Aqueous solutions of hydrolyzed polyacrylamide polymer (HPAM) are now regularly used to support excavation in underground construction, e.g., trenches and pile bores, because of their high viscosity and shear-thinning nature. To ensure excavation stability, it is crucial to maintain the high fluid viscosity and therefore to be able to measure it accurately on site. In this work, two tests commonly used in civil engineering are assessed, i.e. the Fann viscometer and the Marsh funnel. These tests measure viscosity in different ways; the Fann viscometer uses a torque measurement, while the Marsh funnel measures the discharge time for a fixed flow volume. We employed high-resolution computational fluid dynamics (CFD) techniques to simulate both test processes and assessed the extent to which the test predictions by the standard measurement protocols can capture the rheological parameters. The results suggest that the Fann viscometer is reliable at very low shear rates; however, inertial effects lead to an overestimation of apparent viscosity at high rotational velocities. Though a correction factor can be successfully applied to correct Fann viscometer data in the case of Newtonian fluids, it is unlikely that a single correction factor can be used for HPAM fluids. The Marsh funnel time is very sensitive to the infinite-shear-rate viscosity for HPAM fluids. This presents a challenge to its use as an in-situ quality control measure as the infinite-shear-rate viscosity is similar for HPAM with different concentrations. The critical assessment offered in this work supports the development of new experimental approaches for quality control when polymer fluids are used in ground engineering construction projects.

Geological storage of carbon dioxide (CO2) and hydrogen (H2) is a promising strategy for reducing carbon emissions, underscoring the need to optimize the storage performance. While most prior research has focused on single injections in non-deformable media, less attention has been given to cyclic injection regimes and porous systems that deform under injection-induced pressures. This study experimentally investigates the effects of particle deformability (i.e., soft and hard materials) and injection schemes (i.e., single and cyclic injections) on gas trapping behaviors using a quasi-2D Hele-Shaw cell. In soft media, cyclic injections notably enhance gas retention compared to single injection, with trapped gas evolving from large cavities to interconnected ganglia and ultimately to discrete blobs. These trapped clusters reduce local porosity and promote further trapping in subsequent cycles. In contrast, the hard case, where traditional capillary trapping dominates, shows residual saturation remaining independent of injection cycling. Comparisons with the Land model and prior studies indicate that hard granular media follow the predicted hysteresis trajectory, consistent with experiments under ambient pressure. By contrast, soft media exhibit enhanced trapping efficiency, resembling behavior observed under MPa-level pore pressure, suggesting that elevated pressure may induce pore-scale deformation even in nominally rigid rocks. Importantly, a strong correlation between pressure trends and residual saturation suggests that pressure monitoring may serve as a proxy for estimating gas entrapment. These insights provide guidance for optimizing geological gas storage strategies, particularly for improving the efficiency and reliability of H2 withdrawal and long-term CO2 sequestration security.

Hypothesis: Wettability, quantified by the contact angle, is a key property of porous media influencing the capillary pressures, the fluid–solid interfacial area, and eventually reaction and mass transfer processes. Recent advances in imaging enable the direct extraction of contact angles from 3D image data. However, available extraction methods often produce non-physical extreme angles that obscure the true statistics. We suspect that the implementation of physical constrains can filter out the errors accompanied by voxelization and image noise.Experiments: We propose a novel geometrical to physical compliance extraction based on the X-ray imaging (CT) experiments. This model is validated against an ideal geometrical model and compared with reported methods on the same specimens under distinct wettability conditions.Findings: We have demonstrated that our algorithm yields a more physically meaningful and robust measurement of the distribution of contact angles. Each extracted angle matches the ideal geometrical model with a pointwise deviation of ≤ 2°. In real porous systems, our physics-constrained procedure preserves the expected wettability ordering across different conditions while markedly suppressing spurious extreme tails and yielding a tighter central peak, thereby indicating effective removal of non-physical artifacts induced by voxelization and segmentation. We further visualize spatially resolved contact angle fields, revealing the 3D wettability heterogeneity. Moreover, our work yields consistent statistics from small representative volumes to larger domains, thereby enabling seamless upscaling to continuum descriptions and providing robust inputs and potential value for modeling capillarity-driven transport, interfacial area evolution, and interface-controlled mass transfer and reactions in general porous systems.

