Advanced nuclear magnetic resonance (NMR) imaging methodologies have been developed to determine porous media properties associated with fluid flow processes. This dissertation presents the development of NMR experimental and analysis methodologies, called NMR probes, particularly for determination of porosity, permeability, and pore-size distributions of porous media while the developed methodologies can be used for other properties. The NMR relaxation distribution can provide various information about porous systems having NMR active nuclei. The determination of the distribution from NMR relaxation data is an ill-posed inverse problem that requires special care, but conventionally the problem has been solved by ad-hoc methods. We have developed a new method based on sound statistical theory that suitably implements smoothness and equality/inequality constraints. This method is used for determination of porosity distributions. A Carr-Purcell-Meiboom-Gill (CPMG) NMR experiment is designed to measure spatially resolved NMR relaxation data. The determined relaxation distribution provides the estimate of intrinsic magnetization which, in turn, is scaled to porosity. A pulsed-field-gradient stimulated-echo (PFGSTE) NMR velocity imaging experiment is designed to measure the superficial average velocity at each volume element. This experiment measures velocity number distributions as opposed to the average phase shift, which is conventionally measured, to suitably quantify the velocities within heterogeneous porous media. The permeability distributions are determined by solving the inverse problem formulated in terms of flow models and the velocity data. We present new experimental designs associated with flow conditions to enhance the accuracy of the estimates. Efforts have been put forth to further improve the accuracy by introducing and evaluating global optimization methods. The NMR relaxation distribution can be scaled to a pore-size distribution once the surface relaxivity is known. We have developed a new method, which avoids limitations on the range of time for which data may be used, to determine surface relaxivity by the PFGSTE NMR diffusion experiment.

Pipeline natural gas is the primary fuel of choice for distributed fuel cell-based applications. The concentration of sulfur in odorized natural gas is about 30 ppm, with acceptable levels being <1 ppm for catalyst stability in such applications. Packed bed technology for desulfurization suffers from several disadvantages including high pressure drop and slow regeneration rates that require large unit sizes. We describe a novel Rapid Temperature Swing Adsorption (RTSA) system utilizing hollow fibers with polymer 'binder', impregnated with high loadings of sulfur selective sorbent 'fillers'. Steam and cooling water can be utilized to thermally swing the sorbent during the regeneration cycles. An impermeable, thin polymer barrier layer on the outside of fiber sorbents allows only thermal interactions with the regeneration media, thereby promoting consistent sorption capacity over repeated cycles. A simplified flow pattern minimizes pressure drop, porous core morphology maximizes sorption efficiencies, while small fiber dimensions allows for rapid thermal cycles.

Up-to-date methods of numerical modelling of random fields were applied to investigate some features of solute transport in saturated porous media with stochastic hydraulic conductivity. The paper describes numerical experiments which were performed and presents the first results.

This paper presents the theoretical background of a phenomenological biphasic material approach at large strains based on the theory of porous media as well as its numerical realization within the context of an adaptive mixed finite element formulation. The study is aimed at the simulation of coupled multiphysics problems with special focus on biomechanics. As the materials of interest can be considered as a mixture of two immiscible components (solid and fluid phases), they can be modeled as saturated porous media. For the numerical treatment of according problems within a finite element approach, weak formulations of the balance equations of momentum and volume of the mixture are developed. Within this context, a generalized Lagrangean approach is preferred assuming the initial configuration of the solid phase as reference configuration of the mixture. The transient problem results in weak formulations with respect to the displacement and pore pressure fields as well as their time derivatives. Therefore special linearization techniques are applied, and after spatial discretization a global system for the incremental solution of the initial boundary value problem within the framework of a stable mixed U/p-c finite element approach is defined. The global system is solved using an iterative solver with hierarchical preconditioning. Adaptive mesh evolution is controlled by a residual a posteriori error estimator. The accuracy and the efficiency of the numerical algorithms are demonstrated on a typical example.

