Microfluidic Hydrodynamic of Gas-Liquid flow in Single Microchannel and Porous Media with Microchannel Network

Abstract: In this thesis, a microfluidics platform with high-speed imaging system was built to investigate gas-liquid flow in single microchannel and interfacial instability in porous media with microchannel network:The mass transfer of slug flow in the rectangular and square microchannels was experimentally studied by using water as liquid phase and CO2 as gas phase. Depending on flow rates, flow patterns including slug flow, bubbly flow, and annular flow were observed in rectangular and square microchannels. Flow pattern map was proposed and compared with the maps in the literatures. By using digital image processing, the bubble volume especially deformed bubbles in rectangular and square microchannels was calculated based on 2D projection and 3D slicing, correspondingly. Scaling laws including important parameters of bubbles were derived to provide the guidance of microreactor design. Mass transfer coefficients were calculated based on bubble volume. The empirical correlations involving dimensionless numbers were fitted to precisely predict mass transfer coefficients. Further, to be universality, a semi-theoretical model considering length ratio of liquid and gas phases was developed topredict measured mass transfer coefficients in square microchannel precisely.The gas-liquid displacement in porous media with microchannel network was experimentally investigated. By varying capillary numbers Ca and viscosity ratios M in a wide range, flow pattern involving viscous fingering (VF), capillary fingering (CF) and crossover zone (CZ) can be observed. Finger morphologies at breakthrough moment and steady state in three different flow regions was visualized. The main difference between VF and CF is that the gas stops invading in CF region after breakthrough, whereas in VF region gas can continue to expand until almost all the liquid phase is displaced. Invasion velocity, phase saturation and fingering complexity were quantified based on digital image processing. Fingering dynamical behaviors in different flow pattern before and after breakthrough was investigated. Time evolution of fingering displacement after breakthrough demonstrated an unobserved circle, consisting of new finger generation, cap invasion, breakthrough and finger disappearance. The circle repeats until steady state. Finally, local dynamical invasion behavior was studied and a stepwise way of gas invasion was exposed.