ABSTRACT
We study the interfacial evolution of immiscible two-phase flow within a capillary tube in the partial wetting regime using direct numerical simulation. We investigate the flow patterns resulting from the displacement of a more viscous fluid by a less viscous one under a wide range of wettability conditions. We find that beyond a wettability dependent critical capillary number, a uniform displacement by a less viscous fluid can transition into a growing finger that eventually breaks up into discrete blobs by a series of pinch-off events for both wetting and nonwetting contact angles. This study validates previous experimental observations of pinch-off for wetting contact angles and extends those to nonwetting contact angles. We find that the blob length increases with the capillary number. We observe that the time between consecutive pinch-off events decreases with the capillary number and is greater for more wetting conditions in the displaced phase. We further show that the blob separation distance as a function of the difference between the inlet velocity and the contact line speed collapses into two monotonically decreasing curves for wetting and nonwetting contact angles. For the phase separation in the form of pinch-off, this work provides a quantitative study of the emerging length and timescales and their dependence on the wettability conditions, capillary effects, and viscous forces.
ABSTRACT
HYPOTHESES: The interfacial dynamics in natural porous media are affected not only by the interplay between viscous and capillary forces but also the solid surface wettability. It has been hypothesized that the wettability alteration induced by changes in the water salinity is primarily caused by electric double-layer force expansion, which strongly affects the multiphase flow dynamics. SIMULATIONS: We investigate the effect of water ionic composition and surface roughness on pore-scale wettability alteration. Multiphase hydrodynamics is numerically captured by a lubrication approximation describing the evolution of thin-films coupled with a multiscale level-set approach. An oil blob mobilized by water within a single pore is considered as a case study. The effect of brine ionic composition is accounted for by an electric double-layer through the water ionic strength and zeta-potential parameters. FINDINGS: We demonstrate that high-salinity water thin-films collapse to an adsorbed nanometer layer, leading to a large pressure drop during mobilization of the blob induced by the attractive surface forces. However, low-salinity water thin-films are stable due to the repulsive electric double-layer forces, leading to less pressure drop during mobilization of the blob. The novelty of this work lies in efficiently capturing the nanoscale effects of the electric double-layer in pore-scale multiphase flow displacements. Our quantitative investigations provide fundamental insights into the efficiency of low-salinity waterflooding.
ABSTRACT
As shown in computer vision, the power of deep learning lies in automatically learning relevant and powerful features for any perdition task, which is made possible through end-to-end architectures. However, deep learning approaches applied for classifying medical images do not adhere to this architecture as they rely on several pre- and post-processing steps. This shortcoming can be explained by the relatively small number of available labeled subjects, the high dimensionality of neuroimaging data, and difficulties in interpreting the results of deep learning methods. In this paper, we propose a simple 3D Convolutional Neural Networks and exploit its model parameters to tailor the end-to-end architecture for the diagnosis of Alzheimer's disease (AD). Our model can diagnose AD with an accuracy of 94.1% on the popular ADNI dataset using only MRI data, which outperforms the previous state-of-the-art. Based on the learned model, we identify the disease biomarkers, the results of which were in accordance with the literature. We further transfer the learned model to diagnose mild cognitive impairment (MCI), the prodromal stage of AD, which yield better results compared to other methods.
