Bulletin of the American Physical Society
75th Annual Meeting of the Division of Fluid Dynamics
Volume 67, Number 19
Sunday–Tuesday, November 20–22, 2022; Indiana Convention Center, Indianapolis, Indiana.
Session T23: Porous Media Flows: General |
Hide Abstracts |
Chair: Samantha McBride, Princeton University; Boris Stoeber, The University of British Columbia Room: 231 |
Monday, November 21, 2022 4:10PM - 4:23PM |
T23.00001: Characterizing flow through artificial deformable porous media Raunak Basak, Boris Stoeber Porous deformable materials are ubiquitous in nature. Fluid flow through any deformable porous medium causes deformation of the medium and in turn, this deformation affects the fluid flow. Flow through deformable porous media is relevant to applications such as soil consolidation, CO2 sequestration, infiltration and transport processes in human tissue including for drug delivery. This work aims to investigate certain aspects of such flow using a purposely designed artificial deformable porous medium. The matrix is made with polydimethylsiloxane using the solvent casting and particulate leaching technique. The setup comprises a tightly sealed chamber and a mechanism to strain the porous matrix and the pressure difference and the resultant flow rate through the matrix are measured. We derive strain-permeability and porosity-permeability relationships from the measurements for comparison with existing models, and we observe that the matrix permeability varies exponentially with strain. The setup also permits observation of flow-induced deformation of the matrix as a function of driving pressure and fluid viscosity. Our ability to manipulate individual parameters of this flow problem allows us to better understand the complexities of poroviscoelastic flow. |
Monday, November 21, 2022 4:23PM - 4:36PM |
T23.00002: Prediction of 3D Velocity and Temperature Field of Reticulated Foams using Deep Learning Danny Ko, Hangjie Ji, Y. Sungtaek Ju Data-driven deep learning models are emerging as a new method to predict the flow and transport behavior of porous media and to drastically reduce the required computational power. Previous deep learning models, however, experience difficulty or require an additional computation to predict the full 3D velocity field which is essential in the characterization at the pore-level and subsequent transport analysis of porous media. In this study, we design a deep learning model and incorporate a physics-informed loss function to relate the spatial information of the 3D binary image to the full 3D velocity field of a porous medium. We demonstrate that our model, trained only with synthetic porous media, can predict the full 3D velocity field of real reticulated foams which have different microstructures from typical sandstones that are studied in previous works. We also show that our loss function is able to enforce the mass conservation in incompressible flow. Our study provides the framework for predicting the full 3D velocity field of porous media and conducting subsequent transport analysis for various engineering applications. As an example, we conduct heat transfer analysis using the predicted velocity field and demonstrate the capability and advantage of our deep learning model in consecutive transport analysis. |
Monday, November 21, 2022 4:36PM - 4:49PM |
T23.00003: A Hybrid Experimental-Numerical Approach to Study the Evolution of Porous Media during Biomineralization. Sina Nassiri, SeyedArmin MotahariTabari, Nariman Mahabadi In this study, pore-scale characteristics of the interactions between biogeochemically induced carbonate precipitation and the accompanying impacts on the hydraulics properties of porous media are investigated by using a hybrid experimental-numerical approach. The experimental studies consist of a two-dimensional transparent microfluidic chip to visualize the biomineralization process at a pore-scale. A reactive solution is flushed in 10 cycles through the microfluidic chip while the process is recorded by time-sequential high-resolution imaging. The fluid flow response in the pore structure due to clogging was traced via dyed water injection and recording the video in the early stages of the fluid entrance to the pore structure and its percolation through the pore network. To obtain tempo-spatial characteristics of velocity fields affected by mineralization in the pore space, Particle Image Velocimetry (PIV) method is adopted by injecting fluorescent micro-sized spheres into the water saturated pore network. The experimental images were analyzed using a developed image processing algorithm to detect carbonate minerals within the pore space. Based on processed images, a finite element CFD model was developed to explore the impact of bioclogging on the hydrodynamics of the porous media including the evolution of pore structure morphology, porosity-permeability relations and of fluid percolation pathways. |
Monday, November 21, 2022 4:49PM - 5:02PM |
