reading input files.

Context: crystallin materials, simulation of twins in Titanium

I have a configuration, result of a previous simulation, which I would like to use as an input file for another simulation. This configuration respects periodicity, I check this with the visualization tool OVITO i.e. using the common neigh. analysis and looking at the energy per atom ( I check there were no strange energy excess at the boundary).

What I did is writing a small program to create an input file for LAMMPS from the dump file of the previous simulation. In ths input file number of atoms, box limits of my triclinic simulation box and atom typology + coordinates are specified.

I run the new simulation using this file as input and what I noticed is that from the really begin of the simulation the box limits change and I loose periodicity in the direction X and Y. This is strange, because I am pretty sure that periodicity is respected in the initial configuration so I dont’t understand how it can be lost so quickly. The lost of periodicity can be seen using OVITO and looking at the energy per atom, all the atoms at the boundary have not the good energy value …

Find attached the dump fie with the initial configuration (macle), the input file for the new run (Ti_macle) I wrote and the LAMMPS script (macle.inp).

Thank You for Your help in advance, I wish a good Monday !

Carolina

ps. a naive question … using FIX NVT should imply no change in the box volume but possible change in the box limits, am I right ? But I think that if periodic boundary conditions in the 3 dimensions are used the system is forced also to not change its shape.

macle.inp (1018 Bytes)

Ti_macle (1.53 MB)

macle (978 KB)

Hello LAMMPS users,
I have a question about reading input files.

​first off, please do not attache such large files​. keep in mind you are
e-mailing this to a very large number of people. better to compress them,
even better to create smaller examples and best to place them somewhere on
a file hosting service (google drive, dropbox, etc.) and provide the links
to them.

Context: crystallin materials, simulation of twins in Titanium

I have a configuration, result of a previous simulation, which I would
like to use as an input file for another simulation. This configuration
respects periodicity, I check this with the visualization tool OVITO i.e.
using the common neigh. analysis and looking at the energy per atom ( I
check there were no strange energy excess at the boundary).

What I did is writing a small program to create an input file for LAMMPS
from the dump file of the previous simulation. In ths input file number of
atoms, box limits of my triclinic simulation box and atom typology +
coordinates are specified.

I run the new simulation using this file as input and what I noticed is
that from the really begin of the simulation the box limits change and I
loose periodicity in the direction X and Y. This is strange, because I am
pretty sure that periodicity is respected in the initial configuration so I
dont't understand how it can be lost so quickly. The lost of periodicity
can be seen using OVITO and looking at the energy per atom, all the atoms
at the boundary have not the good energy value ...

Find attached the dump fie with the initial configuration (macle), the
input file for the new run (Ti_macle) I wrote and the LAMMPS script
(macle.inp).

​the problem is, that the "box boundary" information in a dump file is not
the same as the "box​ length" values in the data file.
when i load your dump file into VMD and then write out a data file from
that, i get the following info in the data file header:

LAMMPS data file. CGCMM style. atom_style atomic generated by VMD/TopoTools
v1.7 on Mon Sep 04 11:16:36 EDT 2017
15552 atoms
0 bonds
0 angles
0 dihedrals
0 impropers
1 atom types
0 bond types
0 angle types
0 dihedral types
0 improper types
2.462990 115.167320 xlo xhi
0.835051 118.835122 ylo yhi
-0.477319 20.292666 zlo zhi
-0.627570 -0.479541 0.977388 xy xz yz

​which results in a box length in x direction of ​112.704330 instead of the
113.811430 you have in your data file.

for more details on this subject, please see:
http://lammps.sandia.gov/doc/Section_howto.html#howto-12

axel.