The recently developed technique of diffusive gradients in thin films (DGT) for speciation measurement of analytes in the environment was further developed through the development of series of new binding phases including poly(acrylamide-co-acrylic acid) copolymer hydrogel (PAM-PAA), poly(acrylamidoglycolic acid-co-acrylamide) (PAAGA-PAM) hydrogel, the Whatman P81 cellulose phosphate ion exchange membrane (P81) and a liquid binding phase of poly(4-styrenesulfonate) (PSS). A new diffusion layer, cellulose dialysis membrane, was also employed for the liquid binding phase DGT. PAM-PAA copolymer hydrogel was prepared by the controlled hydrolysis of polyacrylamide (PAM) in an alkaline solution of 10% sodium hydroxide. The capacity of the copolymer hydrogel to bind various metal ions was tested under a range of uptake conditions. Ions such as Cu2+ and Cd2+ were bound more strongly to the copolymer hydrogel than the competing ions such as Na+, K+, Ca2+ and Mg2+. Metals bound to the copolymer hydrogel can be efficiently eluted in 2 M HNO3 solution (>94%). Application of this new binding material to DGT technique was validated in a synthetic lake water (Windermere, Lake District, UK) with a recovery of 99.0% for Cu2+. PAAGA-PAM hydrogel was prepared by copolymerising 2-acrylamidoglycolic acid with acrylamide. The metal ion binding properties of the hydrogel were characterised for Cu2+, Cd2+ and competing ions under various experimental conditions. The hydrogel was shown to bind Cu2+ and Cd2+ strongly under non-competitive binding conditions, with binding capacities of 5.3 and 5.1 micromol cm-2, respectively. The binding capacity of each metal decreased, under competitive binding conditions (with a range of metal ions present at 17.8 mN), to 1.3 and 0.17 micromol cm-2, respectively, indicating a strong selective binding towards Cu2+. The metal ions were readily recovered (>94%) by eluting with 2 M HNO3. Finally, the copolymer hydrogel was tested as a binding phase with the DGT technique. A linear mass vs. time relationship was observed for Cu2+ in Windermere water with a recovery of close to 100%. The use of a commercially available solid ion exchange membrane (P81) as the binding phase in DGT analysis was demonstrated. P81 is a strong cation exchange membrane. Its performance characteristics as a new binding phase in DGT measurement of Cu2+ and Cd2+ were systematically investigated. Several advantages over the conventional ion exchange resin-embedded hydrogel based binding phases used in DGT were observed. These include: simple preparation, ease of handling, and reusability. The binding phase preferentially binds to transition metal ions rather than competing ions. Within the optimum pH range (pH 4.0 - 9.0), the maximum non-competitive binding capacities of the membrane for Cu2+ and Cd2+ were 3.22 and 3.07 micromol cm-2, respectively. The suitability of the new membrane-based binding phase for DGT applications was validated experimentally. The results demonstrated excellent agreement with theoretically predicted trends. The reusability of this binding phase was also investigated. Application of a liquid binding phase and a dialysis membrane diffusive layer were proposed for the first time. The binding phase was a 0.020 M solution of poly(4-styrenesulfonate) (PSS) polyelectrolyte using a specially designed DGT device. The binding properties of Cd2+, Cu2+, and a range of alkali and alkaline earth metal ions to the PSS solution were characterised. The PSS behaved like a cation exchanger with preferential binding to Cd2+ (6.0 micromole ml-1, log K = 9.0) and Cu2+ (2.5 micromole ml-1, log K = 8.1) under competitive binding conditions. The DGT devices were successfully validated for Cd2+ and Cu2+ in Windermere water. The speciation performance of the solid and liquid binding phases developed in this study was investigated in solutions containing ethylenediaminetetraacetic acid disodium salt (EDTA), humic acid (HA), glucose (GL), dodecylbenzenesulfonic acid (DBS) and tannic acid (TA) with Cu2+ and Cd2+. The ratios of labile metals over total metals were at good agreement with calculated theoretical values using Stability Constants Database. The results indicated that the DGT-labile concentration measured by DGT with these binding phases is essentially free metal ion concentration in the sample. All newly developed DGT binding phases were successfully applied for environmental speciation. The field deployments were carried out in both freshwater and salt-water test sites.

