Made available in DSpace on 2014-10-09T12:44:01Z (GMT). No. of bitstreams: 0 / Made available in DSpace on 2014-10-09T14:07:20Z (GMT). No. of bitstreams: 1 06913.pdf: 2715369 bytes, checksum: 9d927e16226a25d1d362ba0ebc83502c (MD5) / Tese (Doutoramento) / IPEN/T / Instituto de Pesquisas Energeticas e Nucleares - IPEN/CNEN-SP

Submitted by Claudinei Pracidelli (cpracide@ipen.br) on 2016-06-22T11:21:31Z No. of bitstreams: 0 / Made available in DSpace on 2016-06-22T11:21:31Z (GMT). No. of bitstreams: 0 / O tratamento de câncer possui diversas modalidades. Uma delas é a utilização de fontes de radiação como principal protagonista do tratamento. A radioterapia e a medicina nuclear são exemplos desse tipo de tratamento. Por utilizarem a radiação ionizante como principal ferramenta para a terapia, há a necessidade de se efetuar diversas simulações do tratamento a fim de maximizar a dose nos tecidos tumorais sem ultrapassar os limites de dose nos tecidos sadios circunvizinhos. Os sistemas utilizados na simulação desses tipos de terapia recebem o nome de Sistemas de Planejamento Dosimétrico. A medicina nuclear e a radioterapia possuem seus próprios sistemas de planejamento dosimétricos devido a grande diversidade das informações necessárias às suas simulações. Os sistemas de planejamento em radioterapia são mais consolidados do que os de medicina nuclear e por tal motivo um sistema que aborde tanto os casos de radioterapia como de medicina nuclear contribuiria para significativos avanços na área de medicina nuclear. Dessa forma, o objetivo do trabalho foi modelar um Sistema de Planejamento Dosimétrico com o uso do código de Monte Carlo MCNP6 Monte Carlo N-Particle Transport Code que permitisse incorporar os casos de radioterapia e medicina nuclear e que fosse extensível a novos tipos de tratamentos. A modelagem desse sistema resultou na construção de um Framework, orientado a objetos, nomeado IBMC o qual auxilia no desenvolvimento de sistemas de planejamento que necessitam interpretar grandes quantidades de informações com o objetivo de escrever o arquivo base do MCNP6. O IBMC permitiu desenvolver de maneira rápida e prática sistemas de planejamento para radioterapia e medicina nuclear e os resultados foram validados com sistemas já consolidados. Ele também mostrou alto potencial para desenvolver sistemas de planejamento de novos tipos de tratamentos que utilizam a radiação ionizante. / Tese (Doutorado em Tecnologia Nuclear) / IPEN/T / Instituto de Pesquisas Energeticas e Nucleares - IPEN-CNEN/SP

Doctor of Philosophy / Department of Biological & Agricultural Engineering / Ronaldo G. Maghirang / Nanostructured particles possess unique chemical and physical properties, making them excellent candidates for air purification, smoke clearing, and obscuration. This research was conducted to investigate the aerodynamic, charging, and infrared (IR) extinction properties of nanostructured particles. Specific objectives were to: (1) measure the size distribution and concentration of aerosolized nanostructured particles; (2) evaluate their IR extinction properties; (3) determine their relative chargeability; and (4) numerically model their transport in enclosed rooms. The size distribution and concentration of two nanostructured particles (NanoActive® MgO and MgO plus) were measured in an enclosed room. The particles differed in size distribution and concentration; for example, the geometric mean diameters of NanoActive® MgO and MgO plus were 3.12 and 11.1 [Mu]m, respectively. The potential of nanostructured particles as IR obscurants was determined and compared with other particles. Four groups of particles were considered: nanostructured particles (NanoActive® MgO plus, MgO, TiO[subscript2]); nanorods (MgO, TiO[subscript2]); conventional particles (NaHCO[subscript3] and ISO fine test dust); and common obscurants (brass, graphite, carbon black). The extinction coefficients of the nanostructured particles were generally significantly smaller than those of the other particles. Graphite flakes had the greatest mass extinction coefficient (3.22 m[superscript2]/g), followed by carbon black (1.72 m[superscript2]/g), and brass flakes (1.57 m[superscript2]/g). Brass flakes had the greatest volume extinction coefficient (1.64 m[superscript2]/cc), followed by NaHCO[subscript3] (0.93 m[superscript2]/cc), and ISO fine test dust (0.91 m[superscript2]/cc). The relative chargeability of nanostructured particles was also investigated. Selected particles were passed through a Teflon tribocharger and their net charge-to-mass ratios were measured. Tribocharging was able to charge the particles; however, the resulting charge was generally small. NanoActive® TiO[subscript2] gained the highest net charge-to-mass ratio (1.21 mC/kg) followed by NanoActive® MgO (0.81 mC/kg) and ISO fine test dust (0.66 mC/kg). The transport of NanoActive® MgO plus and hollow glass spheres in an enclosed room was simulated by implementing the discrete phase model of FLUENT. In terms of mass concentrations, there was reasonable agreement between predicted and measured values for hollow glass spheres but not for NanoActive® MgO plus. In terms of number concentration, there was large discrepancy between predicted and measured values for both particles.