Fibrous porous media are widely used in various industries such as biomedical engineering, textiles, paper, and alternative energy. Often these porous materials are formed into composite materials, using subsequent manufacturing steps, to improve their properties. There is a strong correlation between system performance and the transport and mechanical properties of the porous media, in raw or composite form. However, these properties depend on the final pore structure of the material. Thus, the ability to manufacture fibrous porous media, in raw or composite forms, with an engineered structure with predictable properties is highly desirable for the optimization of the overall performance of a relevant system. To date, the characterization of the porous media has been primarily based on reverse design methods i.e., extracting the data from existing materials with image processing techniques. The objective of this research is to develop a methodology to enable the virtual generation of complex composite porous media with tailored properties, from the implementation of a fibrous medium in the design space to the simulated coating of this media representative of the manufacturing space. To meet this objective a modified periodic surface model is proposed, which is utilized to parametrically generate a fibrous domain. The suggested modeling approach allows for a high-degree of control over the fiber profile, matrix properties, and fiber-binder composition. Using the domain generated with the suggested geometrical modeling approach, numerical simulations are executed to simulate transport properties such as permeability, diffusivity and tortuosity, as well as, to directly coat the microstructure, thereby forming a complex composite material. To understand the interplay between the xxiii fiber matrix and the transport properties, the morphology of the virtual microstructure is characterized based on the pore size, chord length and shortest path length distributions inside the porous domain. In order to ensure the desired properties of the microstructure, the fluid penetration, at the micro scale, is analyzed during the direct coating process. This work presents a framework for feasible and effective generation of complex porous media in the virtual space, which can be directly manufactured.

A distinctive field in the coatings industry is the coating of porous media, with broad applications in paper, apparel, textile, electronics, bioengineering, filtration and energy sector. A primary industrial scale process that can be used to coat porous media in a fast and flexible manner is slot die extrusion. A major concern when coating porous media with a wetting fluid is fluid penetration into the substrate. Although some level of penetration is desirable to obtain specific material properties, inadequate or excessive fluid penetration can negatively affect the strength, functionality or performance of the resulting material. In spite of its apparent industrial importance, limited modeling and experimental work has been conducted to study fluid penetration into porous media during fabrication. The effects of processing parameters on the penetration depth, the effects of penetration on material quality, and the method to predict and control the penetration depth are not well understood. This dissertation is composed of two parts. Part I is an applied study for coating onto porous media. This part focuses on the first objective of this dissertation which is to elucidate clearly the feasibility, advantages and disadvantages of the direct coating method as a potential fabrication route for membrane electrode assembly (MEA). MEA samples are fabricated using both traditional and the direct coating methods. Then, the quality and performance of the MEA samples are examined. Experimental results in Part I demonstrate that it is feasible to fabricate MEAs using the direct coating method. However, Nafion® solution penetrates into the catalyst layer during the coating process and causes lower performance of fuel cells, which is the motivation for Part II of this thesis. The objective of Part II is to fundamentally understand the fluid penetration process and predict the penetration depth when directly coating porous media, using a comprehensive approach. A series of computational and analytical models are developed to predict the penetration depth for both Newtonian and non-Newtonian fluids with or without capillary pressure. Finally the accuracy of developed models are validated through experiments. The relative error between the predicted and experimentally measured penetration depth is generally lower than 20%.

With a number of advantages hitherto unrecognized, nanoparticle-stabilized emulsions and foams have recently been proposed for enhanced oil recovery (EOR) applications. Long-distance transport of nanoparticles is a prerequisite for any such EOR applications. The transport of the particles is limited by the degree to which the particles are retained by the porous medium. In this work, experiments that quantify the retention and provide insight into the mechanisms for nanoparticle retention in porous media are described. Sedimentary rock samples (Boise sandstone and Texas Cream limestone) were crushed into single grains and sieved into narrow grain size fractions. In some cases, clay (kaolinite or illite) was added to the Boise sandstone samples. These grain samples were packed into long (1 ft – 15 ft) slim tubes (ID = 0.93 cm) to create unconsolidated sandpack columns. The columns were injected with aqueous dispersions of silica-cored nanoparticle (with and without surface coating) and flushed with brine. The nanoparticle effluent concentration history was measured and the nanoparticle recovery was calculated as a percentage of the injected nanoparticle dispersion. Fifty experiments were performed in this fashion, varying different experimental parameters while maintaining others constant to allow direct comparisons between experiments. The parameters analyzed in this thesis are: specific surface area of the porous medium, lithology, brine salinity, interstitial velocity, residence time, column length, and temperature. Our results indicate that retention is not severe, with an 8% average of the injected amount, for all our experiments. From the parameters analyzed, specific surface area was the most influential variable, with a linear effect on nanoparticle retention independently of lithology. Salinity increased nanoparticle retention slightly and delayed nanoparticle arrival. Velocity, residence time and length are coupled parameters and were studied jointly; they had a minor effect on retention. Temperature had a marginal effect, as we observed an approximate 2% increase in retention at 80°C compared to 21°C. Both surface coated and bare silica nanoparticles were successfully transported, so surface coating does not appear to be a prerequisite for transport for the particle and rock systems studied. / text

