The uncertainty inherent to DEM simulations of interlocking particles

In industrial applications, the handling of heterogeneous mixtures of phases and materials poses challenges for direct measurements and experiments, necessitating complementary modeling approaches. The Discrete Element Method (DEM) is commonly used for simulating the flow of granular systems, typically with spherical particles. However, certain applications, such as recycled polymers and batteries, require alternative non-convex particle representations in DEM simulations. Tetrapods are a promising shape candidate for modeling the flow behavior of such materials, as well as the associated uncertainty. We investigate the impact of the tetrapods' properties on the outcome and uncertainty inherent to DEM-based simulations. We demonstrate that tetrapods are effective for modeling interlocking materials, with their shape and size parameter significantly affecting interlocking behavior. Most interestingly, we can correlate the shape and size of tetrapods to the uncertainty inherent to our simulations. Specifically, we find that this uncertainty is positively correlated with both tetrapod size and the interlocking parameter ξ/D that quantifies their non-convexity. Lastly, we provide guidelines for selecting optimal tetrapod parameter sets for accurately modeling materials based on mean and variability measured in experiments.