L’algorithme Adaptive Multilevel Splitting (AMS) a récemment fait son apparition dans la littérature de mathématiques appliquées, en tant que méthode de réduction de variance pour la simulation Monte Carlo de chaı̂nes de Markov. Ce travail de thèse se propose d’implémenter cette méthode de réduction de variance adaptative dans le code Monte-Carlo de transport de particules TRIPOLI-4,dédié entre autres aux études de radioprotection et d’instrumentation nucléaire. Caractérisées par de fortes atténuations des rayonnements dans la matière, ces études entrent dans la problématique du traitement d’évènements rares. Outre son implémentation inédite dans ce domaine d’application, deux nouvelles fonctionnalités ont été développées pour l’AMS, testées puis validées. La première est une procédure d’encaissement au vol permettant d’optimiser plusieurs scores en une seule simulation AMS. La seconde est une extension de l’AMS aux processus branchants, courants dans les simulations de radioprotection, par exemple lors du transport couplé de neutrons et des photons induits par ces derniers. L’efficacité et la robustesse de l’AMS dans ce nouveau cadre applicatif ont été démontrées dans des configurations physiquement très sévères (atténuations du flux de particules de plus de 10 ordres de grandeur), mettant ainsi en évidence les avantages prometteurs de l’AMS par rapport aux méthodes de réduction de variance existantes. / The Adaptive Multilevel Splitting algorithm (AMS) has recently been introduced to the field of applied mathematics as a variance reduction scheme for Monte Carlo Markov chains simulation. This Ph.D. work intends to implement this adaptative variance reduction method in the particle transport Monte Carlo code TRIPOLI-4, dedicated among others to radiation shielding and nuclear instrumentation studies. Those studies are characterized by strong radiation attenuation in matter, so that they fall within the scope of rare events analysis. In addition to its unprecedented implementation in the field of particle transport, two new features were developed for the AMS. The first is an on-the-fly scoring procedure, designed to optimize the estimation of multiple scores in a single AMS simulation. The second is an extension of the AMS to branching processes, which are common in radiation shielding simulations. For example, in coupled neutron-photon simulations, the neutrons have to be transported alongside the photons they produce. The efficiency and robustness of AMS in this new framework have been demonstrated in physically challenging configurations (particle flux attenuations larger than 10 orders of magnitude), which highlights the promising advantages of the AMS algorithm over existing variance reduction techniques.

[pt] O processo de revestimento por extrusão é um método muito utilizado na manufatura de diversos produtos. Ele pertence à uma classe de processos de revestimento chamada revestimento com vazão pré-fixada: para operações em regime permanente, a espessura da camada de líquido revestida é definida pela vazão na entrada do equipamento e pela velocidade do substrato. Para diversas aplicações, o líquido de revestimento é uma suspensão de partículas. A abordagem comumente usada é analisar o escoamento como se o fluido fosse Newtoniano, onde sua viscosidade é avaliada pela concentração média de partículas. Porém, dados experimentais mostram que a distribuição de partículas não é uniforme em escoamentos com cisalhamento. Além disso, a distribuição de partículas no filme após o processo de revestimento afetará a estrutura final do filme e, consequentemente, a qualidade final do produto. Portanto, é importante entender os fundamentos do processo de revestimento com suspensão de partículas para melhor descrever este processo e prever comportamentos desse escoamento. O objetivo do presente trabalho é estudar diferentes mecanismos de migração de partículas em suspensão que podem afetar a distribuição de partículas no filme revestido; como mecanismos de difusão e sedimentação. O modelo apresentado leva em consideração a dependência da viscosidade pela concentração local de partículas e o efeito de Marangoni. Dois modelos diferentes foram usados para estudar o escoamento de Marangoni. No primeiro, o transporte de partículas entre a superfície e o escoamento é considerado muito mais rápido que a difusão no escoamento de tal forma que o fluxo líquido é zero e a concentração de partículas na região do escoamento próxima à superfície. No segundo, um modelo mais completo, os efeitos de adsorção e dessorção de partículas pela interface é considerado. O sistema final de equações e não-linear e, com as condições de contorno apropriadas, é resolvido pelo método de Galerkin em elementos finitos e pelo método de Newton. O escoamento é bi-dimensional e, portanto, elementos bi-dimensionais são utilizados para descrever o domínio do problema. Entretanto, para avaliar a concentração ao longo da superfície, foi necessário criar um elemento uni-dimensional. O acoplamento entre estas duas classes de elementos também é discutido. Os resultados mostram que a distribuição de partículas do filme revestido é uma forte função da espessura do filme e das propriedades da suspensão, como a densidade das partículas e coeficientes de difusão. O escoamento de Marangoni pode afetar o padrão de recirculação e a distribuição de concentração de partículas. Além disso, os resultados mostram que a adsorção e dessorção de partículas na interface afetam fortemente a distribuição de partículas. / [en] Slot coating is a common method in the manufacturing of a wide variety of products. It belongs to a class of coating method known as premetered coating: in a steady state operation, the thickness of the coated liquid layer is set by the flow rate fed to the die and the speed of the substrate moving past, and is independent of other variables of the process. For many applications, the coating liquid is a particle suspension. The commom simplified approach is to study the flow as Newtonian and evaluate its viscosity based on the average particle concentration. However, experimental data shows that particle distribution is in fact non-uniform in shear flows. Moreover, particle distribution along the film thickness during the coating process will affect the final fim structure and consequently product performance. Hence, it is important to understand the fundamentals of coating process with particle suspension to better describe and predict the behavior of the flow and the particle distribution in the coated layer. The goal of the present work is to study different particle migration mechanisms in a suspension that may affect particle distribution in the coated film; such as diffusive mechanisms and sedimentation. The model presented takes into account the viscosity dependency and local particle concentration and surface-active particles, i.e. Marangoni effect. Two different approaches were used to study Marangoni flow. First, the bulk-interface transport