[lammps-users] Need suggestion regarding shear simulations

LAMMPS LAMMPS Mailing List Mirror
1

Dear Lammps users,

I am working on an 3-D periodic atomistic system and want to calculate its mechanical properties. I am planning to calculate Lame Constants for my system so that I can predict various moduli, etc. In order to do so, I need to apply tensile and shear strain onto the system and see its response stresses.

I am having confusion on how to apply a shear strain. I looked at the fix_deform command but the terms (styles) xy, yz and xz were not clear enough to me. My approach is quite simple

  1. Start with a minimized system
  2. Apply tensile or shear strain (within one step, 0.1% far within elastic limits)
  3. Minimize the new system/Store various stress components for the minimized system. (During this minimization, I want top and bottom surfaces to be fixed).

If anyone is aware of any other approach, I would appreciate hearing about that too.

I am quite sure that fix_deform should be the way to go but I am not sure what keywords should I use. For example, “remap”. Moreover, it seems like fix_deform does not work with minimization algorithm.

I will highly appreciate if someone who has done such calculations could guide me here.

Thanks in advance.

2

Vikas,
Look into the conversation few days/weeks back on triclinic tilt factor where the lammps terms are explained. You can also apply tetragonal strain on the system to find shear components as an alternative approach to pure shear. (See ‘Ab-initio studies of pressure induced phase transitions in BaO’ by Uludogan et. al , 2002. ). Hope this helps.

Thanks
Arnab

3

If you want to deform the system to some new state, then relax it,
and repeat. You probably want to use the displace_box command,
and do minimizations in between.

Displace_box is similar to fix_deform, but does the deformation all
at once.

Steve

4

Dear Steve, Peter, and Arnab…
Thank you very much for the suggestions. I will try them out once…

Regards,
Vikas