Polymer fluids are widely used in subsurface and geotechnical engineering applications. While the steady shear rheology of polymer fluids is known to be reasonably captured by a Carreau-like shear-thinning model, it is still not fully understood how their elastic rheological characteristics, beyond shear-thinning behavior alone, influence their flow in porous media. In this study, we numerically investigate these effects using direct, pore-scale numerical simulations. By comparing data from simulations using the FENE-P model, which incorporates viscoelastic effects, with data from corresponding simulations using the Carreau model, which captures only shear thinning, we confirm that fluid elasticity can induce recirculation upstream of restrictions, leading to a reduction in polymer fluid conductance in porous media. As this recirculation is controlled by the geometric conditions, we conducted detailed comparisons between a two-dimensional model, a three-dimensional model mimicking microfluidics experiments, and an axisymmetric model, analogous to a constricted capillary tube. We also simulate flow in an ordered packing of uniform spheres to develop an understanding of the implications for flow in a 3D porous material. We find that these flows are regulated by the interplay between shear-thinning and elasticity effects. When the shear-thinning effect is sufficiently strong, the effects of elasticity are suppressed. In subsurface applications, viscoelastic effects are significant due to pore-scale confinement and fluid rheology itself, requiring explicit consideration in modeling, pilot design, and performance forecasting.

Polymer fluids present a number of advantages over bentonite-based alternatives for excavation support. Much is still to be learned about the permeation of these non-Newtonian fluids in soils, especially more permeable soils, and there is uncertainty around the support forces imparted to the soil. This study develops a numerical framework based on a pore-network modeling approach to quantify the grain drag forces induced by polymer flows in sand packings. The model captures the shear-thinning rheology of polymer fluids, and the emergent data reveal the distinct flow dynamics and their influence on grain-scale interactions. The shear-thinning rheology of these fluids amplifies flow heterogeneity, leading to a nonlinear relationship between individual grain drag and the total applied pressure gradient if it exceeds a critical threshold. Implications of these fundamental observations for engineering practice are demonstrated through an example application to excavation support during diaphragm wall construction.

Spontaneous droplet movement has gained increased interest in many applications, including microfluidics and microfabrication. This study focuses on the numerical investigation of driving mechanisms of spontaneous droplet motion. The numerical model using the phase-field method was validated by available experimental data. In this study, a heterogeneous wettability condition is implemented to reproduce contact angle hysteresis for the accurate prediction of spontaneous droplet dynamics. Through analysing the capillary pressure within the droplet, the driving mechanism is identified as being governed by the pressure difference between the two interfaces which depends on channel configuration, wettability, and contact angle hysteresis. The impact of channel deformability was further studied, revealing that channel deformability leads to significant changes in velocity or even reversed droplet movement direction. This study provides a novel numerical framework for controllable spontaneous droplet movement in flexible channels.

Density-driven segregation, extensively studied in a simple rotating drum, is enriched with a wide range of underlying physics. Diverse symmetrical segregation patterns formed by mixing two types of dry mono-sized grains have been revealed due to variations in heavy and light grain densities rho h and rho l and rotating speeds omega. We engender experimentally a nearly complete segregation, not occurring in dry conditions of the same rho h, rho l, and omega, in submerged states. Furthermore, based on the experiment-validated simulations, using coupled computational fluid dynamics and the discrete element method, it is found the mixing index can be well predicted over a wide parameter space in the effective density ratio, D = (rho h - rho f)/(rho l - rho f) with rho f being the fluid density. Specifically, with increasing D well-mixed states transit to fully segregated states with a rising number of vortices and more severe asymmetrical patterns. When the global Reynolds number Reg is enlarged, the vortex area of heavy particles shrinks for lower D, while the area of light particles gradually saturates; meanwhile, for higher D a new vortex with a continuously expanded area can be encountered in the light particle zone. These results improve our understanding of segregation transitions, especially in submerged granular systems, and shed new light on various science and engineering practices.

