Shear Deformation: Monomers and Simulation Box Deforming Out of Sync

Dear LAMMPS Users,

Hello. This is Masato Koizumi, and I am an incoming student at UCLA. I am using LAMMPS to do my molecular simulations.

I modeled a simulation box containing polymer chains. Initially, the simulation box is a cube. I used fix deform command to apply shear such that the box will deform into a parallelepiped**.** I have outputted dump files so that I could view the shear deformation in OVITO.

When OVITO reads the dump files, I see the instantaneous shear deformation of my polymer matrix without any problems

I wanted to add bonds in between the modeled monomers, and since the dump files only provides the x, y and z coordinates of the monomers, I have converted the dump files into a data file following the format used in data files outputted by the write_data command. Therefore, the first portion of my converted data file looks like the following:

LAMMPS data file via dump file, version 10 Feb 2015, timestep = 200000
250000 atoms
1 atom types
263728 bonds
1 bond types
-3.2590800000000002e+01 4.5627099999999999e+01 xlo xhi
-3.2590800000000002e+01 3.2590800000000002e+01 ylo yhi
-3.2590800000000002e+01 3.2590800000000002e+01 zlo zhi
1.3036300000000001e+01 0.0000000000000000e+00 0.0000000000000000e+00 xy xz yz

However, when OVITO read my converted data file, the monomers and the simulation box geometry is out of sync like the following:


The simulation box is now deforming differently compared to how it was reading the original dump files. I did not change the xy, xz and yz values but when I converted the dump files into a data file without changing the simulation box parameters and the timesteps, the simulation box is deforming differently.

I would greatly appreciate if anyone could provide me with any hints or advices in resolving this problem. Thank you.


Masato Koizumi

Mitch can likely answer (CCd).




Thanks for your mail. I can think of a few different ways to resolve this. Since this is essentially a visualization problem and not a simulation problem I am going to keep my answers to thing done in OVITO rather than LAMMPS.

First thing to try is to go back and use your dump files with just coordinate information and add the ‘Create Bonds’ modification within OVITO and supply it with some cutoff radius to draw bonds. Make sure OVITO knows your cell is periodic by checking the boxes under Input -> Simulation Cell. This way bonds will be created through the periodic boundaries without any trouble. Also, the header in your data files only has one bond type, so if you want OVITO to only create bonds to a subset of atoms, use the ‘Expression Select’ command before the ‘Create Bonds’. Order of operations in OVITO is bottom to top if you are using the graphical interface.

If you really like the data files with bond information in them, you can adjust the simulation cell boundaries within OVITO in Input -> Simulation Cell. The LAMMPS tilt factors go in the upper triangle of the 3x3 matrix that OVITO supplies.

From the picture you sent, and description of the problem, I would also suggest looking at your fix deform command to make sure you aren’t adjusting the length of x in addition to the tilt angles in the x and y planes.

Hope that helps, if you continue to see strange results please provide us with the LAMMPS input file that generates this data.


Mitchell Wood

Postdoctoral Appointee

Sandia National Labs - 1444 Multiscale Science