Browsing by Author "Blois, Gianluca"
Now showing 1 - 2 of 2
- Results Per Page
- Sort Options
Item A particle-based image segmentation method for phase separation and interface detection in PIV images of immiscible multiphase flow(IOP Publishing, 2021-06) Li, Yaofa; Blois, Gianluca; Kazemifar, Farzan; Christensen, Kenneth TParticle image velocimetry (PIV) is a valuable tool for experimentally studying multiphase flows. In order to distinguish the flow dynamics of each individual phase, proper image segmentation must be performed. In immiscible multiphase flows, the fidelity of phase segmentation is directly linked to the accuracy of the interface detection. This paper reports a novel method for robust phase separation and phase boundary identification applicable to particle images acquired via PIV. The method, which requires a seeding density differentiation between the phases, is based on particle detection and triangular meshing. In this method, tracer particles in all seeded phases are first identified, and the coordinates of the particle locations are then used to formulate a 2D unstructured mesh, where the triangular grids provide the basis for phase separation and the outer edges provide a basis for interface detection. This paper presents a parametric analysis on synthetic particle images to assess the performance of the method and to compare the results to existing approaches. In addition, an application to experimentally generated images is reported. These results show that this method can successfully track the complex interface evolution in a particularly challenging flow system consisting of immiscible multiphase flow in porous media.Item Pore-Scale Dynamics of Liquid CO2–Water Displacement in 2D Axisymmetric Porous Micromodels Under Strong Drainage and Weak Imbibition Conditions: High-Speed μPIV Measurements(Frontiers Media SA, 2021-08) Li, Yaofa; Blois, Gianluca; Kazemifar, Farzan; Molla, Razin S.; Christensen, Kenneth T.Resolving pore-scale transient flow dynamics is crucial to understanding the physics underlying multiphase flow in porous media and informing large-scale predictive models. Surface properties of the porous matrix play an important role in controlling such physics, yet interfacial mechanisms remain poorly understood, in part due to a lack of direct observations. This study reports on an experimental investigation of the pore-scale flow dynamics of liquid CO2 and water in two-dimensional (2D) circular porous micromodels with different surface characteristics employing high-speed microscopic particle image velocimetry (μPIV). The design of the micromodel minimized side boundary effects due to the limited size of the domain. The high-speed μPIV technique resolved the spatial and temporal dynamics of multiphase flow of CO2 and water under reservoir-relevant conditions, for both drainage and imbibition scenarios. When CO2 displaced water in a hydrophilic micromodel (i.e., drainage), unstable capillary fingering occurred and the pore flow was dominated by successive pore-scale burst events (i.e., Haines jumps). When the same experiment was repeated in a nearly neutral wetting micromodel (i.e., weak imbibition), flow instability and fluctuations were virtually eliminated, leading to a more compact displacement pattern. Energy balance analysis indicates that the conversion efficiency between surface energy and external work is less than 30%, and that kinetic energy is a disproportionately smaller contributor to the energy budget. This is true even during a Haines jump event, which induces velocities typically two orders of magnitude higher than the bulk velocity. These novel measurements further enabled direct observations of the meniscus displacement, revealing a significant alteration of the pore filling mechanisms during drainage and imbibition. While the former typically featured burst events, which often occur only at one of the several throats connecting a pore, the latter is typically dominated by a cooperative filling mechanism involving simultaneous invasion of a pore from multiple throats. This cooperative filling mechanism leads to merging of two interfaces and releases surface energy, causing instantaneous high-speed events that are similar, yet fundamentally different from, burst events. Finally, pore-scale velocity fields were statistically analyzed to provide a quantitative measure of the role of capillary effects in these pore flows.