Cycle injection schemes are often encountered in underground hydrogen storage (UHS), and the involved hysteresis directly impacts storage and extraction efficiency. The geological formation generally has hierarchical features containing multiple‐level pore sizes. Nevertheless, we still lack a comprehensive understanding of this phenomenon and the pore‐scale mechanism behind the geometry affects saturation hysteresis and its cyclic responses. In this work, by 3D printing technology, we fabricated a hierarchically structured porous media with dual permeability and uniform one for comparison. Gas‐liquid injection cycles were performed to investigate the impact of hierarchical structure on invasion behavior. The phase morphology shows the preferential invasion in 1st‐order structure and the capillary trapping in 2nd‐order structure, which are supported by the phase saturation at each level of the hierarchical structure. Furthermore, ganglion motion is suppressed in the hierarchical structure. Through analyzing local invasion behaviors, the connect‐jump invasion mode is identified as the primary reason for this suppression. Then, the hysteresis effect was quantified based on the Land model, revealing a weaker hysteresis effect in the hierarchical structure compared with the uniform structure, indicating that the hierarchical structure has a lower storage and extraction efficiency in UHS. Finally, the upward trend of relative permeability with saturation was fitted by the van Genuchten model. The model parameter in the hierarchical structure is higher than that in the uniform structure, which is caused by extra pore space in 2nd‐order structure. The findings improve the understanding of hysteresis effect and can promote optimizing strategies for storage and extraction in UHS.

Non-Newtonian fluid flows in porous media are critical in various subsurface and geotechnical engineering applications. However, accurately predicting such flows remains challenging due to the complex fluid rheology and intricate pore structures. This study focuses on polymer fluids with shear-thinning rheology with the motivation of advancing understanding of polymer support fluids for ground engineering applications. To address the limitations of existing models, we derive a theoretical conductance model for polymer flow in a capillary tube, based on which a customized pore-network method is developed. Our simulations reveal three distinct flow regimes, highlighting the impact of the rheology on flow dynamics. Notably, flow heterogeneity amplifies as the shear-thinning feature directs more flow through wider pores, where the effective viscosity decreases more significantly compared to narrower ones. A generalized Darcy's law is formulated for non- Newtonian fluids, validated through pore-network modeling on 60 distinct sphere packings. The proposed framework is adaptable to abroad range of non-Newtonian fluids, offering valuable insights for scaling up to field-scale applications.

Spontaneous and directional droplet transport has attracted considerable attention for its potential applications such as self-cleaning surfaces and microfluidics. Previous studies on the droplet motion between rigid or parallel flexible channels established that the movement direction can be predetermined by the initial channel configuration and wettability. However, in this study, we experimentally show that the direction of droplet movement in nonparallel deformable channels can be carefully tuned, even under the same geometric and wetting conditions. This advanced feature arises from the interplay between initial system conditions and the subsequent solid-liquid interaction, depending on the initial channel configuration, channel flexibility, droplet conditions, wettability, and contact angle hysteresis. Combining the capillary pressure determined from the Young-Laplace equation with the structural deformation described by Euler-Bernoulli beam theory, we developed a general mathematical model capable of accurately predicting the droplet movement direction under a wide range of conditions. The model does not need any fitting parameters and is validated by our experiments. Additionally, our results lead to the identification of a phase diagram encompassing three distinct modes of droplet movement: toward the free end, toward the fixed end, and a stationary state. The proposed phase diagram serves as a predictive tool, offering new insights into controlling the direction of spontaneous droplet motion in flexible channels, with applications in the design and optimization of microfluidic devices.

Preventing excavations from collapsing before inserting the permanent structure is essential in many geotechnical projects. Support fluids, most commonly comprising suspensions of bentonite clay, are one way to achieve this support. Compared to bentonite slurries, support systems that use polymer fluids are economical and have a smaller environmental impact. Currently, a poor understanding of how these support fluids achieve support hinders the broad application of these materials in ground engineering. One key issue is quantifying the fluid-particle interaction force, which ultimately contributes to the overall excavation support. This research exploits Computational Fluid Dynamics (CFD) in a detailed study of the fluid-particle interaction. The results show that the fluid-particle interaction forces for polymer support fluids may be up to 10 4 times that for water at very low seepage velocities. It gives an insight into the support mechanism using polymer support fluids.