We developed a diffusive load balancing scheme that equalizes the available capacities of nodes in a peer-to-peer (P2P) system. These nodes may have different resource capacities, geographic locations, or availabilities (i.e., length of time being part of the peer-to-peer system). The services on these nodes may have different service times and arrival rates of requests. Using the diffusive scheme, the system is able to maintain similar response times for its services. Our scheme is a modification of the diffusive load balancing algorithms proposed for parallel computing systems. This scheme is able to handle services with heterogeneous resource requirements and P2P nodes with heterogeneous capacities. We also adapted the diffusive scheme to clustered peer-to-peer system, where a load balancing operation may move services or nodes between clusters. After a literature survey of this field, this thesis investigates the following issues using analytical reasoning and extensive simulation studies. The load balancing operations equalize the available capacities of the nodes in a neighborhood to their averages. As a result, the available capacities of all nodes in the P2P system converge to a global average. We found that this convergence is faster when the scheme uses neighborhoods defined by the structure of the structured P2P overlay network rather than using randomly selected neighbors. For a system with churn (i.e. nodes joining and leaving), the load balancing operations maintain the standard deviation of the available capacities of nodes within a bound. This bound depends on the amount of churn and the frequency of load balancing operations, as well as on the capacities of the nodes. However, the sizes of the services have little impact on this bound. In a clustered peer-to-peer system, the size of the bound largely depends on the average cluster size. When nodes are moved among clusters for load balancing, the numbers of cluster splits and merges are reduced. This may reduce the maintenance cost of the overlay network.

Resonant Magnetic Perturbation (RMP) fields produced by external control coils are considered a viable option for the suppression of Edge Localized Modes (ELMs) in present and future tokamaks. Repeated reversals of the toroidal phase of the I-coil magnetic field in RMP shot 147170 on DIII-D has generated uniquely different edge pedestal profiles, implying different edge transport phenomena. The causes, trends, and implications of RMP toroidal phase reversal on edge transport is analyzed by comparing various parameters at 0 and 60 degree toroidal phases, with an I-coil mode number of n=3. An analysis of diffusive and non-diffusive transport effects of these magnetic perturbations it the plasma edge pedestal for this RMP shot is characterized by interpreting the ion and electron heat diffusivities, angular momentum transport frequencies, ion diffusion coefficients, and pinch velocities for both phases.

In this thesis, we consider a projection-based stabilization method for solving buoyancy driven flows (natural convection problems). The method consists of adding global stabilization for all scales and then anti-diffusing these effects on the large scales defined by projections into appropriate function spaces. In this way, stabilization acts only on the small scales. We consider two different variations of buoyancy driven flows based on the projection-based stabilization. First, we focus on the steady-state natural convection problem of heat transport through combined solid and fluid media in a classical enclosure. We present the mathematical analysis of the projection-based method and prove existence, uniqueness and convergence of the approximate solutions of the velocity, temperature and pressure. We also present some numerical tests to support theoretical findings. Second, we consider a system of combined heat and mass transfer in a porous medium due to the natural convection. For the semi-discrete problem, a stability analysis of the projectionbased method and a priori error estimate are given for the Darcy-Brinkman equations in double-diffusive convection. Then we provide numerical assessments and a comparison with some benchmark data for the Darcy-Brinkman equations. In the last part of the thesis, we present a fully discrete scheme with the linear extrapolation of convecting velocity terms for the Darcy-Brinkman equations.

Double diffusive phenomena occur in many natural systems and in a variety of engineering applications, such as solar ponds. In a solar pond, the erosion of gradient zone due to convection and diffusion is the main cause of inefficiency and hence, requires lot of maintenance. In this thesis, control of erosion of the gradient zone in a solar pond situation has been studied through experiments and transient numerical simulations. A laboratory solar pond was setup in a tank of size 19cm X 19cm X 19cm, and was heated from below. A polymer additive, Carboxy Methyl Cellulose (CMC) was used to increase the viscosity of the working fluid. With 0.01% CMC, viscosity enhancement of the working fluid by 25 to 30 times was achieved, without changing any other properties. The transient momentum, energy and species conservation equations along with continuity equation were solved numerically, using the SIMPLER algorithm with 2 mm grid spacing. The experiments conducted with the addition of CMC, showed delays in the onset of convection and reduced erosion of the gradient zone. The erosion rates obtained from the numerical simulations agreed with the experimental observation. The impact of viscosity on the onset of convection, kinetic energy, convection pattern, frequency of bursts and erosion rate of the gradient zone, were simulated. Numerical simulations revealed that, there is a nonlinear relationship between the viscosity and erosion rate of the gradient zone. Increase in viscosity by a factor of 15 reduced the erosion of gradient zone completely, indicating the dominant role played by Turbulent entrainment. The present work indicates that the control of erosion of gradient zone in solar pond using the polymer additive shows lot of promise. The use of the polymer additive as a method to increase viscosity will lower the maintenance costs and, increase the reliability and efficiency of solar ponds, with less than 5% increase in the initial cost.

