Body:

Hi everyone,

I came across an interesting paper recently where the authors used MD simulations in Materials Studio to generate stress–strain curves for CB/NBR composites. The curves they showed clearly go through an elastic region, a yield point, and then a fracture/drop-off. I've attached the relevant text and figures from the paper below for reference.

Here's what's been bugging me a bit. My understanding is that the COMPASS force field is non‑reactive, meaning it shouldn't be able to model actual covalent bond breaking. So I'm trying to wrap my head around why the stress–strain curves still show a distinct yield and eventual "fracture" behavior in the simulation. What's actually happening at the molecular level in that drop‑off stage?

Also, the way they set up the simulation caught my eye. They applied a constant strain rate along the Z‑axis, but for the X and Y directions they mention applying a compressive pressure. I vaguely remember seeing an answer from a teacher on this forum a while back, suggesting that applying 1 atm in the lateral directions allows the box to relax freely and mimic a realistic Poisson's ratio effect. Is that what's going on here, or is it a different technique entirely?

Any insight from those of you with more MD experience would be massively appreciated. Thanks a ton in advance!

Attachment note:

The mechanical properties of the CB/NBR models were evaluated through axial stretching simulations using molecular dynamics. The simulation protocol involved applying compressive pressure along the x and y axes while maintaining tensile strain in the z direction. The model was subjected to incremental stretching at a constant strain rate of 0.02 per step over 100 consecutive steps, with the stretching configuration illustrated in Fig. 2(d).