Support fluids are widely used to provide temporary support to open excavations. Polymer support fluids have advantages over conventional bentonite slurries in many situations; they are cheaper and can have a lower environmental impact. This work employs direct numerical simulations to investigate the flow of a shear-thinning polymer support fluid whose rheology can be described by the Carreau model in an idealized granular material comprising spheres on a lattice packing. We show that particle radii can be calibrated to simulate a permeation behavior similar to that observed in physical permeability tests on sand specimens. The numerical results indicate that the seepage force induced by polymer flow is up to four orders of magnitude higher than that by water at a low seepage rate, providing a physical explanation of how polymer fluids support excavation in civil engineering.

Carbon dioxide (CO2) and hydrogen (H2) storage in geological formations are two key approaches to reducing carbon emissions, with capillary trapping serving as one of the primary trapping mechanisms. Understanding the behaviour of immiscible fluid-fluid displacement in porous media is crucial for optimizing trapping efficiency. However, previous studies have primarily focused on single injection scenarios in non-deformable particles, with limited attention given to cyclic scenarios and conditions where the porous media may deform under injection-induced stress. To address this gap, this study experimentally investigates the effects of solid deformability (i.e., hard and soft particles) on trapping behaviours during single and cyclic injections under quasi-2D conditions using a Hele-Shaw cell. In soft media, gas bubbles evolve from cavities to small blobs during cyclic injections, leading to a noticeable increase in residual saturation compared to single injection. In contrast, hard granular media exhibit pore invasion from the onset, with residual saturation remaining independent of the number of injection cycles. These findings demonstrate that solid deformability plays a critical role in governing the dependence of residual trapping on injection schemes. These insights offer valuable guidance for developing more effective and optimized strategies for geological gas storage.

Controversy exists regarding whether grain morphology reduces or enhances the drag of a single grain in creeping flows; further complication occurs when orientation dependence for aspherical grains comes into play. To quantify influences of shape irregularity and orientation, this study numerically investigates the drag in creeping flows for fractally rough grains depicted by Spherical Harmonics. As the grain surface becomes more angular with increasing relative roughness and fractal dimension, a stronger drag reduction is observed; this observed morphology-dependent reduction indicates Stokes formula is insufficient. The rotational dependence helps to explain the paradox wherein drag enhancement is always encountered insettling-grain experiments popular in geophysics. For the drag of a given rough grain at various orientations, extracted distributions of the overall drag, and its two components, i.e., skin friction and pressure drag, universally adhere to the Weibull patterns. Moreover, we identify a universal power law between drag coefficients and a newly proposed area-related number, sufficiently taking into consideration both grain roughness by surface area and orientational dependence by projected area perpendicular to the flow direction. This law is primarily controlled by relative roughness, with less sensitivity to fractal dimension. The determined expression enables precise estimation of fluid-particle interactions in upscaling simulations.

Carbon dioxide and hydrogen storage in geological formations at Gt scale are two promising strategies toward net-zero carbon emissions. To date, investigations into underground hydrogen storage (UHS) remain relatively limited in comparison to the more established knowledge body of underground carbon dioxide storage (UCS). Despite their analogous physical processes can be used for accelerating the advancements in UHS technology, the existing distinctions possibly may hinder direct applicability. This review therefore contributes to advancing our fundamental understanding on the key differences between UCS and UHS through multi-scale comparisons. These comparisons encompass key factors influencing underground gas storage, including storage media, trapping mechanisms, respective fluid properties, petrophysical properties, and injection scenarios. They provide guidance for the conversion of our existing knowledge from UCS to UHS, emphasizing the necessity of incorporating these factors relevant to their trapping and loss mechanisms. The article also outlines future directions to address the crucial knowledge gaps identified, aiming to enhance the utilisation of geological formations for hydrogen and carbon dioxide storage.

