Computational Fluid Dynamics (CFD) is a field that is constantly advancing. Its advances in terms of capabilities are a result of new theories, faster computers, and new numerical methods. In this thesis, advances in the computational fluid dynamic modeling of moving bodies and combustive flows are investigated. Thus, the basic theory behind CFD is being extended to solve a new class of problems that are generally more complex. The first chapter that investigates some of the results, chapter IV, discusses a technique developed to model unsteady aerodynamics with moving boundaries such as flapping winged flight. This will include mesh deformation and fluid dynamics theory needed to solve such a complex system. Chapter V will examine the numerical modeling of a combustive flow. A three dimensional single vane burner combustion chamber is numerically modeled. Species balance equations along with rates of reactions are introduced when modeling combustive flows and these expressions are discussed. A reaction mechanism is validated for use with in situ reheat simulations. Chapter VI compares numerical results with a laminar methane flame experiment to further investigate the capabilities of CFD to simulate a combustive flow. A new method of examining a combustive flow is introduced by looking at the solutions ability to satisfy the second law of thermodynamics. All laminar flame simulations are found to be in violation of the entropy inequality.

Le travail présenté s'intéresse au développement d'une méthodologie d'étude et d'optimisation des opérations de cristallisation en solution, par refroidissement en cuve agitée, de substances d'intérêt pharmaceutique présentant un polymorphisme cristallin. L'opération de cristallisation de référence, servant de support dans le cadre de ce travail, est l'opération de cristallisation de purification de l'Étiracetam crude. Il s'agit du step 3 dans la chaîne de production du Lévétiracetam, principe actif du Keppra, médicament commercialisé par la Société UCB. L'opération de cristallisation de référence se découpe en deux étapes. La première d'entre elles, l'étape de refroidissement, se caractérise par l'apparition de cristaux, dits de morphe II. La deuxième étape, l'étape de maturation, se caractérise par la transition polymorphe, dans des conditions isothermes, des cristaux de morphe II en des cristaux, dits de morphe I, correspondant à la forme cristallographique d'intérêt pharmaceutique. Cette étude est abordée via une approche multi-échelle et en suivant les lignes conductrices et méthodologies proposées par la Food and Drug Administration (FDA). Ainsi, dans un premier temps, nous nous intéressons à la caractérisation des deux formes cristallographiques, de leur milieu environnant et des interactions entre eux. Nous déterminons ainsi que les cristaux des deux formes cristallographiques sont des composés racémiques. Nous mettons également en évidence la nature énantiotrope du système polymorphe et déterminons la température de transition solide-solide séparant le domaine de stabilité thermique des deux morphes. Enfin, les courbes de solubilité, tant thermodynamiques que métastables, de ces derniers, dans du méthanol, sont déterminées. Toute cette étude se base sur l'analyse de résultats obtenus au moyen de nombreuses techniques analytiques et méthodes expérimentales de mesure aussi diversifiées les unes que les autres. Nous développons, par ailleurs, trois méthodes de caractérisation des substances cristallines, lesquelles permettent la détermination précise de l'enthalpie fusion via l'analyse des courbes DSC expérimentales, la détermination de la température de transition solide-solide via des essais de stabilité thermique des cristaux en suspension et enfin la détermination de la solubilité d'un morphe métastable via l'utilisation de la relation thermodynamique établie, sur base de la connaissance de la solubilité thermodynamique du morphe stable. Dans un deuxième temps, l'objectif est de comprendre les phénomènes physico-chimiques sous-jacents à l'opération de cristallisation, tant lors de l'étape de refroidissement que lors de celle de maturation. Cette étude s'effectue via le couplage entre des études fondamentales, des méthodes expérimentales et un modèle mathématique. L'originalité des expériences réalisées se trouve dans l'utilisation simultanée de quatre sondes de mesure afin de suivre l'évolution temporelle de divers paramètres-clés de l'opération de cristallisation : la température, la concentration de la solution, la distribution granulométrique ainsi que la forme cristallographique des cristaux en suspension. Le modèle mathématique développé, traduisant le comportement de la suspension, lors de l'étape de maturation et, plus particulièrement, au cours de la transition polymorphe, repose sur le couplage entre des équations de bilan de population pour les deux morphes et un bilan de matière pour le soluté. Cette étude nous permet essentiellement de proposer un mécanisme détaillé pour la transition polymorphe, laquelle, médiée par la solution, s'effectue par dissolution-recristallisation. Cette étude met, par ailleurs, en évidence l'importance considérable que joue la germination primaire sur la limitation des cinétiques des mécanismes sous-jacents à l'opération de cristallisation, que ce soit lors de l'étape de refroidissement au cours de laquelle apparaissent les premiers cristaux de morphe II, ou lors de l'étape de maturation au cours de laquelle se déroule la transition polymorphe entre les cristaux de morphe II, qui se dissolvent, en les cristaux de morphe I, qui croissent. Nous montrons également qu'une surface suffisante, dite critique, est nécessaire pour amorcer l'emballement des phénomènes. Enfin, dans un troisième temps, nous nous focalisons sur l'étude de l'influence des paramètres opératoires sur les temps caractéristiques des phénomènes physico-chimiques, et dès lors sur les distributions granulométriques des cristaux produits, et ce, en vue d'optimiser l'opération de cristallisation. Cette optimisation sous-entend essentiellement la diminution de la durée de l'opération de cristallisation tout en contrôlant la distribution granulométrique des tailles des cristaux de morphe I récupérés à l'issue de l'opération de cristallisation. Cette étude s'effectue via la réalisation d'un nombre limité d'expériences, dont les points expérimentaux sont définis par un plan d'expériences. Les résultats expérimentaux montrent, entre autres, qu'il existe de réelles compétitions, d'une part, entres les phénomènes physico-chimiques, la germination et la croissance, et d'autre part, entres les facteurs cinétiques et thermodynamiques. Ce travail s'intéresse également à deux études annexes. L'objectif de la première d'entre elles est de caractériser l'écoulement et le mélange générés par