Hyunggwi Song

Korea Advanced Institute of Science and Technology

Papers

Total Citations

H-Index

About

Hyunggwi Song is a researcher whose work bridges mechanical metamaterials and biomechanics, with a focus on designing materials that exhibit precise, programmable mechanical behaviors. His most notable contribution is the introduction of a **Constrained Generative Inverse Design Network (CGIDN)** for creating customizable mechanical metamaterials with tailored strain-dependent Poisson’s ratios. This framework, detailed in his 2024 paper (27 citations), employs a **PCA-weighted loss function** to significantly enhance deep neural network training efficiency and accuracy, enabling high-fidelity inverse design validated through finite element analysis and experiments. This work has advanced the field of metamaterial design by offering a data-driven pathway to achieve desired mechanical properties on demand. Earlier, Song explored biomechanics in his 2016 paper (14 citations), where he demonstrated that a **springy pendulum model** can effectively describe swing leg kinetics during human walking, providing insights into gait dynamics. His research, while still early in its citation trajectory, shows clear impact in both computational design and biomechanical modeling. Song’s work is particularly valuable for students and researchers interested in the intersection of machine learning, material science, and biomechanics, as it offers practical tools for inverse design and a deeper understanding of human movement.