2026-10-12T15:31:49Z https://www.scidb.cn/oai

10.57760/sciencedb.j00125.00281 2026-08-13T17:01:30Z

2026-08-13 Molecular dynamics simulation of mechanical properties in carbon fibers: a quantitative study on orientation, porosity, and amorphous microstructures doi:10.57760/sciencedb.j00125.00281 en Establishing quantitative structure–property relationships in carbon fibers (CFs) is crucial for designing next-generation high-performance materials. Guided by experimental data, we constructed a series of CF models incorporating varying y-axis orientations(0°–26°), pore characteristics (content and orientation), crystalline/amorphous structures (content and graphitization degree), and their coexistence using large-scale ReaxFF molecular dynamics simulations. Nonlinear regression was employed to relate structural features to mechanical properties quantitatively. Results indicate that the axial Young’s modulus is primarily governed by axial stress and sp2 carbon content, while tensile strength is dictated by the combined contributions of sp2 and sp/sp3 hybridized carbons. Both tensile strength and modulus decrease with increasing porosity up to 25%; at a given porosity, pores aligned perpendicular to the fiber axis are more detrimental than axially aligned ones. An anomalous increase in elongation at break occurs beyond 15% porosity, attributed to a faster decline in modulus relative to strength. Exponential functions effectively describe the correlations between pore features (content, size, coalescence) and mechanical performance. Increasing the amorphous content from 5 to 36.87 wt% also leads to exponential reductions in strength and modulus. The predicted modulus (450–1200 GPa) for models featuring coexisting amorphous structures and pores aligns closely with experimental data. Divergent decay trends between strength and modulus give rise to distinct CF performance series. This computational framework offers valuable guidance for the targeted structural design of high-performance CFs. CFs; microstructures; mechanical properties; molecular dynamics; numerical fitting Yuanyuan Ma RESTRICTED dataset Science Data Bank