I have a box and I'm gonna compute elastic constant
using stress/strain curve. when I use this commands
fix 2 all deform 1 x scale 5 y volume z volume remap x
or
fix 2 all deform 1 x scale 5 y volume remap x
or
fix 2 all deform 1 x scale 5 remap x
I can't see any necking and failure.
I know it's impossible somehow to stretch a real metal cube to it's 5 times
length
but before it the box should reach failure.
It's whereas when I set a velocity to one side of cube I can see necking and
failure.
Hi, Reza. Strain localization in particle-based sims. is (very) often suppressed by the periodic boundary conditions. If you use the “volume” command, and pbc’s, ask yourself, how would a neck form? And what stress do you get? Is it well above the yield stress? (it should be). In any case, necking is notoriously hard to capture with 3D-pbcs. You could use a “column” with 1D-pbcs, but then will need a big system.
And anyway, if you want to compute the elastic constant, a deformation to 1% strain (as opposed to 500%) would be enough. So, think you have some background reading to do.