is considered much faster than bulk diffusion such that the net flux is zero and surface concentration is equal bulk concentration. The second, more complete model, considers adsorption and desorption fluxes from the interface. The non-linear system of equations, with appropriate boundary conditions, is solved by Galerkin Finite Element Method and Newton s Method. The flow is two-dimensional and therefore two-dimensional elements are used to describe flow domain. Nonetheless, in order to evaluate particle concentration along the interface a one-dimensional element was created. The coupling between those two classes of elements of discussed. The results show that the particle distribution on the coated layer is a strong function of the film thickness and suspension properties, such as particle density and diffusion coefficients. The Marangoni flow associated with surface tension gradient due to particle concentration variation along the interface can change the recirculation pattern and particle concentration distribution. Furthermore, the results show that particle adsorption and desorption from interface have a strong effect on the particle distribution.

Solving radiation shielding problems, i.e. deep penetration problems, is a challenging task from both computation time and resource aspects in field of nuclear engineering. This is mainly because of the complexity of the governing equation for neutral particle transport - Linear Boltzmann Equation (LBE). The LBE includes seven independent variables with presence of integral and differential operators. Moreover, the low successive rate of radiation shielding problem is also challenging for solving such problems. In this dissertation, the Detector Response Function (DRF) methodology is proposed and developed for real-time and accurate radiation shielding calculation. The real-time capability of solving radiation shielding problem is very important for: (1) Safety and monitoring of nuclear systems; (2) Nuclear non-proliferation; and (3) Sensitivity study and Uncertainty quantification. Traditionally, the difficulties of solving radiation problem are: (1) Very long computation time using Monte Carlo method; (2) Extremely large memory requirement for deterministic method; and (3) Re-calculations using hybrid method. Among all of them, the hybrid method, typically Monte Carlo + deterministic, is capable of solving radiation shielding problem more efficiently than either Monte Carlo or deterministic methods. However, none of the aforementioned methods are capable of performing "real-time" radiation shielding calculation. Literature survey reveals a number of investigation on improving or developing efficient methods for radiation shielding calculation. These methods can be categorized by: (1) Using variance reduction techniques to improve successive rate of Monte Carlo method; and (2) Developing numerical techniques to improve convergence rate and avoid unphysical behavior for deterministic method. These methods are considered clever and useful for the radiation transport community. However, real-time radiation shielding calculation capability is still missing although the aforementioned advanced methods are able to accelerate the calculation efficiency significantly. In addition, very few methods are "Physics-based" For example, the mean free path of neutrons are typically orders of magnitude smaller than a nuclear system, i.e. nuclear reactor. Each individual neutron will not travel too far before its history is terminated. This is called the "loosely coupled" nature of nuclear systems. In principle, a radiation shielding problem can be potentially decomposed into pieces and solved more efficient. In the DRF methodology, the DRF coefficients are pre-calculated with dependency of several parameters. These coefficients can be directly coupled with radiation source calculated from other code system, i.e. RAPID (Real-time Analysis for Particle transport and In-situ Detection) code system. With this arrangement, detector/dosimeter response can be calculated on the fly. Thus far, the DRF methodology has been incorporated into the RAPID code system, and applied on four different benchmark problems: (1) The GBC-32 Spent Nuclear Fuel (SNF) cask flooded with water with a $^3$He detector placed on the cask surface; (2) The VENUS-3 experimental Reactor Pressure Vessel (RPV) neutron fluence calculation benchmark problem; (3) RPV dosimetry using the Three-Mile Island Unit-1 (TMI-1) commercial reactor; and (4) A Dry storage SNF cask external dosimetry problem. The results show that dosimeter/detector response or dose value calculations using the DRF methodology are all within $2sigma$ relative statistical uncertainties of MCNP5 + CADIS (Consistent Adjoint Driven Importance Sampling) standard fixed-source calculation. The DRF methodology only requires order of seconds for the dosimeter/detector response or dose value calculations using 1 processor if the DRF coefficients are appropriately prepared. The DRF coefficients can be reused without re-calculations when a model configuration is changed. In contrast, the standard MCNP5 calculations typically require more than an hour using 8 processors, even using the CADIS methodology. The DRF methodology has enabled the capability of real-time radiation shielding calculation. The radiation transport community can be greatly benefited by the development of DRF methodology. Users can easily utilize the DRF methodology to perform parametric studies, sensitivity studies, and uncertainty quantifications. The DRF methodology can be applied on various radiation shielding problems, such as nuclear system monitoring and medical radiation facilities. The appropriate procedure of DRF methodology and necessary parameters on DRF coefficient dependency will be discussed in detail in this dissertation. / Doctor of Philosophy / Since the beginning of nuclear era, enormous amount of radiation applications have been proposed, developed, and applied in our daily life. The radiation is useful and beneficial when they are under control. However, there will be some "unwanted radiation" from these applications, which have to be shielded. For this, radiation shielding has become a very important task. To effectively shield the unwanted radiations, studying the thickness and design of the shields is important. Instead of directly performing experiments, computation is a more affordable and safer approach. The radiation shielding computation is typically an extremely difffficult task due to very limited "communication" between the radiation within the shield and detector outside the shield. In general, it is impractical to simulate the radiation shielding problems directly because the extremely expensive computation resources. Most of interactions of radiation are within the shield while we are only interested in how many of them penetrate through the shield. This is typically called "deep penetration" problems in the radiation transport community.

