The Polymer Society of Korea

Abstract

This presentation introduces a physics-based multiscale modeling framework for predictive design of polymer materials, spanning both dynamic polymer networks and composite interface mechanics. In Part 1, a quantum-mechanics-coupled Monte Carlo molecular dynamics framework is developed to simulate dynamic covalent bond exchange reactions in vitrimer systems. The approach enables quantitative prediction of how monomer chemistry and bond-exchange kinetics influence network topology and thermomechanical properties. To improve computational efficiency and expand compositional exploration, large-scale simulation datasets are further utilized to construct AI-based surrogate models for rapid structure–property prediction. In Part 2, a multiscale interface characterization framework is presented for carbon fiber reinforced polymer (CFRP) composites. Realistic carbon fiber surface topologies obtained from atomic force microscopy are directly incorporated into finite element interface models to investigate load transfer mechanisms under multi-directional loading conditions. The analysis reveals distinct roles of chemical bonding and mechanical interlocking, as well as anisotropic interfacial behavior governed by surface morphology and cohesive properties.