The relative dispersion of one fluid particle with respect to another is fundamentally related to the transport and mixing of contaminant species in turbulent flows. The most basic consequence of Kolmogorov's 1941 similarity hypotheses for relative dispersion, the Richardson-Obukhov law that mean-square pair separation distance grows with the cube of time at intermediate times in the inertial subrange, is notoriously difficult to observe in the environment, laboratory, and direct numerical simulations (DNS). Inertial subrange scaling in size parameters like the mean-square pair separation requires careful adjustment for the initial conditions of the dispersion process as well as a very wide range of scales (high Reynolds number) in the flow being studied. However, the statistical evolution of the shapes of clusters of more than two particles has already exhibited statistical invariance at intermediate times in existing DNS. This invariance is identified with inertial-subrange scaling and is more readily observed than inertial-subrange scaling for seemingly simpler quantities such as the mean-square pair separation Results from dispersion of clusters of four particles (called tetrads) in large-scale DNS at grid resolutions up to 4096 points in each of three directions and Taylor-scale Reynolds numbers from 140 to 1000 are used to explore the question of statistical universality in measures of the size and shape of tetrahedra in homogeneous isotropic turbulence in distinct scaling regimes at very small times (ballistic), intermediate times (inertial) and very late times (diffusive). Derivatives of fractional powers of the mean-square pair separation with respect to time normalized by the characteristic time scale at the initial tetrad size constitute a powerful technique in isolating cubic time scaling in the mean-square pair separation. This technique is applied to the eigenvalues of a moment-of-inertia-like tensor formed from the separation vectors between particles in the tetrad. Estimates of the proportionality constant "g" in the Richardson-Obukhov law from DNS at a Taylor-scale Reynolds number of 1000 converge towards the value g=0.56 reported in previous studies. The exit time taken by a particle pair to first reach successively larger thresholds of fixed separation distance is also briefly discussed and found to have unexplained dependence on initial separation distance for negative moments, but good inertial range scaling for positive moments. The use of diffusion models of relative dispersion in the inertial subrange to connect mean exit time to "g" is also tested and briefly discussed in these simulations. Mean values and probability density functions of shape parameters including the triangle aspect ratio "w," tetrahedron volume-to-gyration radius ratio, and normalized moment-of-inertia eigenvalues are all found to approach invariant forms in the inertial subrange for a wider range of initial separations than size parameters such as mean-square gyration radius. These results constitute the clearest evidence to date that turbulence has a tendency to distort and elongate multiparticle configurations more severely in the inertial subrange than it does in the diffusive regime at asymptotically late time. Triangle statistics are found to be independent of initial shape for all time beyond the ballistic regime. The development and testing of different schemes for parallelizing the cubic spline interpolation procedure for particle velocities needed to track particles in DNS is also covered. A "pipeline" method of moving batches of particles from processor to processor is adopted due to its low memory overhead, but there are challenges in achieving good performance scaling.

The project was aimed at describing the mixing phenomena relevant for both safety analysis, particularly in steam line break and boron dilution scenarios, and mixing phenomena of interest for economical operation and the structural integrity. Measurement data from a set of mixing experiments, gained by using advanced measurement techniques with enhanced resolution in time and space help to improve the basic understanding of turbulent mixing and to provide data for Computational Fluid Dynamics (CFD) code validation. Slug mixing tests simulating the start-up of the first main circulation pump are performed with two 1:5 scaled facilities: The Rossendorf coolant mixing model ROCOM and the VATTENFALL test facility, modelling a German Konvoi type and a Westinghouse type three-loop PWR, respectively. Additional data on slug mixing in a VVER-1000 type reactor gained at a 1:5 scaled metal mock-up at EDO Gidropress are provided. Experimental results on mixing of fluids with density differences obtained at ROCOM and the FORTUM PTS test facility are made available. Concerning mixing phenomena of interest for operational issues and thermal fatigue, flow distribution data available from commissioning tests (Sizewell-B for PWRs, Loviisa and Paks for VVERs) are used together with the data from the ROCOM facility as a basis for the flow distribution studies. The test matrix on flow distribution and steady state mixing performed at ROCOM comprises experiments with various combinations of running pumps and various mass flow rates in the working loops. Computational fluid dynamics calculations are accomplished for selected experiments with two different CFD codes (CFX-5, FLUENT). Best practice guidelines (BPG) are applied in all CFD work when choosing computational grid, time step, turbulence models, modelling of internal geometry, boundary conditions, numerical schemes and convergence criteria. The BPG contain a set of systematic procedures for quantifying and reducing numerical errors. The knowledge of these numerical errors is a prerequisite for the proper judgement of model errors. The strategy of code validation based on the BPG and a matrix of CFD code validation calculations have been elaborated. Besides of the benchmark cases, additional experiments were calculated by new partners and observers, joining the project later. Based on the "best practice solutions", conclusions on the applicability of CFD for turbulent mixing problems in PWR were drawn and recommendations on CFD modelling were given. The high importance of proper grid generation was outlined. In general, second order discretization schemes should be used to minimise numerical diffusion. First order schemes can provide physically wrong results. With optimised "production meshes" reasonable results were obtained, but due to the complex geometry of the flow domains, no fully grid independent solutions were achieved. Therefore, with respect to turbulence models, no final conclusions can be given. However, first order turbulence models like K-e or SST K-w are suitable for momentum driven slug mixing. For buoyancy driven mixing (PTS scenarios), Reynolds stress models provide better results.