les trois systèmes d'agitation au sein desquels se déroule l'opération de cristallisation. Ceci s'effectue en comparant et discutant des résultats obtenus via deux approches complémentaires : des essais expérimentaux et des simulations numériques de mécanique des fluides. Ces résultats montrent que tant l'écoulement que les mécanismes de mélange qu'il conditionne sont très similaires d'un système d'agitation à l'autre. Par ailleurs, nous mettons en évidence, expérimentalement, que l'opération de cristallisation est peu dépendante des trois systèmes d'agitation utilisés, et, en particulier, de l'intensité de l'agitation. La deuxième étude se focalise sur le développement d'une méthode expérimentale originale permettant le suivi en-ligne de l'évolution temporelle, au cours d'opérations du Génie des Procédés, de la viscosité apparente des suspensions, telle qu'elle est définie par Metzner et Otto. L'idée est utiliser, comme lien univoque entre la puissance dissipée mesurée et la viscosité apparente, la relation Np-Re-Fr du système d'agitation utilisé. Cette méthode, appliquée à l'opération de cristallisation de référence, met en évidence l'intérêt du suivi en-ligne de la viscosité apparente de la suspension comme outil de contrôle. ************************************************** The work presented concerns a methodological development and optimization of batch cooling crystallizations in solution of pharmaceutical active ingredients showing a polymorphic behavior. The crystallization process that served as a reference in the context of our work, is the crystallization performed for the purification of crude Etiracetam. This crystallization is performed as a third step in the Leviteracetam production process. The latter is the active compound of Keppra, a drug developed and commercialized by UCB. The reference crystallization can be divided in two distinct steps. At first, a solution of solvent and compound is cooled until a maturation temperature is reached. This step is characterized by the appearance of morphe II crystals. In turn, the second step, the maturation, is characterized by a polymorphic transition under isothermal conditions, with morph II crystals transforming into morph I crystals, morphe I being the compound of pharmaceutical interest. The study presented is performed using a multi-level approach and following the guidelines and methods as suggested by the Food and Drug Administration (FDA). Firstly, we start by characterizing the two crystallographic forms, their environment, as well as their interactions. The results show that both crystallographic forms are racemic compounds that are enantiotropically related. The solid-solid transition temperature, separating the stability range of the two forms, is determined using various techniques. Finally, we describe both the thermodynamic, as well as meta-stable, solubility curves of both forms in methanol. A wide range of known analytical techniques and experimental methods has been used to perform the study presented here. We have furthermore developed three characterization methods for crystalline substances, that allow for a precise determination of the fusion enthalpy through analysis of experimental DSC curves, for a determination of the solid-solid transition temperature through thermal stability testing of different crystal suspensions, and finally for the determination of the solubility of the meta-stable morph through a thermodynamic relation, linking this solubility to that of the stable morph. In a second step, the objective was to understand the physico-chemical phenomena underlying the crystallization, both during cooling, as well as the final maturation. This goal is achieved by coupling fundamental studies, experimental work, and a mathematical model. The originality of the presented work is underlined by a simultaneous use of four online probes, allowing an instantaneous follow-up of the key parameters influencing the crystallization process: temperature, solution concentration, particle size distribution, as well as crystallographic form. The mathematical model developed is based on the coupling of population equations for both morphs with the mass balance equations. The final model translates the behavior of the suspension during the maturation step, and more particularly during the polymorphic transition. A detailed mechanism is presented, explaining the solution mediated polymorphic transformation, occurring through a dissolution-recrystallization process. The importance of the primary nucleation, and more specifically the limitation of the crystallization kinetics due to this nucleation, is highlighted. Both the appearance of morph II crystals during the cooling profile, as well as the polymorphic transformation during the maturation are limited by nucleation. We show that a critical, minimal surface is required for an observable acceleration of the underlying processes to occur. Finally, a third part of the study focuses on the study of the influence of operating parameters on the time needed for the different physico-chemical phenomena to occur, and hence on the particle size distribution of the obtained end product, with the ultimate goal of optimizing these operating conditions. The optimization mainly concerns the reduction of the total operating time, while keeping the particle size distribution of morph I crystals under control. This part of the study is performed through a limited number of experiments that are chosen using a well-defined experimental design. The experimental results show the competition between the physico-chemical nucleation and growth phenomena, as well as between kinetic and thermodynamic factors. In this work, we also discuss two annex studies. The goal of the first of these studies is to characterize the flow and the mixture created using three different stirring systems for the crystallization. We achieve this goal by comparing and discussing results from two complementary sources: experimental trials and numerical fluid mechanics simulations. These results show both the flow and mixture mechanisms to be comparable for the different stirring types studied. We, furthermore, confirm through experimental results that the crystallization process shows almost no dependence on the stirring device used, or on the stirring intensity. The second annex study focalizes on the development of an original experimental design allowing an online follow-up of the apparent viscosity of a suspension, as defined by Metzner and Otto. Central in the idea developed, is the use of the Np-Re-Fr relation as unique link between the dissipated power and the apparent viscosity. Applied to the reference crystallization, this method highlights the significance of an online follow-up of the suspension viscosity to control the crystallization.

