Stress concentration in polymer by introducing a notch

Dear Everyone,

I want to simulate the stress concentration phenomena in polymers by stretching the sample with a notch. Please see the attached figure for the real experiment I did.

Although I have simulated the uni-axial tension tests for such polymers without any notch (i.e. polymer molecules fill the simulation box), I am not sure how to simulate this type test.

My question is how to introduce the notch in the simulation box in Lammps if it is feasible.

Thank you for any comments and suggestions!

Lili Zhang

Stress concentration in polymer.jpg

Dear Everyone,

I want to simulate the stress concentration phenomena in polymers by
stretching the sample with a notch. Please see the attached figure for the
real experiment I did.

Although I have simulated the uni-axial tension tests for such polymers
without any notch (i.e. polymer molecules fill the simulation box), I am not
sure how to simulate this type test.

My question is how to introduce the notch in the simulation box in Lammps if
it is feasible.

it depends on what kind of coordinate data exactly you have available.
did you make your uniaxial stress test with full periodic boundary conditions?

to have the notch properly represented, you will have to turn off
periodic boundaries in that direction, and that also probably means,
that you have to make a test *without* the notch to validata, how much
just this change affects your results.

then you have two choices (similar to what you'd do in an experiment).

1) run a simulation with a prism shaped intenter, perhaps even at
elevated temperature and go on until the notch has formed properly
*and* the polymer sufficiently equilibrated around it.

2) manipulate the polymer block to remove all molecules/atoms
from a prism shaped region to get your notch and then equilibrate
that system before you do your tests. this can be trickier than it
sounds, since you have molecules and not just atoms.

you should be aware that, like in real life experiments, you are not
likely to get the exact same object from either method and take
that into account when you are discussing your results.

axel.

Dear Axel and Steve,

Thanks a lot for your comments. For Axel’s question, yes, I used periodic boundary conditions along 3 directions when I did uni-axial stress simulation without a notch.

The choice (1) seems more doable for my polymer system with very long molecules, since manipulating such entangled chains is very difficult and I have no idea if I can get the notch shape after equilibration.

Does Lammps provide the tools or command to include an indentor? I didn’t find any keyword in the documentation about indentation.

Thanks again,

Best regards,
Lili

See the fix indent command.

Steve