No description available.

The main focus of the current study was to quantify the chemical signals received by a blue crab (Callinectes sapidus) tracking a source in a laboratory flume. To make a direct linkage between tracking behavior and the odorant concentration signal, we developed a measurement system to quantify the instantaneous concentration field surrounding actively tracking blue crabs. A three-dimensional laser-induced fluorescence (3DLIF) system was designed and constructed to measure odorant concentrations around crabs tracking three source types: a continuous release with exit velocity matching the mean local velocity in the flume; a continuous release with a meander created by an upstream cylinder; and a pulsed release switching on and off and with the same mass flow rate as the other two plume types. The meandering and pulsed plumes were introduced to observe the effects of large-scale spatial (meandering) and temporal (pulsed) intermittency on crab tracking. Simultaneous with the chemical concentration measurements, crab position data was recorded for kinematic analysis during post-processing. In addition, concentration measurements were collected for the three plume types without crabs present in order to quantify the statistical characteristics of the plume structure The concentration signals arriving at the antennules and outer chemosensory organs, most notably the legs, were targeted due to the hypotheses that concentration bursts at the antennules mediate upstream movement and that spatial contrast at the leg chemosensors mediates turning. A sampling zone was placed in front of the crab's mouth parts and aligned with the crab carapace orientation to extract odorant bursts at the antennules. The data generally showed an increase in upstream walking speed when high concentration bursts arrive at the antennules location, which agrees with the hypothesis. Measurement of the odorant concentration at the outer chemosensors was less direct and involved placing a box upstream of the crab and sampled earlier in time in order to avoid shadowing interference. Based on the signal at the upstream sampling box, a general bias for turning was observed. Crabs casted transversely in response to the directional bias extracted from the upstream sampling box. A statistical analysis of crab behavioral response to concentrations at the antennules and outer chemosensors can be found in a (future) companion thesis written by Jennifer Page in the School of Biology. Data were also taken for the three plume types in the absence of blue crabs. The continuous plume average statistics displayed Gaussian behavior at nozzle centerline. The meandering plume data conformed to the meandering plume model of Gifford (1959), modified for an induced pseudo-periodic meander. The pulsed plume displayed characteristics intermediate between the cloud dispersion model (Townsend 1951, Chatwin and Sullivan 1979) and the Gaussian dispersion model for a continuous release. For the three plume types, the standard deviation of the concentration fluctuations was greater than the average concentrations, as time records consisted of intermittent high concentrations interspersed with concentrations close to zero.

This thesis aim to provide SSC, Swedish Space Corporation, with a foundation for understanding the key ideas behind water injection during rocket lift-off, including problems to be avoided when simulating the phenomena. This investigation focus on finding approaches suitable for obtaining a rough estimate of the reduction in noise generation, when too expensive equipment required is absent. The main idea was to compare different methods at the end as an alternative suitable way of verifying, since validation data was not available. The setup of the simulations consisted of two cases, one with water injection and the second case was without, and they were simulated the OpenFOAM software while the mesh was constructed using the GMSH software. A 1D analytical prediction model was computed using Matlab to estimate the noise generated. The result of the simulation showed an error of approximately 300-400 m/s within the rocket engine when compared to the Rocket Propulsion Analysis (RPA) software result. The maximum sound pressure level without water injection (SPL) from the analytical prediction model, ended up at approximately 172dB as well as 164dB depending on where it was "recorded". The maximum SPL with water injection was approximately 7dB lower in both recorded locations which was achieved by using optimal initial values. The biggest error observed by researches using this prediction model is approximately +2 dB above the real value. However, the error from this specific setup could not be estimated. The challenges and approximations encountered throughout this investigation is thoroughly discussed within the thesis and despite the absence of accurate results this investigation provides a thorough insight into water injection during rocket lift-off, with the potential of achieving better results using a more advanced solver in OpenFOAM.