The MVP is a special type of Electrical Submersible Pumps (ESPs) manufactured by Baker Hughes, model no. G470, and is capable of handling multiphase flow up to 70% Gas Volume Fraction (GVF). Flows at high GVF cause conventional ESPs to surge. However, the special design of the impeller blades of the MVP ESP enables it to handle higher GVF. Dynamic behavior of the multiphase flow is studied experimentally and theoretically for this pump for the first time. In this work, a Computational Fluid Dynamics (CFD) simulation of an entire pump and detailed experimental analysis are performed. Meshing and CFD simulations are performed using the commercially available software ANSYS Fluent. An experimental facility has been designed and constructed to test the pump at different operating conditions. The pump is modeled and tested at two speeds; 3300 and 3600 rpm, using air-water mixtures with GVFs of 0, 5, 10, 25, 32 and 35%. The flow loop is controlled to produce different suction pressures up to 300psi. Pump pressure head is used to validate the CFD model for both single and two phase flows. Single phase CFD model was validated at 100 psi inlet pressure, while two phase models were validated at 200 psi inlet pressure. CFD simulations can predict the behavior of the pump at different speeds, flow rates, GVFs, and inlet pressures. Different diffuser designs are studied and simulated to improve the multistage pump performance. Enhanced diffuser designs increased the pump pressure head to up to 3.2%.

