The thermo-oxidative aging mechanism of carbon fiber-reinforced epoxy (CF/EP) composites was studied by using reactive molecular dynamics (RMD) simulations with a reactive force field (ReaxFF), traditional molecular dynamics (MD) simulation, and machine learning (ML) methods. The decoupling of the tensile stress applied to the resin from the thermo-oxidative aging reactions in the resin was demonstrated. The tensile strength parallel to the fiber was much higher than that perpendicular to the fiber. The fiber–epoxy interfacial bonding interaction limits the thermal mobility of the resin near the interface, lowering the probability of aging reaction events and thereby prohibiting the material’s aging. The panoramic view of the aging reactions was obtained using a new holistic analysis approach, revealing the high breaking probability of hydroxyl groups, ether bonds, and C–N bonds in the resin skeleton. These chemical bonds were proven to affect the formation of the main products. The total number of chemical bonds in the material was linear with the ultimate tensile strength perpendicular to the fiber. By establishing the relationship between “products–chemical bonds–mechanical properties”, we built the groundwork for novel lifetime prediction model construction based on theoretical and data fusion methods.

中文翻译：

---

碳纤维增强环氧复合材料界面相关热氧化老化全景图


通过使用反应力场 (ReaxFF) 反应分子动力学 (RMD) 模拟、传统分子动力学 (MD) 模拟和机器学习，研究了碳纤维增强环氧树脂 (CF/EP) 复合材料的热氧化老化机理。 ML）方法。证明了施加到树脂上的拉伸应力与树脂中的热氧化老化反应的解耦。平行于纤维的拉伸强度远高于垂直于纤维的拉伸强度。纤维-环氧树脂界面键合相互作用限制了界面附近树脂的热流动性，降低了老化反应事件的可能性，从而阻止了材料的老化。采用新的整体分析方法获得了老化反应的全景图，揭示了树脂骨架中羟基、醚键和C-N键的高断裂概率。这些化学键被证明会影响主要产物的形成。材料中化学键的总数与垂直于纤维的极限拉伸强度呈线性关系。通过建立“产品-化学键-机械性能”之间的关系，我们为基于理论和数据融合方法构建新型寿命预测模型奠定了基础。