RESEARCH

Multiscale Mechanics & Materials Design Laboratory

RESEARCH

Multiscale Modeling & Simulation

Our laboratory possesses strong and well-established expertise in multiscale modeling and analysis, covering a broad spectrum of temporal and spatial scales. The methodological foundation of our research originated from the integration of classical molecular dynamics simulations and micromechanics-based property prediction frameworks. Over the past decade, this foundation has evolved into a comprehensive multiscale modeling capability spanning significantly wider length and time scales.

At the lower-scale end, we employ reactive molecular dynamics simulations based on ReaxFF, enabling rigorous analysis of bond-order evolution in polymers. Through this capability, we perform coupled investigations linking extreme environments, bond-order variations, and material property degradation, allowing us to systematically explore degradation mechanisms that cannot be captured by conventional force-field-based approaches. In parallel, our laboratory has been actively investing in first-principles calculations, driven by industrial demand and collaborative research opportunities related to functional low-dimensional materials. These efforts include the investigation of piezoelectric properties, bottom-up synthesis mechanisms, and energy storage materials.

At higher scales, we have developed capabilities in coarse-grained molecular dynamics, including the development of potential parameters tailored for polymer and composite systems. For process-scale phenomena such as material mixing and diffusion in heterogeneous systems, we also utilize dissipative particle dynamics (DPD) simulations. These methodologies are tightly integrated with the hierarchical multiscale modeling framework for nanocomposites developed by Professor Seunghwa Yang over the past 15 years. This integration enables a fully virtual, computer-driven workflow encompassing particle surface treatment, dispersion evaluation, microstructure formation, and homogenized property prediction—culminating in an all-digital design paradigm for polymer nanocomposites. This digital design framework supports materials development through virtual experiments characterized by solvent-free, mixing-free, and trial-and-error-minimized (UTM-free) approaches, significantly lowering the cost and complexity of nanocomposite design.

Through these multiscale modeling and analysis technologies, our laboratory actively collaborates with leading research institutions, resulting in joint publications and sustained research partnerships, as listed below