No description available.

Fluid dynamics simulations generally require large computational recourses in form of computer power and time. There are different methods for simulating fluid flows that are more or less demanding, but also more or less accurate. Two well known computational methods are the Reynolds Averaged Navier-Stokes (RANS) and Large Eddy Simulation (LES). RANS computes the timeaveraged flow properties, while LES resolve the large structures (eddies) of the flow directly and model the small ones. Hybrid models are combinations of these two models which have been developed to improve the RANS solutions and shorten the simulation time compared to LES computations. One such model is the Scale Adaptive Simulation (SAS) model which uses the RANS model in steady flow regions, such as close to walls, and a LES like model in unsteady regions with large fluctuations. This study was done for evaluating the SAS model compared to Unsteady RANS (URANS) and LES and their performance compared to measurements from an engineering point of view. This was done by running simulations on two different test cases, one external and one internal flow situation. The first one was flow around a wall-mounted cylinder and the second one was flow through an aorta with a coarctation in the descending aorta. The first test case was used to thoroughly evaluate the SAS model by running many simulations with URANS, SAS and LES with different element types, element sizes and flow parameters. The element types that have been analyzed are; tetrahedral, hexahedral and polyhedral. The results were compared with experiments done by Sumner et al. [7, 8, 9, 10]. The second test case was used for evaluating the SAS model even further on another flow situation. For this test case, only two SAS simulations were performed on two different grids; a structured hexahedral and an unstructured polyhedral. These results were compared with Magnetic Resonance Imaging (MRI) measurements obtained from Linköping University. No conclusion of which one of the simulated cases gives the best overall agreement with experimental results could be concluded from the obtained results. The best prediction of the drag coefficient for the cylinder was obtained for the coarsest polyhedral mesh that was run with LES, with the disagreement 0.4 percent. The best prediction of the Strouhal number was obtained for a URANS simulation performed on the coarsest mesh with an improved grid close to the cylinder surface, generating less than one, with a disagreement of 3 percent compared to measurements. For the meshes used, it was found that the polyhedral mesh gave the best overall results and the tetrahedral mesh gave the worst results for the cylinder case. For the aorta case the SAS model produced velocity components that had acceptable agreement with the MRI-measurements, but gave very poor results for the turbulent kinetic energy. The main conclusion of this thesis was that the SAS model performed better than URANS, but took longer time to compute simulations than LES, which was the model that generated the best overall results.

In this thesis work the feasibility of using open source OpenFOAM software as a solver part for Computation Fluid Dynamics in Volvo Technology is studied. Since the structure of the case in OpenFOAM is rather complex, one of the main purposes of this thesis work was also to make the process of using OpenFOAM as user-friendly as possible. The general conclusion that can be drawn from this work is that a very streamlined workflow can be, and has been, designed and created.

Premixed and partially-premixed combustion and ollutant emissions in full-scale gas turbines has been numerically investigated using a massively-parallel Large-Eddy Simulation Combustion Dynamics Model. Through the use of a flamelet library approach, it was observed that CO (Carbon Oxide) and NO (Nitric Oxide) emission can be predicted and match experimental results. The prediction of the CO emission trend is shown to be possible if the influence of the formation of UHC (Unburnt HydroCarbons) via flame extinction is taken into account. Simulations were repeated with two different combustion approach: the G-equation model and the Linear-Eddy Mixing (LEM) Model. Results are similar for these two sets of numerical simulations. The LEM model was used to simulate flame extinction and flame lift-off in a dump combustion chamber. The LEM model is compared to the G-equation model and it was found that the LEM model is more versatile than the G-equation model with regard to accurate simulation of flame propagation in all turbulent premixed combustion regimes. With the addition of heat losses, flame extinction was observed for low equivalence ratio. Numerical simulation of flame propagation with transient inflow conditions were also carried out and demonstrated the ability of the LEM model to accurately simulate flame propagation in the case of a partially-premixed system. In all simulations where flame extinction and flame lift-off was simulated, release of unburnt fuel in the post-flame region through flame extinction was not observed.

