| dc.description.abstract | In the past, the technology to capture a high-resolution image was not mature enough to enable pore-scale modeling and simulation. Additionally, including sufficient pore geometry complexities leads to a heavy computational cost. However, as high-resolution image of a pore structure and high performance computing become more availble nowadays, pore-scale simulation is emerging as a powerful tool to predict the flow properties in a porous media.
One of the most common fluid flow simulation method at the pore-scale is Lattice Boltzmann Modeling (LBM). In this method, the fluid is represented with its particle distribution. The fluid particle distribution undergoes two main processes: streaming and collision. streaming part depicts the movement of the particle distribution between lattices. Collision part depicts the relaxation of the particle distribution towards local equilibrium due to particle collision. The LBM method has many advantages: 1) it can be used directly in a digital pore structure, 2) the implementation is simple and straightforward, and 3) well-suited for parallel computing. For the thesis, we are using the open source software LBPM. The software uses D3Q19 lattice system and boundary condiitons as described by McClure et al. (2018). The method is extended with colour gradient method to model immiscible two-phase flow.
The focus of the thesis is to observe the first-order effect of LBM parameters to the fluid distribution and flow properties, and how they are correlated with experimental studies. Drainage and imbibition process in a microfluidic experiment are simulated. We are particularly interested in observing the typical displacement mechanism that occur in the simulation, with regards to its viscous and capillary forces. | |
