Modeling failure mechanisms of poly(p-phenylene terephthalamide) fiber using reactive potentials
Evidence and attribution¶
Authority of statements
Prose summarizes the article identified by doi and pdf_path. Reactive potentials are used as stated in the abstract; confirm which FF in the Methods section of the PDF.
Summary¶
Reactive MD with bond-order-style potentials studies tensile failure of PPTA (Kevlar-class) in amorphous and crystalline morphologies with vacancy defects. Modulus estimates (~6.7 GPa amorphous; ~350 GPa defect-free crystal) and defect sensitivity (5% N vacancy drops modulus to ~197 GPa in the abstract’s example) are combined with histograms of bonds/angles vs strain to localize failure at amide/phenyl C–N regions under extreme tension. The introduction stresses aramid use in ballistic fabrics where fiber failure absorbs projectile energy, motivating atomistic models that allow bond breaking beyond fixed-bond classical schemes (introduction, extract).
Methods¶
LAMMPS runs use the Strachan et al. ReaxFF set for PPTA-relevant C/H/O/N (papers/ReaxFF_others/1-s2.0-S0927025615004206-main-4.pdf, §2.1–2.3). Models include amorphous packings from an in-house builder (several molecular weights) and crystalline PPTA with C or N vacancies (§2.2), all in 3D periodic supercells with explicit atom positions for polymer and defects. After 300 → 600 K over 10 ps, 10 ps hold at 600 K, and cool to 300 K over 10 ps, quasi-static tensile loading applies 1% z-strain steps with 10 ps relaxations at σ_xx = σ_yy = 0 while L_z is fixed, to 30% engineering strain; histograms every 100 MD steps (§2.3). NVT with a Nosé–Hoover-style thermostat is used (PDF OCR spells “Noose-Hover”). Timestep: N/A — not recovered unambiguously from the indexed §2.3 text—confirm in the journal PDF. No global barostat; in-plane normal stresses are relaxed to zero by the protocol rather than hydrostatic pressure control. No electric field or enhanced sampling.
Force-field choice: NWChem B3LYP/6-31G bond stretches for six PPTA-relevant motifs are compared to ReaxFF scans to justify using the existing Strachan parametrization (§2.1); the paper does not report a new ReaxFF optimization campaign.
Static QM / DFT: NWChem B3LYP/6-31G scans serve QA for the reactive potential, not production AIMD.
Findings¶
Tension is borne mainly by bond stretching in amide–phenyl linkages before ultimate rupture localizes at C–N under large strain—mechanistic localization consistent with reactive bond order (abstract, §3). Histograms of bond lengths, angles, and phenyl metrics versus strain show where response nonlinearizes (§2.3, §3). The abstract quotes ~6.7 GPa (amorphous), ~350 GPa (defect-free crystal), and ~197 GPa with 5% N vacancies, plus a rule-of-mixtures-style blend for effective “fiber” moduli. Vacancy type and concentration are the main sensitivity levers on stiffness and failure progression. Limitations: idealized bulk-like cells omit spun-fiber skin–core texture noted in the Introduction. Corpus honesty: this is an external Comput. Mater. Sci. study (not van Duin-group); comparisons to experiment are only as far as the article’s abstract and discussion take them—use the PDF for numerical tables.
Limitations¶
- Idealized microstructures may omit spun-fiber texture and skin-core gradients.
-
Reactive FF quality depends on the specific parameterization used (see PDF).
-
Strain-rate and system size in MD differ from fiber draw experiments; treat modulus and failure trends as qualitative indicators unless mapped with careful scaling analysis (discussion caveat in polymer mechanics literature).
Relevance to group¶
Polymer mechanics with reactive MD parallels group interests in large-strain failure of organic solids.
Citations and evidence anchors¶
- DOI:
10.1016/j.commatsci.2015.07.010—papers/ReaxFF_others/1-s2.0-S0927025615004206-main-4.pdf.