To realize the transition of our society to a low-carbon future with innovative subsurface energy solutions, understanding the dynamic behavior of gas invading multi-fluid systems in underground pore space is critical. In this work, a joint approach of flow imaging and digital image processing is employed to investigate the fingering dynamics of gas invading multi-fluids in porous media. We examined various gas (G) invasion scenarios of a high-viscosity defending liquid (HL), low-viscosity defending liquid (LL), and their co-existing multi-fluid system, focusing on the viscosity effect. Quantification of phase saturation shows that the displacement efficiency follows the order of G→(L→L) > L→L > G→L, regardless of the varieties in injection flow rate in the viscous-dominated flow regime. In other words, the enhancement in displacement efficiency and potential energy savings are achieved by solely introducing a third phase without the cost of the higher pumping power. When gas invades the HL and LL multi-liquid system, the fingering pattern in G→(HL→LL) and G→(LL→HL) significantly differs and highly depends on the sequential occupation of HL and LL in the pore spaces. The previously unobserved yarn-liked gas pattern in G→(LL→HL) is suspected as the main reason for the fast gas displacement. Through Local dynamics analysis, we identified that the preferential invasion into interconnected LL channels and the inhibitory effect of scattered HL on bypass invasion are the primary mechanisms behind the formation of yarn-liked fingers. We classified two distinct categories of ganglia mobilization and connection in G→(LL→HL), i.e. "catch up to connect" and "expand to connect". Finally, the topological connectivity of the gas finger in G→(LL→HL) is evaluated using Euler number. Euler number shows an ascending trajectory before breakthrough, followed by a rapid descent and stabilization at steady state. This signifies that disconnected ganglia emerge before breakthrough and subsequently expand and reconnect. Our new findings are of great importance for subsurface extraction/storage strategy innovation through enriching multi-fluids injection scenarios.

Unfavorable fluid-fluid displacement, where a low-viscosity fluid displaces a higher-viscosity fluid in permeable media, is commonly encountered in various subsurface processes. Understanding the formation and evolution of the resulting interfacial instability can have practical benefits for engineering applications. Using gradient capillary tubes as surrogate models of permeable media, we numerically investigate interfacial dynamics during gas-driven drainage. Our focus is on understanding the impact of tube geometry on interface stability. In a gradient tube, since the interface shape changes during the drainage process, we measure interfacial stability using the difference between the contact-line velocity Ucl and the meniscus tip velocity Utip. We define instability as a rapid reduction in the contact line velocity Ucl compared to the tip velocity Utip. Beyond the onset of this instability, gas penetrates into the liquid, forming a finger, and entraining a liquid film on the tube wall. The observed stability transition can be rationalized to a large extent by adaptation of an existing theory for cylindrical tubes in terms of a critical capillary number Cacrit. For an expanding tube, simulations suggest that a stability transition from an initially unstable meniscus to a final stable one, with Ucl catching up with Utip, can occur if the local capillary number is initially slightly larger than Cacrit and then drops below Cacrit. The insights gained from this study can be beneficial in estimating the mode and efficiency of subsurface fluid displacement. We numerically investigate the dynamics of a gas-liquid interface during drainage in a gradient capillary tube The observations from our numerical simulations can be rationalized by an adapted theoretical model We find a unique stabilization in drainage along expanding tubes, suppressing film entrainment even when the system is initially unstable

The advance of the wetting front follows the Lucas–Washburn kinetics of t 1/2 scaling and the scaling coefficient strongly depends on the wettability conditions.

Due to their simple geometric configuration and involved rich physics, rotating drums have been widely used to elaborate granular flow dynamics, which is of significant importance in many scientific and engineering applications. This study both numerically and experimentally investigates dry and wet mono-dispersed granular flows in a rotating drum, concentrating on the effects of relative densities, rho s-rho f, and rotating speeds, omega. In our numerical model, a continuum approach based on the two-phase flow and mu(1) theory is adopted, with all material parameters calibrated from experimental measurements. It is found that, in the rolling and cascading regimes, the dynamic angle of repose and the flow region depth are linearly correlated with the modified Froude number, Fr*, introducing the relative density. At the pore scale, flow mobility can be characterized by the excess pore pressure, pf. To quantify the variance of the local pf, it is specifically nondimensionalized as a pore pressure number, K, and then manifested as a function of porosity, 1-phi s. We find K(phi s) approximately follow the same manner as the Kozeny-Carman equation, K proportional to phi(2)(s)(1-phi(s))(3). Furthermore, we present the applicability of the length-scale-based rheology model developed by Ge et al. ["Unifying length-scale-based rheology of dense suspensions," Phys. Rev. Fluids 9, L012302 (2024)], which combines all the related time scales in one dimensionless number G, and a power law between G and 1-phi(s)/phi(c) is confirmed. This work sheds new lights not only on the rigidity of implementing continuum simulations for two-phase granular flows, but also on optimizing rotating drums related engineering applications and understanding their underlying mechanisms.