The betavoltaic battery is a type of micro nuclear battery that harvests beta emitting radioactive decay energy using semiconductors. The literature results suggest that a better model is needed to design a betavoltaic battery. This dissertation creates a comprehensive model that includes all of the important factors that impact betavoltaic battery output and efficiency. Recent advancements in micro electro mechanical systems (MEMS) necessitate an onboard miniaturized power source. As these devices are highly functional, longevity of the power source is also preferred. Betavoltaic batteries are a very promising power source that can fulfill these requirements. They can be miniaturized to the size of a human hair. On the other hand, miniaturization of chemical batteries is restricted by low energy density. That is why betavoltaics are a viable option as a power source for sophisticated MEMS devices. They can also be used for implantable medical devices such as pacemakers; for remote applications such as spacecraft, undersea exploration, polar regions, mountains; military equipment; for sensor networks for environmental monitoring; and for sensors embedded in bridges due to their high energy density and long lifetime (up to 100 years). A betavoltaic battery simulation model was developed using Monte Carlo particle transport codes such as MCNP and PENELOPE whereas many researchers used simple empirical equations. These particle transport codes consider the comprehensive physics theory for electron transport in materials. They are used to estimate the energy deposition and the penetration depth of beta particles in the semiconductors. A full energy spectrum was used in the model to take into account the actual radioactive decay energy of the beta particles. These results were compared to the traditional betavoltaic battery design method of estimating energy deposition and penetration depth using monoenergetic beta average energy. Significant differences in results were observed that have a major impact on betavoltaic battery design. Furthermore, the angular distribution of the beta particles was incorporated in the model in order to take into account the effect of isotropic emission of beta decay. The backscattering of beta particles and loss of energy with angular dependence were analyzed. Then, the drift-diffusion semiconductor model was applied in order to estimate the power outputs for the battery, whereas many researchers used the simple collection probability model neglecting many design parameters. The results showed that an optimum junction depth can maximize the power output. The short circuit current and open circuit voltage of the battery varied with the semiconductor junction depth, angular distribution, and different activities. However, the analysis showed that the analytical results overpredicted the experimental results when self-absorption was not considered. Therefore, the percentage of self-absorption and the source thickness were estimated using a radioisotope source model. It was then validated with the thickness calculated from the specific activity of the radioisotope. As a result, the battery model was improved significantly. Furthermore, different tritiated metal sources were analyzed and the beta fluxes were compared. The optimum source thicknesses were designed to increase the source efficiencies. Both narrow and wide band gap semiconductors for beryllium tritide were analyzed. / PHD / A betavoltaic battery is a type of micro nuclear battery that harnesses electrical energy from radioisotopes using semiconductors. It has high specific energy density and longevity but low specific power. It can be miniaturized to a micron scale size (a size of a human hair) to power micro/nano sensors or devices. They can be used in implantable biomedical devices such as pacemakers, remote areas such as high mountains, undersea, and also in embedded sensors in structures. Chemical and other types of batteries are not suitable at this scale due to their low specific energy density. A betavoltaic battery is an attractive choice in applications where reliability and long service life (up to 100 years) are required. However, their power output is very low (on the scale of microwatts) due to their low specific power. They can aid chemical batteries to increase their lifetime by designing a hybrid battery. In a hybrid battery, a betavoltaic battery can trickle charge a chemical battery to top off the depleted charge. A theoretical analysis of a battery design is useful to improve its power output and efficiency. The literature in this area suggests that a better theoretical model is required to agree well with the experimental results as well as for better design. This model comprehensively included all the important factors that impact betavoltaic battery output and efficiency. All the necessary betavoltaic battery design factors were analyzed in detail in this work in order to maximize the desired output.

