[lammps-users] shear stress experiment


this question is related to the one of Nitin some time in January. I could not find any final result of that thread.

I am just trying to measure stress-strain curves for shearing of solid bulk material at a certain pressure and temperature. So initially, I equilibrated a fully periodic copper block at 300K and 5.1GPa using fix npt.

The problem I am facing now is that neither npt nor press/berendsen work with a triclinic box (at the moment, I can not see any physical reason that would not allow that). Thus, I had no other immediate option, but to keep the dimensions of the box fixed, while shearing it.

Now, when straining the material, let's say in xy, the normal stresses, in particular tyy, rise enormously. This is logical since the atoms need more space as they are forced out of their 'low stress' positions. I am not quite convinced that it is physically correct to keep the volume constant in that case, because it would actually expand. I'd rather run the system still with NPT to keep the normal pressures at the original level.

Is there any software related limitation that makes it impossible to rescale a triclinic simulation box while shearing it?


This is the difference between free-end and fixed-end shear.
Expts are done both ways. I agree it would be useful
to allow in LAMMPS. Nothing is "impossible", it just
means finding time to do it.