Integrated multiscale modelling strategy for shape-memory polymer nanocomposites: Linking surface chemistry-driven filler distribution to macroscopic properties

초록

Obtaining uniform nanoparticle dispersion in polymer matrices is essential for multifunctional properties of shape-memory polymer (SMP) nanocomposites. However, existing computational approaches do not completely account for the effects of nanoparticle surface treatment on nanoparticle dispersion and the impact of this dispersion on nanocomposite properties. This work presents an integrated multiscale modelling framework combining all-atom molecular dynamics (MD), dissipative particle dynamics (DPD), and finite element (FE) homogenization to connect nanoparticle surface chemistry, filler distribution, interphase structure, and macroscopic properties of shape-memory polyurethane (SMPU) nanocomposites. Solubility parameters and mixing energies of SMPU segments and silica nanoparticles were obtained from MD simulations to evaluate surface treatment effects on dispersion. These results informed Flory–Huggins parameters for DPD, enabling quantification of nanoparticle clustering as a function of hard-segment content (HSC) and nanoparticle loading. Based on this methodology, a framework was developed to predict optimal surface treatments for enhanced nanoparticle dispersion. DPD-driven nanocomposite morphologies were then integrated with FE homogenization to estimate mechanical properties. The methodology was further applied to nanocomposites with non-spherical fillers, demonstrating prediction of mechanical performance of cylindrical nanoparticle-reinforced systems as a function of HSC and surface treatment. This approach provides a versatile computational tool for guiding nanocomposite design and optimization.

키워드