<p>Turbulent mixing layers have been researched for many years. Currently, research is focused on studying compressible mixing layers because of their widespread applications in high-speed flight systems. While the effect of compressibility on the shear layer growth rate is well established, there is a lack of consensus over its effect on the turbulent stresses and hence warrants additional research in this area. Computational studies on compressible shear layers could provide a deep cognizance of the dynamics of fluid structures present in these flow fields which in turn would be viable for understanding the effects of compressibility on such flows. However, performing a Direct Numerical Simulation (DNS) of a highly compressible shear layer with experimental flow conditions is extremely expensive, especially when resolving the boundary layers that lead into the mixing section. The attractive alternative is to use Large Eddy Simulation (LES), as it possesses the potential to resolve the flow physics at a reasonable computational cost. Therefore the current work deals with developing a methodology to perform LES of a compressible mixing layer with experimental flow conditions, with resolving the boundary layers that lead into the mixing section through a wall model. The wall model approach, as opposed to a wall resolved simulation, greatly reduces the computational cost associated with the boundary layer regions, especially when using an explicit time-stepping scheme. An in house LES solver which has been used previously for performing simulations of jets, has been chosen for this purpose. The solver is first verified and validated for mixing layer flows by performing simulations of laminar and incompressible turbulent mixing layer flows and comparing the results with the literature. Following this, LES of a compressible mixing layer at a convective Mach number of 0.53 is performed. The inflow profiles for the LES are taken from a precursor RANS solution based on the k-ε and RSM turbulence models. The results of the LES present good agreement with the reference experiment for the upstream boundary layer properties, the mean velocity profile of the shear layer and the shear layer growth rate. The turbulent stresses, however, have been found to be underpredicted. The anisotropy of the normal Reynolds stresses have been found to be in good agreement with the literature. Based on the present results, suggestions for future work are also discussed.</p>

The project was aimed at describing the mixing phenomena relevant for both safety analysis, particularly in steam line break and boron dilution scenarios, and mixing phenomena of interest for economical operation and the structural integrity. Measurement data from a set of mixing experiments, gained by using advanced measurement techniques with enhanced resolution in time and space help to improve the basic understanding of turbulent mixing and to provide data for Computational Fluid Dynamics (CFD) code validation. Slug mixing tests simulating the start-up of the first main circulation pump are performed with two 1:5 scaled facilities: The Rossendorf coolant mixing model ROCOM and the VATTENFALL test facility, modelling a German Konvoi type and a Westinghouse type three-loop PWR, respectively. Additional data on slug mixing in a VVER-1000 type reactor gained at a 1:5 scaled metal mock-up at EDO Gidropress are provided. Experimental results on mixing of fluids with density differences obtained at ROCOM and the FORTUM PTS test facility are made available. Concerning mixing phenomena of interest for operational issues and thermal fatigue, flow distribution data available from commissioning tests (Sizewell-B for PWRs, Loviisa and Paks for VVERs) are used together with the data from the ROCOM facility as a basis for the flow distribution studies. The test matrix on flow distribution and steady state mixing performed at ROCOM comprises experiments with various combinations of running pumps and various mass flow rates in the working loops. Computational fluid dynamics calculations are accomplished for selected experiments with two different CFD codes (CFX-5, FLUENT). Best practice guidelines (BPG) are applied in all CFD work when choosing computational grid, time step, turbulence models, modelling of internal geometry, boundary conditions, numerical schemes and convergence criteria. The BPG contain a set of systematic procedures for quantifying and reducing numerical errors. The knowledge of these numerical errors is a prerequisite for the proper judgement of model errors. The strategy of code validation based on the BPG and a matrix of CFD code validation calculations have been elaborated. Besides of the benchmark cases, additional experiments were calculated by new partners and observers, joining the project later. Based on the "best practice solutions", conclusions on the applicability of CFD for turbulent mixing problems in PWR were drawn and recommendations on CFD modelling were given. The high importance of proper grid generation was outlined. In general, second order discretization schemes should be used to minimise numerical diffusion. First order schemes can provide physically wrong results. With optimised "production meshes" reasonable results were obtained, but due to the complex geometry of the flow domains, no fully grid independent solutions were achieved. Therefore, with respect to turbulence models, no final conclusions can be given. However, first order turbulence models like K-e or SST K-w are suitable for momentum driven slug mixing. For buoyancy driven mixing (PTS scenarios), Reynolds stress models provide better results.

No description available.