Pile foundations that support offshore structures are required to resist not only static loading, but also dynamic loading from waves, wind and earthquakes. The purpose of this study is to gain a better understanding of the behavior of offshore piles under cyclic or dynamic loading using the finite element approach. To achieve this goal, an appropriate constitutive model is required to simulate the behavior of soils and interfaces. The DSC constitutive model is developed for saturated interfaces to study the behavior under severe shear deformation at the soil-pile interface. Monotonic and cyclic simple shear experiments are conducted on Ottawa sand-steel interfaces under drained and undrained conditions using the Cyclic-Multi-Degree-of-Freedom shear device with porewater pressure measurement (CYMDOF-P). The effect of various parameters such as normal stress, surface roughness of steel, type of loading, and the amplitude and frequency of the applied displacement in two-way cyclic loading are investigated. The data from the simple shear tests on saturated interfaces are used to calculate the parameters in the DSC model. The resulting parameters are then used to verify the DSC model by back predicting tests from which parameters are determined and independent tests that are not used in parameters determination. The model predictions, in general, were found to provide a highly satisfactory correlation with the observations. In the context of DSC, the concept of critical disturbance is developed to identify initiation of liquefaction in saturated Ottawa sand-steel interfaces. This method is based on using microstructural changes in material as an indication of liquefaction identification. The finite element method, along with DSC constitutive model, is used to investigate the response of offshore piles to dynamic loading. These include cyclic loading of axially loaded instrumented pile in clay and full-scale laterally loaded pile in sand. The DSC model is used to model the nonlinear behavior of saturated soils and interfaces. A nonlinear dynamic finite element program DSC-DYN2D based on the DSC modeling approach and the theory of nonlinear porous media is used for this purpose. Results from numerical solutions are compared with field measurements. Strong agreement between numerical predictions and field measurements are an indication of the ability to solve challenging soil-structure interaction problems.Based on the results of this research, it can be stated that the finite element-DSC model simulation allows realistic prediction of complex dynamic offshore pile-soil interaction problems, and is capable of characterizing behavior of saturated soils and interfaces involving liquefaction.

Interfacial partitioning tracer tests (IPTT) are one method available for measuring air-water interfacial area (A(ia)).This study used the standard approach comprising tracer injection under steady unsaturated-flow conditions with a uniform water-saturation distribution within the column. Sodium dodecylbezene sulfonate (SDBS) and pentafluorobenzoic acid (PFBA) were used as the partitioning and nonreactive tracers, respectively. Three types of porous media were used for the study: a sandy soil, a well-sorted sand, and glass beads. Initial water saturations, S(w), were approximately 80%, 80%, and 26 % for the soil, sand, and glass beads, respectively. Water saturation was monitored gravimetrically during the experiments. The maximum interfacial areas (A(ia)/(1-S(w))) calculated from the results of the experiments are compared among the three porous media used in this work, and compared to previous air-water interfacial area studies.

We investigate how the formation of heterogeneity and structures in flows through explicit porous microstructures depends upon the geometric and topological observables of the porous medium. Using direct numerical simulations of single-phase, isothermal, laminar fluid flow through realistic three-dimensional stochastically generated pore structures, hereafter referred to as pore spaces, the characteristics of the resulting steady state velocity fields are related to physical characteristics of the pore spaces. The results suggest that the spatially variable resistance offered by the geometry and topology of the pore space induces a highly heterogeneous fluid velocity field therein. Focus is placed on three different length scales: macroscopic (cm), mesoscopic (mm), and microscopic (microns). At the macroscopic length scale, volume averaging is used to relate porosity, mean hydraulic radius, and their product to the permeability of the pore space. At the mesoscopic scale, the effect of a medium's porosity on fluid particle trajectory attributes, such as passage time and tortuosity, is studied. At the final length scale, that of the microscopic in-pore fluid dynamics, finite time Lyapunov exponents are used to determine expanding, contracting, and hyperbolic regions in the flow field, which are then related to the local structure of the pore space. The results have implications to contaminant transport, mixing, and how chemical reactions are induced at the pore-scale. A description of the adopted numerical methods to simulate flow and generate the pore space are provided as well.