Density-dependent groundwater flow in the Floridan aquifer system (FAS) depends on chloride concentrations and fluid temperature. Previous studies addressing the role of chloride concentration and temperatures on groundwater flow in the FAS have relied on observation data or simplified two-dimensional numerical models. A three-dimensional hydrologic analysis of FAS in peninsular Florida was performed using a modified version of SUTRA (SUTRA-MS) capable of simulating multi-species solute and heat transport. SUTRA-MS was developed during this investigation and is capable of reproducing results for several problems with known solutions.The model was developed using available geometric and hydraulic parameter data and calibrated using hydraulic head, chloride concentrations, and temperatures representative of conditions prior to significant groundwater pumpage from the FAS. The calibrated model is capable of reproducing observed pressures and temperatures but in general ov er-simulates chloride concentrations. The inability of the model to simulate observed chloride concentrations suggests chloride concentrations in the FAS are not in equilibrium with current sea level. Previous hydrologic studies of the FAS have attributed anomalous chloride concentrations to incomplete flushing of relict seawater that entered the aquifer during previous sea-level highstands.Three hypothetical, sinusoidal sea-level changes occurring over 100,000-years were used to evaluate how the aquifer responds to sea-level fluctuations. Model results indicate pressure equilibrates most rapidly and is followed by temperatures and then chloride concentrations. Confining unit thicknesses directly affect response times of pressure, temperature, and chloride concentrations in the FAS.Simulation of the system with ("geothermal case") and without ("isothermal case") the geothermal component reveals that the inflow of seawater from the Florida Straits would be similar without the heat f low but the distribution would differ significantly. The addition of heat flow also reduces the asymmetry of the circulation. Simulations evaluating aquifer responses to sea-level fluctuations and the thermal component indicate that the complicated three-dimensional setting of the FAS is a key component of the groundwater flow system and steady state conditions may not exist for relatively thick coastal aquifers that have experienced multiple sea-level cycles.

Bien que souvent présentés comme des configurations atypiques, la classe des milieux diffusifs représente une grande partie des milieux dans lesquels se propagent des ondes aussi bien électromagnétiques qu'acoustiques. Les grandes cavités étant capables de bien approcher ces mêmes caractéristiques, elles sont largement utilisées dans un contexte métrologique, afin d'émuler un grand nombre de configurations pratiques et d'évaluer certaines propriétés de dispositifs électroniques, acoustiques et optiques. Nous nous intéressons à la modélisation statistique de la propagation d'ondes dans les grandes cavités. La pratique courante de modéliser les champs dans une cavité comme diffus est d'abord analysée, afin de montrer comment cette hypothèse n'est pas réaliste en basse fréquence, et les conséquences qui en découlent. L'importance du recouvrement modal et sa nature aléatoire sont discutées, montrant comment l'hypothèse diffusive ne peut pas être décrite comme une propriété certaine. Dans un deuxième temps nous étudions les applications du retournement temporel aux grandes cavités, ce qui nous amène à l'introduction d'une technique généralisée capable de reproduire des fronts d'ondes cohérents dans un environnement diffusif, typiquement regardé comme uniquement capable de supporter une propagation aléatoire.

