Research
We are equipped with rigorous numerical simulation schemes and advanced imaging techniques in conjunction with well-designed experimental methods to explore how geomaterials behave.
GEMs recently focuses on 'deep learning' for applications in geotechnical engineering. Examples include text-mining of geo-documents, the tunnel mapping, discontinuity extraction, simulation-based estimation of safety, and prediction of stress-strain evolution.
[Machine learning & Deep-learning based study]
Morphological characteriztion of granular soils
Image-based estimation of tunnel faces conditions
Evolution of stress-strain
Geomechanical analysis from numerically produced dataset
[Text Mining of Geo-Document by deep learning]
[Adaptive and Functional Landfill liner]
[ Waterless stimulation : Fracture propagation]
[Porous block, swelling aggregate]
[Bio-meditated soils]
[Deep learning-based rock classification]
Identification of igneous rock types
[Hydraulic stimulation: Phase field model]
Waterless fracturing sequesters 'Carbon Dioxide' with the development of microfractures.
The newly developed technique - Time-delayed pressurization - reduces the breakdown pressure and mitigates the acoustic emission.
[Hydraulic stimulation: Imaging]
The 3D x-ray computed tomographic imaging helps identify the fractures and evaluates their morphological characteristics.
[Enzyme-Induced Carbonate Precipitation (EICP) and Soil Improvement]
The carbonate precipitated in soils strengthens soils and its application includes the soil stabilization and dust suppression.
[Lattice-Boltzmann method: Two-phase fluid flow]
[Lattice-Boltzmann Method: Multi-phase fluid flow]
GEMS explores how two-phase fluid flows in the pore by Lattice-Boltzmann method. We evaluate the effect of capillary number, viscosity ratio, and Reynolds number in fluids and see how these variables determine the fate of spatio-temporal evolution of fluids at micro-scale by adopting the virtual 3D images from X-ray computed tomography.
The examples include the geological sequestration of carbon dioxide and its storage optimization, fate and transport of contaminant, gas flow during hydrate production, and hydrocarbon recovery.
[Deep Learning: crack detection]
The hydraulic fracture is detected by 'Deep learning' approach which is now being applied to 'digital geo-images' with the application of 'maintenance and inspection' of infrastructure.
[Carbonation in Oil well cement]
Oil well cement seals the borehole in the CO2 storage site and X-ray CT imaging virtually identifies the reaction zones in conjunction with chemical analysis with the cement matrix.
[Gas hydrate bearing sediment]
Characterization / In-situ testing / chamber development / production
[Micromechanics]
Micro-scale analysis of particle-pore behaviors
[Engineered soils]
Hydrophobic sand and its characterization (initiated at Lehigh Univ.)
Hydro-physical features
[Lightweight Aggregate and fractures in concrete]
Application of LWA in concrete
Discontinuity in cementitious materials
[Particle shape analysis]
Particle shape effect on geophysical and mechanical properties
Shape descriptors
[Environmental Construction Materials]
NOx removal of TiO2-porous block
Drainage and removal capacity