Le polymère poreux (PM) associe les avantages double des matériaux poreux et des polymères, ayant la structure unique de pore, la porosité supérieure et la densité inférieure, ce qui possède une valeur d’application importante dans les domaines de l'adsorption, le soutien de catalyseur, le séparateur de batterie, la filtration, etc. Actuellement, il existe plusieurs façons de préparer le PM, comme la méthode de gabarit, la méthode de séparation de phase, la méthode d'imagerie respiratoire, etc. Chacune des méthodes ci-dessus existe ses propres avantages, mais la préparation à grande échelle de PM à structure de pore contrôlable et aux fonctions spécifiques est toujours un objectif à long terme sur le domaine et l'un des principaux objectifs de ce mémoire. La co-extrusion de microcouche est une méthode pour produire de façon efficace et successive des polymères avec des structures de couches alternées, ayant les avantages de haute efficacité et faible coût. Par conséquent, sur les exigences structurelles de PM de l’application spécifique, ce mémoire a conçu le PM avec une structure spécifique et une co-extrusion de microcouche de manière créative combinée avec la méthode traditionnelle de préparation de PM (méthode de gabarit, méthode de séparation de phase), en combinant les avantages des deux méthodes, les PM avec une structure de pore idéale peuvent être préparés en grande quantité et l’on peut également explorer son application dans les séparateurs de batteries au lithium-ion et l'adsorption d'hydrocarbures aromatiques polycycliques.Le plus important, dans la deuxième partie de cet essai, se trouve que la simulation micro-numérique est utilisée pour étudier le transport et le dépôt de particules dans des milieux poreux pour explorer le mécanisme des matériaux poreux dans les domaines de l'adsorption et du séparateur de batterie. Le code de 3D-PTPO (un modèle tridimensionnel de suivi des particules combinant Python® et OpenFOAM®) est utilisé pour étudier le transport et le dépôt de particules colloïdales dans des milieux poreux, l’on adopte trois modèles (colonne, venturi et tube conique) pour représenter différentes formes de matériaux poreux. Les particules sont considérées comme des points matériaux pendant le transport, le volume des particules sera reconstitué et déposé comme partie de la surface du matériau poreux pendant le dépôt, la caractéristique principale de ce code est de considérer l'influence du volume des particules déposées sur la structure des pores, les lignes d'écoulement et le processus du dépôt des autres particules. Les simulations numériques sont d'abord conduites dans des capillaires simples, le travail de chercheurs de Lopez et d’autres est réexaminé en établissant un modèle géométrique tridimensionnel plus réaliste et il explore les mécanismes cachés derrière les règles de transmission et de dépôt. Par la suite, des simulations numériques sont effectuées dans des capillaires convergents-divergents pour étudier la structure des pores et l'effet de nombre Peclet sur le dépôt de particules. Enfin, l’on étudie l’effet double de l'hétérogénéité de surface et de l'hydrodynamique sur le comportement de dépôt de particules. / Due to the combination of the advantages of porous media and polymer materials, polymeric porous media possess the properties of controllable porous structure, easily modifiable surface properties, good chemical stability, etc., which make them applicable in a wide range of industrial fields, including adsorption, battery separator, catalyst carrier, filter, energy storage, etc. Although there exist various preparation methods, such as template technique, emulsion method, phase separation method, foaming process, electrospinning, top-down lithographic techniques, breath figure method, etc., the large-scale preparation of polymeric porous media with controllable pore structures and specified functions is still a long-term goal in this field, which is one of the core objectives of this thesis. Therefore, in the first part of the thesis, polymeric porous media are firstly designed based on the specific application requirements. Then the designed polymeric porous media are prepared by the combination of multilayer coextrusion and traditional preparation methods (template technique, phase separation method). This combined preparation method has integrated the advantages of the multilayer coextrusion (continuous process, economic pathway for large-scale fabrication, flexibility of the polymer species, and tunable layer structures) and the template/phase separation method (simple preparation process and tunable pore structure). Afterwards, the applications of the polymeric porous media in polycyclic aromatic hydrocarbons adsorption and lithium-ion battery separator have been investigated.More importantly, in the second part of the thesis, numerical simulations of particle transport and deposition in porous media are carried out to explore the mechanisms that form the theoretical basis for the above applications (adsorption, separation, etc.). Transport and deposition of colloidal particles in porous media are of vital important in other applications such as aquifer remediation, fouling of surfaces, and therapeutic drug delivery. Therefore, it is quite worthy to have a thorough understanding of these processes as well as the dominant mechanisms involved. In this part, the microscale simulations of colloidal particle transport and deposition in porous media are achieved by a novel colloidal particle tracking model, called 3D-PTPO (Three-Dimensional Particle Tracking model by Python® and OpenFOAM®) code. The particles are considered as a mass point during transport in the flow and their volume is reconstructed when they are deposited. The main feature of the code is to take into account the modification of the pore structure and thus the flow streamlines due to deposit. Numerical simulations were firstly carried out in a capillary tube considered as an element of an idealized porous medium composed of capillaries of circular cross sections to revisit the work of Lopez and co-authors by considering a more realistic 3D geometry and also to get the most relevant quantities by capturing the physics underlying the process. Then microscale simulation is approached by representing the elementary pore structure as a capillary tube with converging/diverging geometries (tapered pipe and venturi tube) to explore the influence of the pore geometry and the particle Péclet number (Pe) on particle deposition. Finally, the coupled effects of surface chemical heterogeneity and hydrodynamics on particle deposition in porous media were investigated in a three-dimensional capillary with periodically repeating chemically heterogeneous surfaces.

