Mechanisms of Shock Dissipation in Semicrystalline Polyethylene

Abstract

Semicrystalline polymers are lightweight, multiphase materials that exhibit attractive shock dissipation characteristics and have potential applications as protective armor for people and equipment. For shocks of 10 GPa or less, we analyzed various mechanisms for the storage and dissipation of shock wave energy in a realistic, united atom (UA) model of semicrystalline polyethylene. Systems characterized by different levels of crystallinity were simulated using equilibrium molecular dynamics with a Hugoniostat to ensure that the resulting states conform to the Rankine–Hugoniot conditions. To determine the role of structural rearrangements, order parameters and configuration time series were collected during the course of the shock simulations. We conclude that the major mechanisms responsible for the storage and dissipation of shock energy in semicrystalline polyethylene are those associated with plastic deformation and melting of the crystalline domain. For this UA model, plastic deformation occurs primarily through fine crystallographic slip and the formation of kink bands, whose long period decreases with increasing shock pressure.

Nucleation Patterns of Polymer Crystals Analyzed by Machine Learning Models

“…In contrast to the classification methods based on the picture of extended chain parts (stems), which occur with a very low probability in the molten state only, we do not have to make any presumptions about the conformational state of the corresponding sequence. Both stem length based and local orientation order parameters show bimodal distributions that can be employed for defining decision boundaries manually, or in unsupervised fashion, for instance, via the fuzzy c-means cluster algorithm . The resulting crystallinity measures, however, are not neccessarily compatible in terms of sensitivity for crystal domains especially at early nucleation times.…”

“…Both stem length based and local orientation order parameters show bimodal distributions that can be employed for defining decision boundaries manually, 27 or in unsupervised fashion, for instance, via the fuzzy c-means cluster algorithm. 28 The resulting crystallinity measures, however, are not neccessarily compatible in terms of sensitivity for crystal domains especially at early nucleation times. As we will discuss, different emphases on local conformation and alignment of the immediate environment influence the crystalline fraction, and the shape of the resulting cluster surfaces.…”