We developed a diffusive load balancing scheme that equalizes the available capacities of nodes in a peer-to-peer (P2P) system. These nodes may have different resource capacities, geographic locations, or availabilities (i.e., length of time being part of the peer-to-peer system). The services on these nodes may have different service times and arrival rates of requests. Using the diffusive scheme, the system is able to maintain similar response times for its services. Our scheme is a modification of the diffusive load balancing algorithms proposed for parallel computing systems. This scheme is able to handle services with heterogeneous resource requirements and P2P nodes with heterogeneous capacities. We also adapted the diffusive scheme to clustered peer-to-peer system, where a load balancing operation may move services or nodes between clusters. After a literature survey of this field, this thesis investigates the following issues using analytical reasoning and extensive simulation studies. The load balancing operations equalize the available capacities of the nodes in a neighborhood to their averages. As a result, the available capacities of all nodes in the P2P system converge to a global average. We found that this convergence is faster when the scheme uses neighborhoods defined by the structure of the structured P2P overlay network rather than using randomly selected neighbors. For a system with churn (i.e. nodes joining and leaving), the load balancing operations maintain the standard deviation of the available capacities of nodes within a bound. This bound depends on the amount of churn and the frequency of load balancing operations, as well as on the capacities of the nodes. However, the sizes of the services have little impact on this bound. In a clustered peer-to-peer system, the size of the bound largely depends on the average cluster size. When nodes are moved among clusters for load balancing, the numbers of cluster splits and merges are reduced. This may reduce the maintenance cost of the overlay network.

Coordenação de Aperfeiçoamento de Pessoal de Nível Superior - CAPES / In this work, we investigate the pattern formation of a reaction-diffusion generalist predator-prey system with constant alternative food source and a Holling type III functional response. The local model can exhibit up to six equilibria, bistability, limit cyles and several bifurcations. We analyse diffusive instability conditions and verify the spatial distribution of preys and predators in a two-dimensional habitat with reflective zero-flux boundary conditions. We investigate the spatial behavior for three parameters combinations: Case 1: single coexistence equilibrium; Case 2: bistability; and Case 3: bistability and limit cycle. This work is divided as follows: nn Chapter 1, we present the stability analysis of the local model; in Chapter 2, we consider the movement by diffusion and find diffusive instability conditions; lastly, in Chapter 3 we perform extensive numerical simulations in order to investigate the spatial distribution of preys and predators. / Neste trabalho, investigamos a formação de padrões em um modelo de reação-difusão presa-predador generalista com fonte de alimento alternativa constante e resposta funcional Holling tipo III. O modelo local pode apresentar até seis equilíbrios, biestabilidade, ciclos limite e diversas bifurcações. Analisamos as condições de instabilidade difusiva e verificamos a distribuição espacial de presas e predadores em um habitat bidimensional com fronteira reflexiva. Investigamos o comportamento espacial para três combinações de parâmetros: Caso 1: equilíbrio de coexistência único; Caso 2: biestabilidade; e Caso 3: biestabilidade e ciclo limite. O trabalho está dividido da seguinte forma: no Capítulo 1, apresentamos a análise de estabilidade do modelo local; no Capítulo 2, inserimos a difusão no modelo e encontramos as condições de instabilidade difusiva; por fim, no Capítulo 3, realizamos diversas simulações numéricas a fim de investigar a distribuição espacial de presas e predadores.

We developed a diffusive load balancing scheme that equalizes the available capacities of nodes in a peer-to-peer (P2P) system. These nodes may have different resource capacities, geographic locations, or availabilities (i.e., length of time being part of the peer-to-peer system). The services on these nodes may have different service times and arrival rates of requests. Using the diffusive scheme, the system is able to maintain similar response times for its services. Our scheme is a modification of the diffusive load balancing algorithms proposed for parallel computing systems. This scheme is able to handle services with heterogeneous resource requirements and P2P nodes with heterogeneous capacities. We also adapted the diffusive scheme to clustered peer-to-peer system, where a load balancing operation may move services or nodes between clusters. After a literature survey of this field, this thesis investigates the following issues using analytical reasoning and extensive simulation studies. The load balancing operations equalize the available capacities of the nodes in a neighborhood to their averages. As a result, the available capacities of all nodes in the P2P system converge to a global average. We found that this convergence is faster when the scheme uses neighborhoods defined by the structure of the structured P2P overlay network rather than using randomly selected neighbors. For a system with churn (i.e. nodes joining and leaving), the load balancing operations maintain the standard deviation of the available capacities of nodes within a bound. This bound depends on the amount of churn and the frequency of load balancing operations, as well as on the capacities of the nodes. However, the sizes of the services have little impact on this bound. In a clustered peer-to-peer system, the size of the bound largely depends on the average cluster size. When nodes are moved among clusters for load balancing, the numbers of cluster splits and merges are reduced. This may reduce the maintenance cost of the overlay network.