Fluid-Structure Interaction
Fluid-structure interaction research comprising the following:
-
Modeling hydrodynamic, mechanical, and thermal stresses in complex porous media, microfluidic devices, biobased systems and microchannels using Fluid-Structure Interaction (FSI) approach.
-
Prediction of fluid occupancy of multiphase fluids and stress-dependent permeability of solid matrices using fully coupled models involving mesh-fitted approaches such as arbitrary Lagrangian-Eulerian (ALE) and non-fitted methods such as immersed boundary method (IBM).
-
Development of multiphase resolved and unresolved CFD-DEM coupling for fluid-solid interaction in discrete particles.
Multiphase Flow FSI
Fagbemi et al. (2018)
Artificial Intelligence for Geosystems
Fagbemi et al. (under review)
Artificial intelligence is applied for various scientific and engineering applications such as computational fluid dynamics, predicting material properties, and enhancing digital images of materials:
-
AI for advancing pore-scale multiphase flow simulation efforts in hydrological systems, microfluidics, etc.
-
Mapping complex process-structure-property relationships for various materials, geomaterials, etc., using advanced imaging analysis and AI.
-
Image super-resolution and segmentation using Generative Adversarial Networks (GAN).
-
Development of various image registration techniques using hybrid MPI-GPU systems that are portable on large computer clusters.
Image Enhancement
Image super-resolution for multi-scale CT imaging for image enhancement, image super-resolution:
-
Apply different network architectures such as GAN to increase the resolution of images using perceptual loss functions.
-
Apply resolution increase for up to 6X the original resolution with the aim of retaining and improving pore throat connections.
-
Implementing codes in optimized Python-based machine learning models such as TensorFlow and PyTorch to train and validate thousands of 2D and 3D images.
Fagbemi et al. (under review)