Seoul National University · Engineering
Professor Jae-Joon Song's research lab specializes in rock mechanics and fluid flow characterization in porous media, with a strong focus on geological CO₂ storage, non-Darcy flow, and pore-scale modeling. The lab integrates advanced imaging techniques such as micro-CT scanning with experimental mechanics and computational modeling to investigate fluid-rock interactions, permeability evolution, and failure mechanisms in fractured and porous rocks. Key research directions include the development of pore channel models from reconstructed microstructures, effective pressure law applications under varying confining and pore pressures, and the analysis of inertial flow effects using the Forchheimer equation. The lab also explores 3D-printed rock analogs to simulate natural rock behavior under controlled conditions, enabling detailed study of crack propagation and strain localization.
Figures are computed from collected data and may differ slightly.
Pore-scale modeling with a reconstructed rock microstructure has become a dominant technique for fluid flow characterization in rock thanks to technological improvements in X-ray computed tomography (CT) imaging. A new method for the construction of a pore channel model from micro-CT image analysis is suggested to improve computational efficiency by simplifying a highly complex pore structure. Ternary segmentation was applied through matching a pore volume experimentally measured by mercury intr
Abstract The liquid permeability of rock with distilled water or brine is different from that obtained using gas by variation in the confining pressure P c and pore pressure P p . In this study, as part of the research on CO 2 geological storage, the permeability of sandstone was measured using supercritical CO 2 , and the effect of P c and P p on this permeability was analyzed. For applying the effective pressure law to the analysis, an effective pressure coefficient for permeability was derive
Open papers in the app to read, cite, and organize with AI.