The fluid flow and tracer transport in a single rock fracture during shear processes has been an important issue in rock mechanics and is investigated in this thesis using Finite Element Method (FEM) and streamline particle tracking method, considering evolutions of aperture and transmissivity with shear displacement histories under different normal stresses, based on laboratory tests. The distributions of fracture aperture and its evolution during shear were calculated from the initial aperture fields, based on the laser-scanned surface roughness features of replicas of rock fracture specimens, and shear dilations measured during the coupled shear-flow-tracer tests in laboratory performed using a newly developed testing apparatus in Nagasaki University, Nagasaki, Japan. Three rock fractures of granite with different roughness characteristics were used as parent samples from which nine plaster replicas were made and coupled shear-flow tests was performed under three normal loading conditions (two levels of constant normal loading (CNL) and one constant normal stiffness (CNS) conditions). In order to visualize the tracer transport, transparent acrylic upper parts and plaster lower parts of the fracture specimens were manufactured from an artificially created tensile fracture of sandstone and the coupled shear-flow tests with fluid visualization was performed using a dye tracer injected from upstream and a CCD camera to record the dye movement. A special algorithm for treating the contact areas as zero-aperture elements was used to produce more accurate flow field simulations by using FEM, which is important for continued simulations of particle transport, but was often not properly treated in literature. The simulation results agreed well with the flow rate data obtained from the laboratory tests, showing that complex histories of fracture aperture and tortuous flow channels with changing normal stresses and increasing shear displacements, which were also captured by the coupled shear-flow tests of fracture specimens with visualization of the fluid flow. From the obtained flow velocity fields, the particle transport was predicted by the streamline particle tracking method with calculated flow velocity fields (vectors) from the flow simulations, obtaining results such as flow velocity profiles, total flow rates, particle travel time, breakthrough curves and the Péclet number, Pe, respectively. The fluid flow in the vertical 2-D cross-sections of a rock fracture was also simulated by solving both Navier-Stokes (NS) and Reynolds equations, and the particle transport was predicted by streamline particle tracking method. The results obtained using NS and Reynolds equations were compared to illustrate the degree of the validity of the Reynolds equation for general applications in practice since the later is mush more computationally efficient for large scale problems. The flow simulation results show that the total flow rate and the flow velocity predicted by NS equations are quite different from that as predicted by the Reynolds equation. The results show that a roughly 5-10 % overestimation on the flow rate is produced when the Reynolds equation is used, and the ideal parabolic velocity profiles defined by the local cubic law, when Reynolds equation is used, is no longer valid, especially when the roughness feature of the fracture surfaces changes with shear. These deviations of flow rate and flow velocity profiles across the fracture aperture have a significant impact on the particle transport behavior and the associated properties, such as the travel time and Péclet number. The deviations increase with increasing flow velocity and become more significant when fracture aperture geometry changes with shear. The scientific findings from these studies provided new insights to the physical behavior of fluid flow and mass transport in rock fractures which is the scientific basis for many rock mechanics problems at the fundamental level, and with special importance to rock engineering problems such as geothermal energy extraction (where flow rate in fractures dominates the productivity of a geothermal energy reservoir) and nuclear waste repositories (where radioactive nuclides transport through fractures dominates the final safety evaluations) in fractured rocks. / Vätskeflödet och spårämnestransporten i en enskild bergsspricka under skjuvningsprocesser har varit ett viktigt ämne inom bergmekanik. I denna avhandling undersöks ämnet med hjälp av finita element metoden (FEM) och en strömlinjebaserad partikelspårningsmetod. Hänsyn tas till utveckling av öppningar och transmissivitet med skjuvningens förflyttningshistoria under olika normala belastningar baserat på laboratorietester. Fördelningen av spricköppningar och deras utveckling under skjuvning beräknades från de initiala öppningsfälten baserat på det laserscannade provets ytas grovhetskännetecken sam tskjuvningsöppningar uppmätta under de kopplade skjuvning-flöde-spårämneslaboratorietesterna som utförts med nyutvecklad testapparatur i Nagasaki Universitet i Nagasaki, Japan. Tre bergssprickor i granit med olika grovhetskarakteristika användes