T23.00004: Prediction of the velocity distribution in porous media from the pore volume distribution Vi Nguyen, Dimitrios V Papavassiliou The average pore velocity and the local velocity are important in porous media applications so that predicting the velocity distribution enables the prediction of unexpected issues that may occur media. In this study, velocity magnitude and pore size distributions were examined in twelve different porous media with porosity from 13% to 85%, including sandstones, carbonates, synthetic silica, sphere packings and fiber scaffolds. The flow field was simulated by lattice Boltzmann computations, [1,2], while pore networks were extracted by a maximal ball algorithm [3]. The velocity magnitude and pore volume were normalized by the mean velocity magnitude and mean pore volume, respectively. Two-sample Kolgomorov-Smirnov goodness-of-fit tests were performed, and the results showed that the velocity magnitude distribution and pore volume distributions in a porous medium are similar within statistical accuracy. This finding allows the prediction of the velocity distribution in porous media when the pore size distribution is known. |
Monday, November 21, 2022 5:02PM - 5:15PM |
T23.00005: Formation and collapse of gas cavities in a soft porous medium Oliver Paulin, Liam Morrow, Matthew Hennessy, Chris W MacMinn Gas bubbles can form and grow in otherwise liquid-saturated granular media due to various physical processes, such as corrosion or the microbial decomposition of organic matter. The gas bubbles are typically non-wetting to the solid grains; as such, it is energetically costly for the gas to invade the narrow pore throats between grains. If the solid skeleton is sufficiently soft, the gas can instead displace the solid grains to open macroscopic cavities. An increase in external confining stress can suppress the formation of these cavities and even trigger the collapse of existing cavities, forcing the gas into the pore space. Here, we investigate this problem experimentally using a packing of hydrogel beads as a model soft porous medium. We vary the external confining stress to study the formation and collapse of gas cavities within this system. We complement our experimental observations with a phase-field model informed by large-deformation poromechanics. We study the formation and collapse of gas cavities within a 1D setting, identifying the confining stress at which cavities collapse, with gas invading the pore space. We also study the impact of the rate of compression as well as the reversibility of cavity formation and collapse under fluctuating confining stress. |
Monday, November 21, 2022 5:15PM - 5:28PM |
T23.00006: Finite PINN Net: Physics-informed deep convolutional neural networks for learning 3D transient Darcy flows in heterogeneous porous media Mohammad Sarabian, Pablo Ruiz Mataran, Ruben Rodriguez Torrado In this study, we have developed a novel and robust physics-informed deep convolutional neural network called "Finite PINN Net" for simulating multiphase, 3D, transient, and compressible Darcy flow in heterogeneous petroleum reservoirs. In the case of porous media, simulating such flows with neural networks (NNs) in highly heterogeneous reservoirs with source/sink terms is challenging due to the non-linearity of the fluid and rock properties. Moreover, enforcing flux continuity across cell boundaries is not viable via automatic differentiation due to the discontinuity of the pressure between two neighboring cells. We have addressed this issue by enforcing the flux continuity using the finite-volume discretization scheme and two-point flux approximation (TPFA) and embedding it into the NN framework. The structure of the Finite PINN Net and training procedure is analogous to the iterative time propagation of the Euler method. The key finding here is that we show, for the first time, the possibility of simulating 3D transient multiphase flows in highly heterogeneous porous media with NNs without using any labeled data. Finally, we validate the predictions of the Finite PINN Net in spatiotemporally resolved pressure and water saturation by comparing against the traditional simulator. |
Monday, November 21, 2022 5:28PM - 5:41PM |
T23.00007: Net Flow Through Soft Porous Media Generated by Periodic Mean-Zero Pressure Gradient Jacob Stein, Jeffrey Tithof Characterization of fluid transport through soft porous media is crucial to understanding and predicting mass transport in a variety of biological and industrial contexts. Recent studies have shown complex hysteretic behavior in the material deformation of soft porous media, which leads to a variable, nonlinear relationship between fluid flux and pressure gradients such that Darcy's law is insufficient. We add fluorescent microspheres to fluid passing through a bed of densely-packed hydrogel particles and track their motion to simultaneously characterize the fluid flow and porous media deformation. When subjected to asymmetric, mean-zero temporally periodic pressure gradients, a non-zero directional flow is observed through the porous media. We quantify the magnitude of the directional flow as a function of the pressure gradient and time scale of pressure variation. Our observations offer insight into potential mechanisms of fluid transport for numerous biological systems, such as the interstitial spaces of the body and brain tissues. |
Monday, November 21, 2022 5:41PM - 5:54PM |
T23.00008: Mass transport in Channels with Porous Walls Alex J Warhover, Marc A Guasch, Michael F Schatz, Roman O Grigoriev Mass transport between the bulk and a surface arises in a variety of applications ranging from catalytic converters to oxygen transport in the lungs. These settings typically involve channels with porous walls to maximize surface-to volume ratio. Mass flux is controlled by three separate physical processes: transport from the bulk to the channel walls, transport through the pores, and finally absorption/desorption at the surface. By using a combination of simulation and experiments, we investigate transport in the channel and the pores under laminar flow conditions. The impact of porosity on oxygen, carbon dioxide, and water concentrations in a gaseous mixture is quantified for a range of flow speeds, pore fraction, and tortuosity. |