Flotation is a complex, multiphase process used to separate minerals. Four problems central to the fundamentals of the flotation process were studied. A multiphase grid turbulence experiment was conducted to verify particle collision models. The slip velocities of solid particles and bubbles were measured using Digital Particle Image Velocimetry (DPIV). The experimental results were compared with the predictions from empirical and theoretical collision models. Time-resolved DPIV was used to measure the turbulent velocity field in a Rushton turbine around the impeller region. The turbulence quantities were found by removing the periodic component from the blade passing, which is a dominant part of the measured velocities near the impeller. We provide evidence that larger, biased dissipation and turbulent kinetic energy values are estimated in the vicinity of the impeller due to the periodic component of the blade passage. The flow was found to be anisotropic close to the impeller. Vortex detection revealed that the tip vortices travel in a nearly radial direction from the impeller for small Reynolds numbers and with a wider distribution for higher Reynolds numbers. The rise of a buoyant bubble and its interaction with a free liquid surface was experimentally investigated using Time-Resolved Digital Particle Image Velocimetry as a function of bubble size, and surfactant concentration of the fluid medium. It is shown that the presence of a surfactant significantly affected the characteristics of the velocity field during the rise and interaction with the free surface. This difference is attributed to the adsorption coverage of the surfactant at the bubble-fluid interface. Wake profiles were compared. The presence of large vortices were observed and found to play a significant role. Finally, Numerical and experimental results of stable and unstable foams are presented by comparing liquid fractions and bubble sizes. There was good agreement between the experiments and numerical modeling in free drainage and forced drainage experiments. In addition, foam coarsening was measured and characterized experimentally. Each of the problems investigated have added to the understanding in the underlying physics of the flotation process and can lead to more accurate modeling. The ultimate goal of this work is to contribute to the design of more effective and efficient flotation machines. / Ph. D.

Cette thèse s’inscrit dans le domaine de la simulation numérique appliquée à la propulsion. Le moteur à détonation rotative (RDE) fait partie des candidats susceptibles de remplacer nos actuels moyens de propulsion grâce à l’augmentation du rendement thermodynamique du moteur. Pour conserver l’avantage de la détonation, l’injecteur doit fournir un mélange dont la qualité doit être la meilleure possible tout en limitant les pertes de pression totale. La présente étude porte sur le développement et l’optimisation numérique d’un injecteur adapté au fonctionnement d’un RDE. L’injection d’hydrogène et d’oxygène gazeux en rapport stoechiométrique est considérée pour une utilisation en propulsion fusée. Le premier objectif est de proposer un concept réaliste d’injecteur permettant de maximiser le mélange des ergols. Le second objectif est de réaliser des études du mélange dans la chambre par des simulations LES (Large Eddy Simulation). Le troisième objectif est de simuler la propagation d’une détonation rotative (RD) alimentée par différents injecteurs en régimes prémélangé et séparé. Deux éléments d’injection sont mis en concurrence. Le premier utilise le principe de jets semi-impactants de H2 et de O2. Le deuxième représente une configuration améliorée. Les simulations de RD avec les deux injecteurs donnent des résultats similaires lorsque l’injection est prémélangée. La part du mélange injecté perdu par déflagration est de 30% et la vitesse de propagation de la RD est proche de la vitesse théorique CJ. Pour les injections séparées de H2 et O2, l’injecteur amélioré permet de conserver un bon niveau de mélange dans la chambre, contrairement à l’injecteur à semi-impact qui produit une forte stratification des ergols dans la chambre. En conséquence, la vitesse de propagation de la RD est proche de la vitesse CJ avec l’injecteur amélioré et limitée à 80% de la vitesse CJ avec l’injecteur à semi-impact. / This thesis pertains to the domain of numerical simulation for propulsion applications. The rotating detonation engine (RDE) appears to be a good candidate to replace our current means of propulsion thanks to the increase of the thermodynamic efficiency. To preserve the advantage given by the detonation mode, the injector must provide the best possible mixing of the propellants together with acceptable total pressure losses. This numerical study deals with developing and optimizing an injector adapted to the operation of a RDE. Injection of gaseous H2 and O2 at stoichiometric ratio is considered to be suitable for rocket propulsion application. The first goal is to propose an efficient injector design so that the mixing between the propellants is maximized. The second goal is to perform simulations of the mixing process in the chamber by LES (Large Eddy Simulation) computations. The third goal consists in computing the propagation of a rotating detonation (RD) fed by different injectors in premixed and separate regimes. This study allows the comparison of two injection elements. The first one uses the principle of semi-impinging jets of H2 and O2. The second one represents an improved configuration. RD simulations with both injectors provide similar results when premixed injection is considered. The part of the injected mixture that burns by deflagration is 30% and the detonation velocity remains close the theoretical CJ velocity. In the regime of separate injection of H2 and O2, the improved injector enables to keep a high mixing efficiency in the chamber whereas the semi-impingement injector produces a strong stratification of the propellants in the chamber. As a consequence, the detonation velocity is close to the CJ velocity with the improved injector and limited to 80% of the CJ velocity with the semi-impingement injector.