som utgångsprover från vilka nio gipskopior gjordes. Kopplade skjuvning-flödes tester utfördes sedan under tre normala belastningstillstånd (två nivåer med konstant normal last (KNL) och en konstant normal styvhetstillstånd (KNS). För att visualisera spårämnestransporten tillverkades en transparent övre del av sprickproverna av akryl och en nedre del av gipsbaserat på en kostgjord spänningsspricka i sandsten och de kopplade skjuvning-flödes testerna med vätskevisualisering utfördes med färgspårämne injekterat uppströms och en CCD kamera monterad ovanför för att registrera färgens rörelse. En särskild algoritm användes för att behandla kontaktytorna som nollöppningsämnen användes för att åstadkomma mer exakta flödesfältssimuleringar med FEM. Detta är viktigt för kontinuerliga simuleringar av partikelflöden men uppmärksammas oftast inte tillräckligt i litteraturen. Simuleringsresultaten överensstämde väl med de flödesnivådata som erhölls från laboratorietesterna vilket visade att komplexa historier av spricköppningar och invecklade flöden överensstämde med ändrade normala belastningar och ökande skjuvningsförflyttningar, vilket även fångades av de kopplade skjuvning-flödestesterna av sprickproverna genom visualisering av vätskeflödet. Från de erhållna flödesfälten förutsågs partikeltransporten genom en strömlinjebaserad partikelspårningsmetod med kalkylerade flödeshastighetsfält (vektorer) från flödessimuleringarna genom vilka resultat som flödeshastighetsprofiler, totala flödesnivåer,partikeltransporttid, genombrottskurvor samt Pécletnumret, Pe, erhölls. Vätskeflödet i det vertikala tvådimensionella tvärsnittet av en bergsspricka simulerades även genom att både Navier-Stokes (NS) och Reynoldsekvationerna löstes och partikeltransporten förutsågs genom den strömlinjebaserade partikelspårningsmetoden. Resultaten som erhöllsmed NS och Reynoldsekvationerna jämfördes för att illustrera graden av tillförlitlighet för Reynoldsekvationen för allmänna tillämpningar i praktiken då den senare är betydligt mer beräkningseffektiv för storskaliga problem. Resultaten från flödessimuleringarna visar att den totala flödesnivån och den totala flödeshastigheten förutsedda med NS ekvationer är helt annorlunda motsvarande värden som förutsågs med Reynoldsekvationen. Resultaten visar att en ca 5-10 % för hög uppskattning av flödesnivån erhålls då Reynoldsekvationen används och de ideala parabola hastighetsprofilerna, som definieras av den lokala kubiklagen när Reynoldsekvationen används, inte längre är giltiga särskilt när sprickytornas grovhetskarakteristika ändras med skjuvning. De här avvikelserna i flödesnivå och flödeshastighetsprofiler längs med spricköppningen har en betydande påverkan på partikeltransportuppträdande och de tillhörande egenskaperna såsom rörelsetid och Pécletnummer. Avvikelserna ökar med ökande flödeshastighet och blir mer signifikanta när spricköppningarnas geometri ändras med skjuvning. Forskningsresultaten från dessa studier gav nya insikter i de fysiska uppträdandet av vätskeflöde och masstransporter i bergssprickor vilket är den vetenskapliga basen för många bergmekanikproblem på grundläggande nivå och som har särskild vikt för bergstekniksproblem såsom geotermisk energiutvinning (där flödesnivå i sprickor dominerar produktiviteten för en geotermisk energikälla) och kärnavfallsförvaringsplatser (där transporten av radioaktiva nuklider genom sprickor dominerar den slutgiltigasäkerhetsutvärderingen) i sprickigt berg. / QC 20100803

Im Grundwasser mobile Feststoff-Partikel stehen im Verdacht, den Transport schwerlöslicher Schadstoffe zu begünstigen. Die Partikel-Konzentration im Wasser lässt sich aber nur mittels einer aufwändigen Probenahme bestimmen. Vor diesem Hintergrund wurden im Rahmen dieser Arbeit Feld- und Laborversuche durchgeführt. In Feldversuchen zur Probenahme wurde festgestellt, dass für eine Stabilisierung der Partikel-Anzahl-Konzentration (CN in Partikel>Nachweisgrenze/L) die Dauer der Probenahme ausschlaggebend ist, nicht das insgesamt geförderte Volumen. Eine hohe Förderrate hat dabei nicht automatisch eine stärkere Mobilisierung von Partikeln aus dem Partikel-Depot im Umfeld der Messstelle zur Folge, wie es häufig postuliert wird. Die Gewinnung einer repräsentativen Partikelprobe macht eine Probenahme-Dauer von mindestens 5, häufig sogar über 10 oder 12 Stunden erforderlich. Empfehlungen zur Probenahme wurden erarbeitet. Es wurde aber festgestellt, dass sich stabilisierte Werte für CN als Funktion der Förderdauer (t in min) nach CN = a ? t^(-½) ungefähr abschätzen lassen, wenn der Parameter a über CN nach ca. 60 Minuten ermittelt wird. Bei Wässern mit Sauerstoff-Konzentrationen über ca. 1 mg/L kann überdies auch die Verteilung der Partikel auf einzelne Größenklassen bereits nach dieser Zeit ermittelt werden; bei sauerstoffärmeren Wässern verändert sie sich später noch. Angesichts einer starken Variation von CN einer Messstelle ist es generell aber nur möglich, eine Größenordnung für die Hintergrundkonzentration anzugeben. Von Messstelle zu Messstelle sind große Unterschiede bei CN festzustellen. Die Konzentration ist dabei nicht eindeutig abhängig von der Lithologie