Monday, November 21, 2022 5:54PM - 6:07PM |
T23.00009: Super resolution-assisted pore flow field prediction using neural networks Xu-Hui Zhou, James McClure, Cheng Chen, Heng Xiao Direct pore-scale simulations of fluid flow through porous media are computationally expensive to perform for realistic systems. Previous researches have shown that geometry of the microstructure of porous media can be used to predict the velocity fields therein using neural networks. Such trained neural networks, however, perform poorly for unseen porous media, particularly those with a large degree of heterogeneity. In this work, we propose that incorporating a coarse velocity field into the input of neural networks is an effective method for enhancing the prediction accuracy. The velocity field is simulated on a coarsened mesh with a low computational cost. More importantly, it contains global physics information that it can remedy the ill-conditioning produced by the usage of segmented porous media due to the limitation of GPU memory. Numerical results show that incorporating the coarse velocity field greatly improves the prediction accuracy of the fine velocity field when compared to the prediction based on geometric information alone, especially for the porous media with a large interior vuggy pore space. The feasibility of the method is further demonstrated by testing the trained network on real rocks with non-spherical solid grains and much more complex microstructure. |
Monday, November 21, 2022 6:07PM - 6:20PM |
T23.00010: Dynamics of fluid-driven fractures in the viscous-dominated regime Sri Savya Tanikella, Emilie Dressaire Hydraulic fracturing is a natural and industrial process that creates tensile fractures in rock using |
Monday, November 21, 2022 6:20PM - 6:33PM |
T23.00011: Low-resolution magnetic resonance velocimetry in porous media: comparison with Navier Stokes Sid BECKER, Martin Bruschewski, Sam Flint Magnetic Resonance Velocimetry (MRV) in flows through porous media require resolution to distinguish sub-pore scale flow charactersitics. Less information is lost at smaller a voxel size, but this comes at the cost of incresed measurement effort. At some point there are doiminishing retruns associated with smaller voxel size compared to the additional measurement effort. In this study, experiments using a glycerol-water mixtures under low Re conditions through a regularly periodic porous matrix. The the prorous matrix was a 3-D polymer print with a pore length 5 mm. Measurements were taken a 15 different voxel sizes in the range 0.42 mm to 4.48 mm. Comparison with simulation show excellent agreement, indicating a low MRV measurement error. It is found that in this experiment a voxel size of 0.96 mm (here 20% of the pore scale length) is sufficient. At higher resolution, the volume-averaged results do not change. At lower resolution (a voxel size above 20% pore scale) systematic errors are observed. In summary, this study shows that even with a relatively low measurement resolution, quantitative three-dimensional velocity fields can be ob-tained through porous flow systems with short measurement times and low measurement uncertainty. |
Monday, November 21, 2022 6:33PM - 6:46PM |
T23.00012: Fluid Flow Prediction in Porous Media using Sparse Data and Physics-Informed PointNet Ali Kashefi, Tapan Mukerji For the first time, we predict Stokes flow on pore scale in porous media by integrating sparse scattered labeled data and physics-informed PointNet (PIPN). PIPN is categorized as weakly-supervised deep learning such that its loss function is constructed by residuals of the continuity and linear momentum equations of incompressible flows and the mismatch between the neural network predictions and sparse observations. Compared to regular physics-informed neural networks, PIPN converges with lower resolutions of inquiry points, saving computational costs. PIPN has advantages over physics-informed convolutional neural networks in case of porous media. First, the PIPN input is exclusively the pore space (rather than both the grain and pore spaces), requiring less GPU memory. Second, density of inquiry point distribution can freely vary over the pore space in PIPN, allowing users to represent the pore space geometry and its boundaries smoothly and realistically. Third, PIPN can be conveniently integrated with spatially-unstructured data without any data interpolation. We examine the performance of the proposed PIPN framework by prediction of the velocity fields and consequently permeability of digital rocks with different spatial correlation lengths and porosities. |
Follow Us |
Engage
Become an APS Member |
My APS
Renew Membership |
Information for |
About APSThe American Physical Society (APS) is a non-profit membership organization working to advance the knowledge of physics. |
© 2024 American Physical Society
| All rights reserved | Terms of Use
| Contact Us
Headquarters
1 Physics Ellipse, College Park, MD 20740-3844
(301) 209-3200
Editorial Office
100 Motor Pkwy, Suite 110, Hauppauge, NY 11788
(631) 591-4000
Office of Public Affairs
529 14th St NW, Suite 1050, Washington, D.C. 20045-2001
(202) 662-8700