An understanding of gas-liquid two-phase interactions, aerosol particle deposition, and heat transfer is needed. Computational Fluid Dynamics (CFD) is becoming a powerful tool to predict aerosol behavior for related design work. In this study, FLUENT 6 is used to analyze the performance of aerosol sampling and concentration devices including inlet components (impactors), cyclones, and virtual impactors. The ω − k model was used to predict particle behavior in Inline Cone Impactor (ICI) and Jet-in-Well impactor (JIW). Simulation provided excellent agreement with experimental test results for a compact ICI. In the JIW, compound impaction is shown to cause the device to have a smaller cutpoint Stokes number than the single impaction unit. The size ratio of the well-to-jet was analyzed to find its influence on the total and side collections. Simulation is used to analyze liquid film, flow structure, particle collection, pressure drop, and heating requirements for a bioaerosol sampling cyclone. A volume of fluid model is used to predict water film in an earlier cyclone. A shell-volume is developed to simulate thin liquid film in large device. For the upgraded version cyclone, simulation is verified to successfully predict cutpoint and pressure drop. A narrowing-jet is shown to describe the flow evolution inside the axial flow cyclone. Turbulent heat transfer coefficients and surface temperatures are analyzed and heaters are designed for this cyclone. A double-outlet cyclone was designed and its pressure drop decreased about 25%, compared with a single-outlet cyclone. A scaled-down 100 L/min cyclone was also designed and tested based on the 1250 L/min unit. CFD is used to design a Circumferential Slot Virtual Impactor (CSVI) which is used for concentration of bioaerosol particles. Simulations showed a 3-D unstable flow inside an earlier version CSVI, which could explain acoustic noise and particle loss observed in the experiment. A smaller CSVI unit was designed using simulation and its flow was shown to be stable. CFD was then used to analyze the wake flow downstream of the posts to reduce particle losses and eliminate flow instabilities caused by wakes. A successful solution, moving the posts outside was developed by the use of CFD.

Computational Fluid Dynamics (CFD) is a field that is constantly advancing. Its advances in terms of capabilities are a result of new theories, faster computers, and new numerical methods. In this thesis, advances in the computational fluid dynamic modeling of moving bodies and combustive flows are investigated. Thus, the basic theory behind CFD is being extended to solve a new class of problems that are generally more complex. The first chapter that investigates some of the results, chapter IV, discusses a technique developed to model unsteady aerodynamics with moving boundaries such as flapping winged flight. This will include mesh deformation and fluid dynamics theory needed to solve such a complex system. Chapter V will examine the numerical modeling of a combustive flow. A three dimensional single vane burner combustion chamber is numerically modeled. Species balance equations along with rates of reactions are introduced when modeling combustive flows and these expressions are discussed. A reaction mechanism is validated for use with in situ reheat simulations. Chapter VI compares numerical results with a laminar methane flame experiment to further investigate the capabilities of CFD to simulate a combustive flow. A new method of examining a combustive flow is introduced by looking at the solutions ability to satisfy the second law of thermodynamics. All laminar flame simulations are found to be in violation of the entropy inequality.

This work examines the ability of two dimensional CFD models to predict the unsteady flow downstream of a cylinder, with and without fairing, in uniform flow. PIV measurements of the flow-field downstream of the cylinder and fairing in uniform flow are first presented. “Slices” of the flow at several locations along the cylinder are compared to show the variation of the flow in the cross-stream direction. Then the PIV flow is compared with RANS and LES simulations of the flow. Velocity time histories are compared and hydrodynamic coefficients are discussed. In a general sense, two dimensional CFD can give a functional approximation of the unsteady flow field downstream of the cylinder or fairing. / text