des Grundwasserleiters oder der Zusammensetzung des Grundwassers. Allerdings konnte eine schwache Korrelation mit dem Redox-Potential (Eh in mV) des Wassers gefunden werden (CN = 1,8?10^6?e^( 0,0087?Eh) [Partikel>2,58µm/L]; r² ≈ 0,46). Hierüber besteht die Möglichkeit, CN eines Grundwassers auch ohne Partikelmessung abzuschätzen. Insgesamt bewegen sich die Partikel-Massen-Konzentrationen (C in mg/L) der beprobten Wässer meist im Bereich von <1 mg/L, öfters sogar <0,1 mg/L, wenn die im Verlaufe von Probenahme und Messung ausgefällten Fe- und Mn-Oxide nicht berücksichtigt werden. Solche Konzentrationen sind vermutlich kaum in der Lage, relevante Mengen selbst sehr hydrophober Schadstoffen aufzunehmen und zu transportieren. Versuche mit Phenanthren ergaben zudem, dass es anscheinend weniger an bereits in Suspension befindliche Partikel sorbiert, sondern vielmehr in sorbiertem oder kristallinen Zustand aus dem Depot erodiert wird. In Laborversuchen wurde weiter gezeigt, dass in natürlichen Sedimenten ein großes Depot mobilisierbarer Partikel vorhanden ist, das Partikel über einen sehr langen Zeitraum kontinuierlich abgeben kann. Durchbruchsversuche ergaben außerdem, dass ein Großteil zugegebener Partikel bei der Passage durch eine Sedimentprobe zurückgehalten und nur sehr allmählich wieder abgegeben wird. Allerdings war auch ein schneller (präferentieller) Durchbruch zu verzeichnen. In Modellrechnungen konnte gezeigt werden, dass sich ein solcher Partikeltransport weder über eine Filterfunktion noch über die Transportgleichung zufrieden stellend berechnen lässt. Daher ist es nötig, einerseits einen bevorzugten Transport und andererseits eine starke Retardation zu berücksichtigen. Letzteres kann am besten über verschiedene Retardationsfaktoren oder ein dynamisches Partikel-Depot mit Anlagerungs- und Ablösungskonstanten geschehen. / Solid particles that are mobile in groundwater are suspected to enhance the transport of hardly soluble contaminants. But particle concentrations in water can only be measured using time-consuming sampling-procedures. On this background field- and laboratory-experiments were conducted in this work. In field experiments on sampling it turned out, that sampling time is crucial for stabilizing particle number-concentration (CN in particles>detection limit/L), not the volume sampled. A high sampling rate does not -as often argued- automatically result into higher mobilization of particles from the particle-depot in the vicinity of the sampling-well. Obtaining a representative particle sample requires a sampling-time of at least 5, often even more than 10 or 12 hours. In this work recommendations on sampling are given. It was noticed that stabilized values of CN can be estimated as function of sampling time (t in min) by CN = A ? t^(-½), if parameter A is calculated using CN after about 60 minutes. In waters having oxygen-concentrations above approximately 1 mg/L even distribution of the particles into size classes can be estimated after this time; in oxygen-poor waters size-distributions stabilized much later. Because of strong variations of CN in a single measuring well it is generally only possible to give the magnitude of the background-particle-concentration. But among several measuring wells CN may differ by several magnitudes. The concentrations do neither definitely depend upon the lithology of the aquifer nor on the groundwater-composition. But a weak correlation to the redox-potential (Eh in mV) can be found (CN = 1.8 ? 10^6 ? e^( 0.0087 ? Eh) [particles>2.58µm/L]; r² ≈ 0,46). Using this connection it is possible to estimate a magnitude of CN of a groundwater without even measuring particles. Over all particle mass-concentrations (C in mg/L) of all sampled groundwaters were almost always <1 mg/L, often even <0.1 mg/L, at least if iron- and manganese-oxides that precipitated during measurements were ignored. Such particle concentrations are probably hardly capable of adsorbing and carrying relevant amounts of contaminants, even very hydrophobic ones. Experiments using phenanthrene in contaminated sand additionally showed that it is probably hardly adsorbed onto already suspended particles, but mostly eroded from the particle depot in adsorbed or crystalline state. In laboratory experiments it was further shown that there is a huge depot of mobilizable particles in natural sediments that can continually release particles over a very long period of time. Break-through-experiments showed in addition that a large part of particles fed into the system are retained during the passage through a sediment sample and that they are re-released only very slowly. However, there also was a fast (preferential) break-through. In model calculations it could be shown that such a particle transport can neither be sufficiently described by the filter-function nor by the transport equation. Because of that it is necessary to take into consideration a preferential transport on the one hand and a strong retardation on the other. The last one can at the best be described by several retardation-factors or a dynamic particle-depot having constant attachment- and detachment rates.